>he doesn't actively avoid konbini staffed by non-japanese
3 months ago
Anonymous
Very hard to do in the major metros. Easy to do out here but becoming more difficult every year
3 months ago
Anonymous
It's not that hard to avoid indian ones.
Harder once you start including koreans and chinese.
3 months ago
Anonymous
Starting to see Muslims all over the place even in the fricking inaka now. You look at any social media involving Japan and the comments are Muslims claiming Japan needs Islam, we will take over Japan, etc. everyone thinks this country is some based anti immigrant heaven but thanks to student and work visas it is becoming anything but
3 months ago
Anonymous
if you need gaijin deportation volunteers, I will fly to japan tomorrow
3 months ago
Anonymous
I'm a gaijin myself so I'd be hypocritical if I said let's go, but at least I try to integrate. The Muslims and central Asians don't even fricking bother
3 months ago
Anonymous
integrated gaijin are fine
but gaijin that go to japan and try to make it not-japan, should be all deported
just like this
3 months ago
Anonymous
Good luck, the Japanese are way too passive.
3 months ago
Anonymous
Being hypocritical is fine. That's how countries are saved from this kind of shit. >b-but muh principles
For 99% of the people pushing for this shit it comes naturally to them to abuse principles like yours.
Yep. Version control, CI/CD, Agile methodology (or even waterfall), design patterns, solid principles etc. are gigantic wastes of time and money that create multiple bullshit jobs and sap the actual programmers' productivity. Corporations have ruined programming, and AI tools will only make it worse.
The best form of programming is 5-8 guys writing what code they want in one place each and then patching it together at the end.
>Agile methodology
I 100% agree with this post
t. worked on a tech project for an organisation that wanted to try agile for the first time on one of the biggest national reforms to date. Pissed 1.5 billion up the wall and the project got shitcanned before anything even made it to release.
I've had better experiences with agile than waterfall, but these are always going to depend on the scale of the project I feel. The waterfall project was just massive, so it was hard to course correct some things.
It depends on the project. I've seen the goods of it making development very easy and straight forward but also the bads where each call was just a huge, paid timewaste with way too many of them. It really depends on how "well" they execute it.
Fair point, but I feel that agile is easier to frick up than most because it's so 'freeform' and if you're not tight with actual sprint outcomes and what the focus is, you're bound to piss away time and resources on nonsense.
Also, agile always seems to have a bunch of fricking zealots who preach about how good it is, without articualating the inherent flaws
>team leader
You can't imagine how much I hate the team "leaders" of the project I'm working in. They are glorified, overpaid seat heaters, because that's all they do. They probably work less than twenty minutes a day.
I work as a lead dev and would have fired you on the spot for saying that version control, CI/CD, design patterns and especially SOLID are redundant.
Yes, there is circlejerky bullshit that nobody should ever care about, but your job as a programmer is to separate the wheat from the chaff.
I get your point on agile, but the rest of them I won't concede. Your code is useless trash if you don't apply SOLID. Sorry, but it's the truth.
>solid defense force
Give it a few years and you finally start to understand what good programming is like. You're not fooling anyone here with your sv web moron opinions
SOLID can make it take more time to add a feature where you have to break everything into more components than neccessary to satisfy the single use princple and create a bunch of unneccessary interfaces to satisfy the dependency inversion principle
Oh no, I have to create two interfaces to communicate between components. The 5 seconds it takes me to create those is definitely such an amazing loss of efficiency that we shouldn't have interfaces at all
Get fricked, next programmer who wants to figure out what functions already exist to access this component, dig through the whole implementation to find out
3 months ago
Anonymous
Interfaces take time to write, take time to read and make your code run slower
The costs add up
3 months ago
Anonymous
>take time to read
lel
"oh man I really wonder which service BusinessObjectServiceInterface.h uses, I guess I'll have to go check the implementation! stupid interfaces ugh why can't I just write it all in one class"
Try writing code with more than 2 people sometime, or a program you have to support for longer than whenever your college course ends
People can definitely take the principles way too far and get way too anal and abstract about what it means for a class to have a "single responsibility" or what is a "high level/low level module".
They are broadly useful tools for when the logic of a program starts to become too complex/entangled and you want to isolate it to prevent unnecessary coupling.
Interfaces take time to write, take time to read and make your code run slower
The costs add up
>Interfaces take time to write
True >take time to read
They save significantly more time by hiding implementation details and decoupling isolated parts of the program >make your code run slower
If interfaces are causing a measurable performance hit to your program you're doing something catastrophically wrong.
Generally the best reason to not use them is when the logic of your program is too simple to warrant it or the area where you're working doesn't change often.
3 months ago
Anonymous
>by hiding implementation details
This is for glue coders.
For actual programmers that's pretty large negative. Particularly once you try to step through it and you have five layers of indirection and can't keep watch windows of out of bound structs.
It really depends what your programming. If you have a solid vision for the program you don't need any of that. Those aspects of the job are to accomodate constant changes because the people responsible for the functionalities and features can't make up their mind, or have no idea, what they want.
frick your patterns, frick SOLID and frick all other superficial bullshit, every SOLID and pattern worshipping moron I've encountered wrote the shittiest code imaginable and I've worked with hundredths of people over 15 years of programming
>I have no idea about the actual work I supervise but if a worker with actual hands-on exp disagrees with me I'd rather lose the worker than give up fraction of my israelite-given authority
The only person I ever had mention SOLID at my workplace was a system architect, great at his specific job and good at being able to look at the important parts of the problem while being able to adapt to new constraints very easily, but not a programmer, never claimed to be one. His adherence to it was indication that it (like so many similar dogmas) is born out of trying to fit programming into a box. In reality it always depends.
A system architect's "job" is knowing the common patterns at the top level and being able to apply them without missing the forest for the trees. SOLID makes sense to them for this reason; it shouldn't make sense to programmers for the exact same reason. Similar to OOP though OOP's complicated by the fact that its roots in smalltalk describe something completely different to what stroustrupp seems to have taken away from it.
Using CI/CD in the same sentence as SOLID is just bizarre to me. Is it useful having a build, test, and deploy process you can kick off with one command without any fiddling with configs? Hell yeah. Is it useful to have it able to run somewhere other than the dev's hardware? Yeah, a lot of the time. Is it useful having it automatically triggered by various events? Maybe, really depends on what you're building and how complicated your deployments are. Should YOU have that(?) for YOUR specific project? Maybe. How long is a piece of string?
What is even version control? To approach this pragmatically; sticking your code in a git repo is always useful even if you just use it to streamline your backup process. I've seen people with IDE plugins that push code back to the upstream repo every time they save a file. Its like those videos of third worlders using a jackhammer like a chisel without even plugging it in, completely bizarre to see, but even so they get value out of it. I find it hard to imagine why you wouldn't use git. Its not like it costs anything.
What a bizarre world view where architects and programmers have opposed philosophies that clash because I guess you're talking about junior programmers that only wanna write all the code while ignoring patterns to make it more robust in case of future change?
>junior programmers that only wanna write all the code while ignoring patterns to make it more robust in case of future change?
I'm not sure in this sentence whether the "to make it more robust" is being ascribed to the junior ignoring the pattern, or the pattern itself.
what I really hate is when I'm told to take durable robust code and make it fragile, because it's "simpler" or because they dont think the code will be used in other ways
3 months ago
Anonymous
Well I hate having to explain that it'd be faster to just write the 500 lines of code myself than to write 2000 lines of documentation so the crack team in south vietnam can do it a fifth as fast with five times the bugs but we can't have everything, anon.
What were they actually asking for?
version control is nice to have, but you have to spend a good amount of time to keep your history clean and will never look at it again.
You can just make your application until you have a stable release and only then start using VC
you add a new feature, its not finished yet but you want to test if its displayed properly in the gui.
You comment the call to handler to make sure the code compiles.
You also notice a minor mistake in another part of the code.
Now you have >an unfinished feature >a temporary change >an unrelated fix
now what do you do,
commit? extract the fix and commit separately, rebase that shit later.
Why would you commit an unfinished feature and some debug shit? You're not supposed to just commit everything at the end of the day.
3 months ago
Anonymous
>commit == push >push == push into production
Here you go https://en.wikipedia.org/wiki/Branching_(version_control)
3 months ago
Anonymous
they really should have named them more logically
savelocal
savetoserver
much simpler to understand
3 months ago
Anonymous
Don't even start about GIT naming, I'm still mad we aren't allowed to call the main branch "master" anymore
3 months ago
Anonymous
OH YEAH I totally forgot about that
[...]
it's dumb to use solid as a solo dev, no one else will use your code, you can modify everything whenever you want and you don't need a dozen unnecessary abstactions
for games, its cool to still have every version of your game that you can build anytime and show others
I really like looking at game prototypes
unfortunately most have been lost because they didn't use source control
3 months ago
Anonymous
>git is github
classic
3 months ago
Anonymous
This. >savetolocal >savetoserver >pushlocaltoserver >pushservertolocal >forcepushservertolocal
Things would be massively easier if commands were named after what they did, instead of being "fast" to write.
3 months ago
Anonymous
The cool thing about bash is you can make your own mnemonics that is a miserable word to spell
3 months ago
Anonymous
>I just commit when everythings done
if you use a modern CI/CD pipeline you are forced to commit to start the build pipeline
I only start using VC when there's big changes
or I bring in another coder who may muck things up
it is nice to be able to regenerate an old build
imagine if we had all of ocarina of time's old code and could play the old builds at any point in development cycle, imagine how much we would learn, instead of just the final result and a few snapshots
even a simple stage all -> commit -> occasionally push to remote routine is better than no version control at all, because it lets you backtrack if you or your software ever frick something up >but you have to spend a good amount of time to keep your history clean and will never look at it again.
???????????????
Patching it together just by eyeballing it? Git provides tools to show differences when debasing in the form of merge conflicts. Catching everything manually is impossible and at the very least would take tediously long, unless it's a toy project with barely any loc
I work in company that deals with natural gas and electricity, for programmer to not know GIT at this level would be a joke, they'd be fired instantly for being a moron.
how is a versioning system where you cannot keep a local copy of your work and update it better?
oh right we're on Gankereddit where it's hip to be contrarian at all times
It's fricking not. I worked almost 5 years with SVN before migrating to Git and it's fricking night and day. I'd never go back and I would question the sanity of anyone suggesting to do so
yep, but when you have 200-500 people working on a project, source control is a must
however git is a convoluted pile of shit that has probably caused more problems with its ambiguous commands which prevent people from using it with 100% confidence
I've had better experiences with agile than waterfall, but these are always going to depend on the scale of the project I feel. The waterfall project was just massive, so it was hard to course correct some things.
it really depends on the project's design and how flexible and non moronic the devs are
>git is a convoluted pile of shit that has probably caused more problems with its ambiguous commands which prevent people from using it with 100% confidence
Look at this loser, he doesn't fricking know how to use git
And most of the time you won't need that, you work on local dev branch, you merge into remote main branch, you pull from remote. For 99% of cases it's all person needs
>Oh weird I did but I guess jira's just buggered again I'll put it in later today once I get this working >>Do nothing and say the same thing tomorrow
They'll get it eventually. Or they wont.
Programming died when all the non programmers found out how much moner programmers made. After 10 years of everyone learning to code being mainstream(thanks obama), tech is just another corporated beaurocracy made up more normalgays. Autistic lesser code gods cant even get jobs in current tech because they can’t pass the interview process while the the dead weight extraverts can easily nail HR interview process
Companies that overvalue the interview process aren't worth working for in the first place. If a company is lorded over by HR karens do you really think you're going to have fun working there? They've done you a favour.
This is why I consider the modern environment of game development to be so fricking dumb. The impression I grew up with with people who were making huge changes in the industry came from roots crammed up in a garage / basement working their ass off because the process of coding is fun and playing their game as they go to iron out the bugs.
git is great though.
i used it for everything related to text documents and not just source code.
honestly i dont want to go back to a time when i didnt know what git was.
i personally like to use lazygit.
>Version control
moron >CI/CD
moron >Agile methodology
Partly agree. It's a broadly useful tool but people get way too ideological about it and don't actually check to see if it's making them more productive. >or even waterfall
That's the alternative to agile moron >design patterns
You don't need to know them (especially not if you're working on your own) but they are good and recognisable solutions to common problems
They absolutely shouldn't be something you focus your time on as a beginner but for big projects they're generally worth understanding. >solid principles
moron
>NOOO SHOW ME THE HECKIN INFOGRAPHS AND METRICS THAT PROVE YOUR POINT
You have corporate brainrot. Just because a tool is useful doesn't mean it's beyond criticism.
Disagree with the first line, agree with the second line.
There's a reason that older games felt more creative/soulful, and it's not simply because there was more uncharted territory. It's because the teams were tight-knit groups where everyone was a generalist instead of a specialist. Games were more cohesive because the developers worked on multiple elements and linked them all together. Whereas now there's hundreds of devs on a team and they all do one specific thing. Nowadays you have to play indie games to get that kind of tight-knit design.
Smaller teams also have a much better sense of what how they can steer the game and what's possible than large teams which lets them be much more dynamic and innovative.
If you get half way into a $300mil game and realise it isn't fun there's basically nothing you can do about it but with a small dev team you can rip it apart completely and only keep what works
Same feeling I got after watching the dev documentaries for Halo 1,2 and 3
In the first two games you really get the feeling it's just a bunch of friends making what they want. They're doing social activities together, everyone is pitching in on everything, programmers have to do voices and storyboards, the composer has to code and help writing the story, the lead dev is adding easter eggs in his spare time.
In halo 3 things are getting a bit more business like, but still relatively casual, but then for Halo 4 it's all corporate, every desk is completely clean, everyone is at their desk typing quietly
A few months ago they fired all the company's scrum masters
I have to do way less work and the projects still aren't deploying on time so it's a net positive for me
version control is just common sense
CI/CD are basically bullshit. but build/deployment automation is good, it just doesn't need to be "continuous". testing is a mixed bag and people can get completely obsessed with it over actually writing code.in the right dose its good.
but yea the rest you can throw into the garbage
Agile is actually a systemic manifestation of demonic forces and corrupts the soul to a near irreparable degree
I am not kidding it legitimately will create evil human beings
In concept, the original version of Agile is actually good. Because it essential was "isolate the programmers in a room, divide the work into little chunks, and don't let any business people disturb them"
the chunking the work people still do, but the rest is completely absent
in practice, agile teams create lots of devs siloing themselves, not working as a team, and hoarding high impact work for themselves
it also unintentionally incentivizes managers/product managers backchanneling new work into the sprint instead of going through the actual process
to go through the process would require them to admit where they fricked up, so they try to slip shit in discretely, or rely on buddy relationships with particular programmers to add extra features
agile basically guarantees that every team devolves into a malthusian dystopia
Agile makes it so managers don’t have to do any planning (which is supposed to be their job) and make the developers do it instead. It’s supposed to give developers “ownership of their tasks” or something. I had to spend a whole day every week in meeting after meeting doing sprint planning, sprint retrospectives, backlog refinement, etc
>it just doesn't need to be "continuous"
it means testers can play a current build and spot problems sooner instead of waiting for the build at the end of the day or however often
and not have to deal with things that have already been fixed
also to push bugfixes to players, it's still essential you can do that asap without much effort before gamebreaking bugs break everything
like everything in the field, it starts by making sense for specific use cases and then incompetent people start copying it without understanding it properly because they think it's some sort of recipe for success
So what this Palworld Phenomenon is is basically what Vidya used to be like 2006 and before where it's was a dozen guys just making their own game and then releasing it to the normalgays instead of making vidya for normalgays
Nobody in their right mind would go against Git. When even the suckless homosexuals use Git you know it's the absolute bare minimum.
Git sucks pretty bad in many respects even for basic stuff and relies on you using tooling to circumvent it.
It's just that so do most of the alternatives.
Personally i'd like something like Fossil but it's also bad with binary files like Git is.
>Blizzard fired
Microsoft fired, blizzard is not a unique, non-microsoft entity anymore and should never be discussed as if they are on this board. Blizzard makes no longer decisions for themselves, whatever is left will just take orders from the largest tech company on the planet.
It's just a funny development anecdote. I don't think it's anons necessarily saying version control is a bad idea, just that they are amused by the apparent inexperience Palworld developers had before getting started. The value of version of control is something you learn with experience, rather than being a strict requirement to make anything.
When you start digging, there's always plenty of funny or off-the-wall stories various developers may have. Often times such stories involve source code or various development practices, where in this case it's version control.
It depends on the project. I've seen the goods of it making development very easy and straight forward but also the bads where each call was just a huge, paid timewaste with way too many of them. It really depends on how "well" they execute it.
I program legacy software at an investment company. My big project after starting was converting the project to a new version that uses text files rather than a monolithic binary file.
Since that finished, I've slowly been teaching the 85 year old guy on our team how to use git, he's almost got it
Tell me the flashdrive story again grampa, it was a rollicker. I liked the part where you used flash drives but most people don't use flash drives. that was the craziest part.
>didnt know Git
Even if they were CLI purists and refused to use something like GH Desktop, they'd only be using the same handful of commands most of the time unless something went wrong. How hard is it to learn just that?
According to the senior dev, he strongly prefers Perforce because of its Unreal engine integration, but Perforce was too expensive for an indie company. So he suggested using SVN instead of Git, which the CEO considered to be outdated and a downgrade from Git. Still the CEO felt he had to take a blind dive and trust the senior dev as it is hard to find someone on his level interested in working for his indie company.
You only need git for large products with multiple teams that work relatively independently from each other.
Also nice projection with your fizbuzz, newbie.
you should have version control for anything where changes might need to be undone or just having undelete protection. or conversely if you know you're going to do something experimental and don't want to waste time commenting and uncommenting code to control which parts are going to run between the production code and the experiment. you can just make a throwaway branch and easily merge updates into it, or not.
you've just never programmed anything worthwhile on your own.
>you've just never programmed anything worthwhile on your own.
I have, and I literally only use git to transfer code between my computers. You're projecting hard
>larp
That's exactly what it is. The interviews are all bullshit >oh actually we're dumb as frick lol
Like no you're not. You hired a bunch of Nintendo Interns. You know EXACTLY what you are doing.
there's a bunch of git frontends, which is absolutely neccesary for comparing files, but it doesn't change the fact that the commands are not totally clear, and the frontends still use the language of git commands instead of clearer english
This moron saw that stupid meme discord screenshot huh? Yes, they did use git but switched over to SVN >To be honest, I was a little hesitant to migrate the engine in the first place because companies that use svn these days have a legacy image. Compared to that, anything like a version control system is fine. Fully trusting his words, I also migrated my version control system from git to svn.
https://note-com.translate.goog/pocketpair/n/n54f674cccc40?_x_tr_sl=ja&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=wapp&_x_tr_hist=true#33883a37-8fdb-4f0d-ab06-853c8aa8be4a
Fricking mouthbreathing morons i swear to god. See an image on the internet and they think it's 100% authentic. Fricking monkeys.
>SVN
Just like Blow and other game devs.
I'm not entirely sure but I think it's because like Perforce it can handle art assets which Git more or less implodes on.
that pic isnt even correct because the holocaust started midway through 1941 and ended in january 1945 when the camps were liberated. So 3 and a half years, 4698 israelites per day.
that pic isnt even correct because the holocaust started midway through 1941 and ended in january 1945 when the camps were liberated. So 3 and a half years, 4698 israelites per day.
Israel is killing hundreds per day in the most in efficient way possible.
4k per day sounds extremely reasonable.
They're destroying bodies, desecrating graves, and blowing up records of birth certificates etc. to facilitate genocide denial.
So I guess you're a fan of that kind of thing?
>shitto, fuccu, we needo programmeru, Shinji-san! >donto worryu, Kaoruo-dono, i foundu sales clerku, he fold programmingu code over 9000 timesu >game sells millions
This shit is beyond comedy. What are these western fricks doing if tatami-san over there on the island has this type of performance with hiring hobbyists?
Not using Git is amateurish as frick. Only people who never worked developing software before would thinks it's a good idea not using it. Forget Linux; Git IS Linus Torvald's greatest accomplishment.
friendly reminder that most software projects written in the 80s and 90s were done without source control, and they still end up being some of the best things ever made for computers
Try testing program to work both on 10 Android versions and Iphones while running code on PC without Virtual Machines. In 80s and 90s barely anyone used tech and there was no need for it
thats the right way to use VMs
and I guess to segregate your dev environment from the rest of the system, cause then you don't have to uninstall a bunch of stuff and can easily restore things from snapshots if updating a tool broke things
but it is just one of many things in modern development that makes things slower
While you've been arguing about version control. Lil' Taro has finished his new character controllers in UE4 blueprints, sent the files to his coworker who merged them all in the current build of the game.
They won in that it sold a lot of copies. As a game it's a barely stable janky pile of doodoo with moronic balance and some of the worst AI pathing and behavior I have ever seen. The amount of times an enemy has just stared ahead at me without moving or attacking is ludicrous, it's a complete failure of programming that is only excused because early access.
>you can download your games whenever you want as long as you want as long as you dont say Black person Black person Black person
This upsets Ganker because of course it does.
but do you know how to rebase a repo? switch branches, revert changes to files or change a repo location
WITHOUT GOOGLE?
and because the syntax is so convoluted, it's just an invitation for mistakes to happen that you won't realise for a while
also many git ui programs are still bad and it's not clear if git is actually doing what you want to do
there's just so much noise and it's so verbose
>keep it simple
Literally impossible with git.
Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
3 months ago
Anonymous
Couldn't you just do an interactive rebase (to the hash of the first commit), choose edit on the first commit, then change the message to what you want it, finish the rebase, then force push?
3 months ago
Anonymous
Unfortunately not. Editing the initial commit message is only possible through commands.
It gets even worse when trying to edit the author.
git commit —amend
BZZZZZZT!
See pic related.
3 months ago
Anonymous
Googled it, you're right, my first idea wouldn't work, I'd have to use git rebase -i -root instead.
Still only had to check one stack overflow thread for that one though.
3 months ago
Anonymous
>edit
that's what "reword" is for.
3 months ago
Anonymous
git commit —amend
3 months ago
Anonymous
No, the first commit, not the latest commit.
3 months ago
Anonymous
>Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
just make a new repo lmao
3 months ago
Anonymous
Unironically your best option if you don't have five years worth of experience in git.
3 months ago
Anonymous
if you have a five year old repo you have no reason to change the first commit message
it's a moronic scenario you made up in your head, literally just make a new repo if you made a typo in the first commit
3 months ago
Anonymous
>if you have a five year old repo you have no reason to change the first commit message
The email address I used when authoring that message is no longer in use.
I wish to change it to my new email address. Why is this not a valid use case according to you?
3 months ago
Anonymous
you're trying to revise history, the commit is five years old
i know your old email address has your deadname in it but you're just gonna have to live with it
3 months ago
Anonymous
>but you're just gonna have to live with it
Not a valid argument. Thank you for conceding.
3 months ago
Anonymous
I was going to say it's not that difficult, that one quick google search gave me the solution, but then I remembered that I have five years worth of experience in git, so that's not really a counterargument.
3 months ago
Anonymous
if you have a five year old repo you have no reason to change the first commit message
it's a moronic scenario you made up in your head, literally just make a new repo if you made a typo in the first commit
Unironically your best option if you don't have five years worth of experience in git.
>Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
just make a new repo lmao
>keep it simple
Literally impossible with git.
Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
Have morons not heard of
git rebase --root
I swear Dunning-Krugers here...
3 months ago
Anonymous
Of course I haven't heard of git rebase --root. How often do you think I need to change the initial commit a repo?
3 months ago
Anonymous
Not often I imagine, but there's this thing called - manual, worth reading.
3 months ago
Anonymous
I'm not going to read through the git manual for fun.
The manual is for looking up specific thingsI don't know how to do.
3 months ago
Anonymous
that's not the only thing it's good for, it also tells what you can do and what workflows it applies to.
3 months ago
Anonymous
multiple times to maintain consistency
>git rebase --root
How the frick is anyone new to git supposed to know what any of that means?
Seriously, why can't we just do: >git edit commit [commit id] -author "[email protected]" -message "Initial commit"
on any commit?
[...] >worth reading
lmao, absolutely not.
You remind me of those unixtards who create thousands of manual pages and then get angry that no one wants to read them to learn how to do the simplest thing.
just study?
3 months ago
Anonymous
>just study?
If I want to do a single small thing, I am not going to read an entire book on the subject.
Instead, I will use something else. It's really as simple as that.
3 months ago
Anonymous
>git rebase --root
How the frick is anyone new to git supposed to know what any of that means?
Seriously, why can't we just do: >git edit commit [commit id] -author "[email protected]" -message "Initial commit"
on any commit?
Not often I imagine, but there's this thing called - manual, worth reading.
>worth reading
lmao, absolutely not.
You remind me of those unixtards who create thousands of manual pages and then get angry that no one wants to read them to learn how to do the simplest thing.
3 months ago
Anonymous
>unixtards
I actually am unixtard, but I don't get angry and I don't write manuals.
3 months ago
Anonymous
Didn't mean to imply all unixgays are the same unixtards who write manuals for fun.
I could've worded it better, my bad.
3 months ago
Anonymous
I know what you mean, considering nowadays most manuals are either complete shit just tells you visit this and this site instead. But generally manuals are useful and when use new software or library and need to look something up first thing I do is check whether there's a manual for it. In unix systems usually there's.
My point being, that if you're software developer - manual should be your best friend.
3 months ago
Anonymous
>but you're just gonna have to live with it
Not a valid argument. Thank you for conceding.
>Seriously, why can't we just do: >git edit commit [commit id] -author "[email protected]" -message "Initial commit" >on any commit?
git hashes depend on all preceding commits in the tree
if you amend a five year old commit, all the following commits are going to get new hashes, which will frick up any script and external link that point to a specific commit
git is trying to prevent you from doing something really fricking stupid
3 months ago
Anonymous
>git hashes depend on all preceding commits in the tree
What? Why? What could possibly be the justification for that moronic decision?
3 months ago
Anonymous
The hash is supposed to be a unique identifier of the commit. If it didn't depend on the history, then two commits with the same content but different histories would have equal hashes.
3 months ago
Anonymous
integrity of the source tree, the current HEAD is a product of every commit in the tree
if this wasn't the case, your link to 4e203fc today could point to different contents tomorrow
3 months ago
Anonymous
integrity of the source tree, the current HEAD is a product of every commit in the tree
if this wasn't the case, your link to 4e203fc today could point to different contents tomorrow
I don't see why you'd just not add a hidden identifier to the hash that can't be amended instead like a counter or a date added. (and treat merges as new ones)
As is the most obvious thing is prevented by design. I guess it makes sneaky bullshit harder in the particular case of a huge open source platform project like Linux but for most everyone else it seems deeply silly.
3 months ago
Anonymous
Don't forget that git was developed specifically for linux kernel source control, it was just later on adopted by normies.
And in kernel development history and tree integrity is extremely important.
3 months ago
Anonymous
>Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
just make a new repo lmao
No, the first commit, not the latest commit.
>git notes add `git log --all --pretty=%h | tail -1` -m "message"
git uses directed acyclic graphs.
Of course you can't just change commits without breaking the whole structure.
It is indeed possible though either through rebasing (will frick over your whole revision)
or by doing a hack by creating a commit and using git replace
when there's a cheat sheet to basically translate git into something every english speaker understands, it makes you wonder why didn't they name the commands better
nobody knows it without google at first.
you just use google until you internalize it but its not like you ever sit in a black box with no internet access to look shit up.
what im saying is even people who love git often dont know how to use git without googling
when you have 2000 employees all using git, people are going to end up making mistakes with git that end up eating time
it's easy to make mistakes with git
nta, but I heard about a game jam that took place on a cross-country train ride where the Internet kept going out, and so all the indie devs were running around panicked asking each other basic programming questions.
It honestly sounded kind of cool, having to do without google for a couple days.
>google how to revert in git >simplest is do x >careful when doing x. it fricked up mine. >the correct way is doing y >y is a convoluted chain of commands
Jesus, feels like being a git manager is a full time job.
>But I really have not been involved. People like Greg and the actual graphics driver guys have been in much more direct contact with Valve. I think it's great to see gaming on Linux, but at the same time, I'm personally not really much of a gamer. >I'm personally not really much of a gamer. >I'm personally not really much of a gamer. >I'm personally not really much of a gamer.
why the frick doesn't linus torvalds play videogames?
what does he think computers are for?
Not using or understanding version control will slow you down massively. Even for projects you're working on individually it's still better than nothing because it makes regression much easier and bugs much faster to resolve.
Blow also doesn't wank off programming as an art moron, he's just as stupid as you and thinks "programming should solve real problems" which to him means use literally 0 abstraction ever until you're forced to because he's too stupid to understand when it's valuable and when it isn't
And if your hard drive corrupts?
Point is it’s just safer to store it on a server vs having it locally.
Development is doable without it but it’s a nice tool to have.
And if your hard drive corrupts?
Point is it’s just safer to store it on a server vs having it locally.
Development is doable without it but it’s a nice tool to have.
Do you test after every single feature?
I'm guessing you're not that forward thinking or else you'd already use version control
So now imagine you implement a batch of features and then you find out that one of them introduces a critical bug.
But oops because you're a fricking moron and your development versioning is a straight line you can't revert that change without also reverting all changes which came after it.
So now you either roll back 2 weeks worth of work or spend half a week trying to find and fix everything that you changed.
Oh but frick now your hard drive with the 3 months of work between august and november is corrupted and there's no way to ever roll back to that patch.
And there's a community patch available to fix the long load times but you have literally no idea what it changed because your tooling of "just copy the files bro" can't identify the diff
There's literally no reason not to use version control. It is just a better and easier alternative to copy and pasting files that guarantees you always know where your history is and what each patch changed.
If I do a major refactor then I make a copy of the old code incase I need to revert by copy and pasting the files into a new folder
Why bother using a version control system
Code needs to be refactored, it's just a fact of life
3 months ago
Anonymous
not all the time
you may need to tidy things up after your first draft so you can easily make changes later, but some people seem to get dopamine rushes from mindlessly editing code that already works well and can be easily modified without breaking anything
homosexual did you not read a single thing that I wrote?
Copying files doesn't allow you to revert individual commits you fricking moron, it only lets you revert back to a specific point in the history of a branch.
You also have to manually track what changed in each commit whereas a version control system would do that for you.
Copying files allows me to do anything I want. What's a commit? What's a branch? If I'm adding something to the program which could break it, I just make a copy
3 months ago
Anonymous
homosexual I will write this out for you in the most grade 3 hin-glish for your stupid ESL ass if I have to: >You add feature 1 to file x.cpp >You then add features 5, 7, 8, 15 etc. to file x.cpp >Oh frick, feature 1 introduced a critical security vulnerability
How do I revert feature 1 without VC?
3 months ago
Anonymous
just remove feature 1?
lmao what is this braindeath
3 months ago
Anonymous
Just remove feature 1.
lmao
If you introduced a bug, fix the bug
If you're making an addition / change that you might want to revert, make a backup copy
It's that easy
[...]
Give the folders descriptive names like "BeforeFeatureX"
>just remove feature 1?
How?
You didn't even track what lines changed when you added feature 1 because you just copy pasted the files
This is compared to: >Right click the commit number >revert changes from this revision
With VC history
3 months ago
Anonymous
>You didn't even track what lines changed
If it's merely a matter of lines then should be able to remember, or look at the code and see what you changed and revert it
3 months ago
Anonymous
>should be able to remember
lol >or look at the code and see what you changed
that's version control...
3 months ago
Anonymous
I don't understand what you're getting at here
If you're making a potentially codebreaking change, make a backup copy of the files, it's that simple
3 months ago
Anonymous
If you're making a potentially codebreaking change, just don't make it moron
3 months ago
Anonymous
>doesn't understand the point when he said "just check the difference between the two versions manually lol" >which is LITERALLY what a version control system does automatically
3 months ago
Anonymous
You don't need to check the difference, just revert back to the copy you made
If it's a minor change you want to revert, try Ctrl+Z
3 months ago
Anonymous
That would also undo all of the changes you made after that point moron
3 months ago
Anonymous
So just go to the particular file you want to revert instead of the whole thing
3 months ago
Anonymous
That would still revert all of the changes to that file that happened later dumbass
3 months ago
Anonymous
Yes, it would
If all you need to change is just one fricking line in a file go and do it manually
3 months ago
Anonymous
And what if you need to revert the changes in 120 files dumbass?
3 months ago
Anonymous
Then go grab your previous backup you made and revert to it
3 months ago
Anonymous
That would revert all subsequent changes in all of those 120 files you fricking moron
3 months ago
Anonymous
Isn't that the point? You want to revert the changes you made and go back to a previous state from a certain point in time?
3 months ago
Anonymous
Yes that was my entire point.
That is why not having version control is bad.
Because with version control, you can just revert that commit and you keep all of the non-breaking changes which came after that point you fricking dumbass
3 months ago
Anonymous
So you're telling me you make changes to 120 files and then after that you don't actually test to see if it works or not?
3 months ago
Anonymous
NTA but this is the most amazing stupidity I've ever seen. Are you 14?
3 months ago
Anonymous
I have a suite of automated tests which run both before and after I push commits to trunk which don't require any input from me because I use VC.
I don't do manual full coverage tests after every single commit because they take too much time and if they fail I can just revert the commit that caused them in a single click because I use vc.
I can't test for every security vulnerability and neither can you because we don't know about every security vulnerability that exists moron. Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them and check if your dependencies were safe.
3 months ago
Anonymous
I don't need to test for security vunerabilities because I program video games not web servers
After I make something I test it to see if it works or not, that's the process, doing partial rollbacks on features like you're talking about happens so rarely
3 months ago
Anonymous
And the rare time it happens you save a massive amount of time.
3 months ago
Anonymous
No I don't
I copy and paste a file over, or I go into a file and copy and paste a function over
This ideal circumstance when you go and change features A, B, C, D, E, then you can just rollback A and it just works again even though you've changed all those other features sounds like a pipe dream. Usually the other changes you've made will be in a similar area so if you just roll back A it's probably not going to work fine with the new features you've introduced
3 months ago
Anonymous
>I copy and paste a file over, or I go into a file and copy and paste a function over
So you waste a bunch of time?
3 months ago
Anonymous
it takes seconds
Security vulnerabilities were just an example moron. It could be segfaults, memory leaks and other crashes could be introduced in this way. >After I make something I test it to see if it works or not
You do literally 0 regression testing? Holy fricking moron >doing partial rollbacks on features like you're talking about happens so rarely
And when it does happen it saves you a frick ton of time
It's also not the only or even biggest advantage of VC, but for some reason because you're a moron with a shovel you can't stop yourself from digging
>You do literally 0 regression testing? Holy fricking moron
What I mentioned is regression testing moron
You've failed to make substanial argument here
3 months ago
Anonymous
>after I make something I test to see if it works or not
You think that's regression testing?
Hallo sar
3 months ago
Anonymous
This is very funny coming from the guy who thinks buffer overflows are an obscure C bug only discovered recently
You're a little baby programmer
3 months ago
Anonymous
Not what I wrote
Doesn't change that you're a moron and completely wrong
3 months ago
Anonymous
You wrote this > Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them
This is a hilariously wrong and naive statement, like factually wrong, not like the opinions we've been discussing here
You're a newbie
3 months ago
Anonymous
Anon I know your entire job is copy pasting replies from chatgpt but when you misquote other people's arguments against me you just expose yourself as a brainless ESL codemonkey.
And misunderstanding what I wrote to present it as an error doesn't change the fact that you think that integration testing is regression testing because you're a fricking moron
You do not need to read the standard to understand buffer overflows
It's a very simple mistake to make when you're programming in C
You don't know because the fact that they exist is widely publicised and modern programmers have access to good, easily searchable documentation and reference material from the internet which was not true historically. You would either have to read the docs yourself or hope whatever printed reference you were using mentioned the pitfall and that you understood it.
It is a very easy mistake to make, which is my entire point, but it's much less easy to make these days. Most IDEs and compilers will even warn you when you're making it these days.
3 months ago
Anonymous
I posted what you said verbatim, you don't know how C works at all buddy
regression testing is making sure the changes you made didn't break anything that already worked
3 months ago
Anonymous
>I posted what you said verbatim, you don't know how C works at all buddy
And that quote about people not knowing about buffer overflows is 100% accurate when they're likely the single most common C security vulnerability in production code. >regression testing is making sure the changes you made didn't break anything that already worked
Huh weird that's literally the opposite of what you said earlier >after I make something I test to see if it works or not
I guess chat gpt must have written that one too
3 months ago
Anonymous
>And that quote about people not knowing about buffer overflows is 100% accurate
No, it's not. Everyone knows what a buffer overflow is. You would know if you took a single class on C at school >Huh weird that's literally the opposite of what you said earlier
You misunderstood what I said
You're an arrogant little webdev who thinks the entire world adheres to your extremely inefficient development practices
3 months ago
Anonymous
>Everyone knows what a buffer overflow is
Anon how old do you think C is? Did you take your first class in C in fricking 1973? >You misunderstood what I said
HAHAHHAHAHAHHAHAHAAH fricking homosexual lmao
3 months ago
Anonymous
I didn't take any classes on C personally, I'm just saying, if YOU took a class on it, you would know what a buffer overflow is from semester one. It's not some super technical thing, you're just writing to memory somewhere you shouldn't be - which is extremely easy to do by mistake in C, which you would know if you had actually used the language
3 months ago
Anonymous
>Everyone knows what a buffer overflow is
Anon how old do you think C is? Did you take your first class in C in fricking 1973? >You misunderstood what I said
HAHAHHAHAHAHHAHAHAAH fricking homosexual lmao
Every single C programmer in existence knew what a buffer overflow was moron, it's just an easy mistake to make. It's like off by one errors, you don't need to read the fricking standard to understand them, they just happen because they are easy to make.
Most people that started coding in C came from coding in fricking assembly for whatever architecture they were hired for back then, you think they didn't know what overwriting memory was?
3 months ago
Anonymous
People who came to C from assembly obviously knew what the underlying code was doing.
People who started with C didn't inherently and frequently fricked up because it's an error which happens silently and not something that most programmers write intentionally.
I feel like you homosexuals have literally never spoken to a single other programmer to understand how people actually learn. Most people just mash keys into the compiler until something works and then they stop testing.
I didn't take any classes on C personally, I'm just saying, if YOU took a class on it, you would know what a buffer overflow is from semester one. It's not some super technical thing, you're just writing to memory somewhere you shouldn't be - which is extremely easy to do by mistake in C, which you would know if you had actually used the language
>you would know what a buffer overflow is from semester one
Which I clearly do given that we're talking about it
But for some reason your esl ass thinks I don't because I'm saying it's easy to make and overlook?
Ok moron.
3 months ago
Anonymous
>Which I clearly do
You don't. You think you have to read the standard to know what it is. You have absolutely no idea
You can't bullshit your way out of this one bro
3 months ago
Anonymous
>You think you have to read the standard to know what it is >You don't know because the fact that they exist is widely publicised and modern programmers have access to good, easily searchable documentation and reference material from the internet which was not true historically. You would either have to read the docs yourself or hope whatever printed reference you were using mentioned the pitfall and that you understood it.
Uh oh ESL bro
Looks like chat gpt missed out on that little bit of context
3 months ago
Anonymous
Historically everyone always fricking knew about buffer overflow attacks because it's incredibly fricking obvious to anyone who knows how to use C
Like I said, you can't bullshit your way out of this, just be a man and admit you didn't know what you were talking about and made something up
3 months ago
Anonymous
>Historically everyone always fricking knew about buffer overflow attacks because it's incredibly fricking obvious to anyone who knows how to use C >source: 13 year old child who openly stated that he has never used C
lmao
3 months ago
Anonymous
I use C all the time, I'm 38
3 months ago
Anonymous
>I didn't take any classes on C
3 months ago
Anonymous
If you think really hard about it you should be able to figure out the difference between "I didn't take any classes on C" and "I use C"
3 months ago
Anonymous
>I didn't take any classes on C >but the only way you could have possibly used C is by taking a class on it >but I've also been using it since 1972
Sure bud
3 months ago
Anonymous
You can use C without taking a class on it
3 months ago
Anonymous
Yet somehow in this reply:
I didn't take any classes on C personally, I'm just saying, if YOU took a class on it, you would know what a buffer overflow is from semester one. It's not some super technical thing, you're just writing to memory somewhere you shouldn't be - which is extremely easy to do by mistake in C, which you would know if you had actually used the language
You seemed to think the only way to learn C was through classes.
C doesn't even have classes you fricking moron
3 months ago
Anonymous
Now you're just lying about what was said
3 months ago
Anonymous
are you trolling or esl
Concession accepted
3 months ago
Anonymous
are you trolling or esl
3 months ago
Anonymous
>he didn't take dicky class
3 months ago
Anonymous
>easy mistake to make
Meh just turn off compiler options like shadow stack if it's really a big issue for you.
Not quoted.
3 months ago
Anonymous
It doesn't not take seconds to find a single function. Unless you want to pretend you have 100% of your code memorized at all times. You remember the extact name of every function in every file? Or are all the games you work on so simple you never change anything you worked on months ago? If you don't use VC, you enjoy wasting time
3 months ago
Anonymous
It takes seconds to copy some files over
Actually going into an old file and digging out some old functions you want to regress to happens pretty rarely, and when it does it doesn't take a long time to do (outside any updates you need to make to it, but VC doesn't save you from that either)
3 months ago
Anonymous
>After I make something I test it to see if it works or not
you can at least be 100% sure something that used to work isn't broken now if you have automated tests
playthroughs are fine for integration testing that automated tests cannot easily cover
and you will thank yourself if you port a game to another system, where many things can break
3 months ago
Anonymous
automated tests are kind of shit for game development
porting things to other systems really only affects about 1% of the codebase
3 months ago
Anonymous
for making sure the ui still works and doesn't get stuck in weird states, especially if features are added, then it's still useful there
though TDD is awful and should never be used for a project where the design evolves rapidly and requires tests to be rewritten constantly
the tests should just be to lock in the correct behaviours
this talk is interesting though
3 months ago
Anonymous
>porting things to other systems really only affects about 1% of the codebase
depends, what if the system you're porting to uses different endian, and you have bit shifting functions and comparisons in your code?
suddenly all those will break
but porting is much simpler now because everything runs on engines and you dont have to store floats as ints and other tricks on certain hardware
3 months ago
Anonymous
>what if the system you're porting to uses different endian
we're talking about things that actually happen in reality
3 months ago
Anonymous
it used to be a real thing that came up
now hardware and software is more standard, so porting is easier
3 months ago
Anonymous
The last time I remember anyone giving a shit about endians was in the 90s
3 months ago
Anonymous
Security vulnerabilities were just an example moron. It could be segfaults, memory leaks and other crashes could be introduced in this way. >After I make something I test it to see if it works or not
You do literally 0 regression testing? Holy fricking moron >doing partial rollbacks on features like you're talking about happens so rarely
And when it does happen it saves you a frick ton of time
It's also not the only or even biggest advantage of VC, but for some reason because you're a moron with a shovel you can't stop yourself from digging
3 months ago
Anonymous
>I don't need to test for security vunerabilities because I program video games not web servers
Hope you never make any multiplayer games.
3 months ago
Anonymous
The surface area for security vunerabilities in multiplayer games is very small
3 months ago
Anonymous
How's the pay at amazon, anon? I head the working conditions can be pretty crap.
3 months ago
Anonymous
That's completely contextual.
3 months ago
Anonymous
Not really, it applies to every multiplayer game, any code that can actually cause security problems is very small and very contained
3 months ago
Anonymous
You're making a hasty generalization with no basis. It completely depends on the implementation. There's countless stories of ACE, even from veteran developers like Valve in CS2.
3 months ago
Anonymous
I hope you don't expect your multiplayer games to be truly secure. Most of them aren't.
3 months ago
Anonymous
>Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them
Buffer overflows have been known about since C was invented, they're a very simple bug to understand
3 months ago
Anonymous
>Buffer overflows have been known about since C was invented
Sure, but you are significantly overestimating the average C and C++ programmer here if you think "known" means "widely known" or "widely understood".
Very few people have ever actually sat down and read the standard or the full definition of the methods they call.
>by hiding implementation details
This is for glue coders.
For actual programmers that's pretty large negative. Particularly once you try to step through it and you have five layers of indirection and can't keep watch windows of out of bound structs.
>For actual programmers that's pretty large negative. Particularly once you try to step through it and you have five layers of indirection and can't keep watch windows of out of bound structs.
If state crosses the boundaries of structs you've fricked up the segregation of your components and they should be more strongly coupled.
Segregation like this should let you ignore what's going on above that method call and isolate errors to individual components that cause them. This is much better than having to keep track of the literally hundreds of attributes that a single class can have which potentially get mutated and affect how the program runs.
3 months ago
Anonymous
You do not need to read the standard to understand buffer overflows
It's a very simple mistake to make when you're programming in C
3 months ago
Anonymous
>You don't need to check the difference
>You didn't even track what lines changed
If it's merely a matter of lines then should be able to remember, or look at the code and see what you changed and revert it
>or look at the code and see what you changed
You were talking about manually checking the difference between two versions
You cannot walk back that post, no matter how hard you try. You accidentally got tangled in your own web of shitposting
Thanks for conceding.
3 months ago
Anonymous
>You cannot walk back that post
I have no idea what you're trying to get across. I very rarely cross-reference code to manually revery something, I was trying to understand this hypothetical situation being presented to me
3 months ago
Anonymous
is LITERALLY what a version control system does automatically
Ah yes, let me just check the difference between now and last week >git -select-version $92ha4 -execute jkaq -rebase-into-now -into $2nc90a -timeframe-in-ms 298127398129
3 months ago
Anonymous
Oh shit, you accidentally deleted three random commits!
Don't worry, you can fix this by executing three other esoteric commands, but you will lose your latest changes... sorry!
3 months ago
Anonymous
>I'm too moronic to use a hammer without breaking my fingers, therefore hammers are bad
git log
git diff <commit from a week ago> HEAD
so hard
3 months ago
Anonymous
>git log
What the frick does this do? Why am I making a log when I want to compare things? >git diff <commit from a week ago> HEAD
Why the frick do I have to write an extremely long string of random numbers and letters? What the frick does HEAD mean?
I can just open my project folder, open the folder of last week, open the file I want to compare and compare it with the already open file. Simple as.
3 months ago
Anonymous
Use SVN then you stupid monkey Black person, it's basically the "backup your file once a day" version control for homosexuals who are too stupid to copypaste a commit ID
3 months ago
Anonymous
Why would I when I can just make a copy of my code?
3 months ago
Anonymous
You gui plugin/program if you don't want to type commit ids, you moronic Black.
3 months ago
Anonymous
Ah so I need a tool to fix my tool so I can use a tool?
No thanks, I'll stick to making daily local backups.
3 months ago
Anonymous
Because between the change that broke things and your current version there might be 50 different changes you've done in the meantime. Version control will tell you exactly what was in the risky change, while your backup will show you all 51 changes forcing you to recall and manually reconstruct what you did. Meanwhile VC saves you time and headspace by doing it for you.
3 months ago
Anonymous
that is basically what VC does, except instead of making a 1GB copy of the project for each backup, it just tracks the 3kb diff of the code that actually changed, with 100% accuracy and not relying on your memory of what you did or didn't modify.
this thread has pretty much proven that people like you are just abject morons. it's not so much that you just lack an understanding of how VC is useful, you are just aggressively and willfully stubborn about it because you're too dumb to figure out how to use it.
3 months ago
Anonymous
Why the frick would you have 1GB of source code? It's text
If you make changes to code, don't test it, then just forget what you fricking wrote, you are a moron
3 months ago
Anonymous
>then should be able to remember
So how many lines of code are in this project you're managing without vc btw? 5? >or look at the code and see what you changed and revert it
You're doing this manually? In 500+ LOC files?
3 months ago
Anonymous
Just remove feature 1.
lmao
3 months ago
Anonymous
If you introduced a bug, fix the bug
If you're making an addition / change that you might want to revert, make a backup copy
It's that easy
how do you know which usb drive has the file/code you want?
Give the folders descriptive names like "BeforeFeatureX"
3 months ago
Anonymous
how do you know which usb drive has the file/code you want?
Self taught doesn't mean not knowing real programming. Someone who has actually shipped hobby shit can be worth 10 CS graduates who have only done their course work. But neither is being a hobbyist guarantee of skill.
>ragtag bunch of jap autists band together to make a janky crafting survival parody game because they felt like it >twitter immediately collapses into a singularity from sheer seethe and goes out of their way to slander it >actual AAA devs are actively shitting on it because "it feels cheated" >same people who spent years complaining about Pokemon's lack of quality are now furiously bootlicking nintendo >people donwright refusing to accept it is successful and make up lies about THE EVIL AI GENERATION >same twitter that claims to support the "little guy" and hates big corporations
I don't know you guys, but this whole thing has been quite the eye opener
bunch of jap autists band together to make a janky crafting survival parody game because they felt like it
the same studio made a moderately successful craftopia 4 years earlier and palworld was advertised multiple times at major game events like tgs, you fell for a meme
>mouthbrrather is so dumb he introduces CVEs in his code because he pulled a third party library to compute if a number is even and log it with colors >"YOU NEED VERSHUN c**tRUL!!"
>ignored that literally everyone on the core team is japanese
The only foreigners credited are some english VAs, some localizers and some outsourcing studio for 3D.
Any principle/strategy/design/whatever that has a catchy acronym as a name is worthless bullshit. There might be exceptions, but it's 100% true for anything *management related.
I get the impression that this entire game hinges on them lucking out and finding the one really experienced dude who knew Unreal Engine inside out and basically became project lead in all but name.
it's dumb to use solid as a solo dev, no one else will use your code, you can modify everything whenever you want and you don't need a dozen unnecessary abstactions
I'm gonna guess you're secure enough in your job that a bunch of college kids who don't know shit don't faze you, but just in case: You're 100% right
I work at a company with 500 devs, delivering a suite of software for over 30 years at this point
I have seen the old code that no one ever had time to refactor. Takes fricking 3 days of reading to understand a tiny part of it, whereas new code following proper patterns is easy to maintain
>how do I remove a feature
Find where you defined it.
Change the name to_getridofthisshit() and compile it.
Clean up all errors that occur.
Remove definition.
Anon, you need to find yourself a good Indian restaurant.
Because once you taste real authentic curry, you will never want to go back to mashed potatoes.
I could actually, cause I just RDP onto the developer PC that's standing underneath my desk at the office
But without version control it gets kinda hard to coordinate with hundreds of other devs
If you need hundreds of devs you're doing something wrong.
Oblivion was made with about 40 programmers.
Half Life 2 had about 50.
It took about 50 developers 10 months to make Halo 2.
Rockstar needed 130 developers for GTA4.
Now dev studios have 600 employees and getting nowhere
free market is always better than corporativism pushed by political agendas like wokeness and DEI, funded by billionaires
nasdaq.com/solutions/corporate-esg-solutions
I will always blame myself for obeying my moronic boomers and teachers instead of trying to learn myself
If I were a zoomer I would just learn to program and learn some useful skill
Git is often not ideal for a game project because the majority of the work usually involves art assets nowadays rather than code work. Having full version history for projects where the full WIP filesize can be easily in the TB range (high quality sculpts for bakes can be 100 mill + vertices) is legitimately awful.
[...]
I'm gonna guess you're secure enough in your job that a bunch of college kids who don't know shit don't faze you, but just in case: You're 100% right
I work at a company with 500 devs, delivering a suite of software for over 30 years at this point
I have seen the old code that no one ever had time to refactor. Takes fricking 3 days of reading to understand a tiny part of it, whereas new code following proper patterns is easy to maintain
I want to say the same.
I work in a big company as well though it's not mainly developing software but still it has a huge IT department with many applications.
Ignoring software development standards can be fine if you are just doing a one time gig that no one will care about updating in the future. Maybe that's what the palworld team had in mind as well (or it was just straight up cluelessness or manager's greed). But even they will notice if they are planning on keeping this game updated for some years, that technical debt is real and that they could have made their lives way easier (read 'develop cheaper') if they adhered to software development standards.
Sure, you are not supposed to read them like the Gospel and you need to tailor them to your needs but just straight up ignoring them will only hurt you in the long run.
This is also something that all the college kids in here will learn as soon as they have their first junior position because everyone hates fixing bad code that doesn't adhere to the necessary standards. So guess who gets all of those grunt tickets. Correct, the junior.
Oh, and not using Git (or svn for the boomers here) and a (at least basic) cicd system for software projects with any kind of relevance is just 1) torturing urself unnecessarily and 2) has a high chance of simply blocking your whole development operation at many occasions for stuff that could have been easily avoided by using version control and a cicd system.
The game is objectively fricking garbage. It's really annoying seeing twats pretend like Palworld is some sort of incredible success story instead of just a random fluke. Every other week streamers decide to shill a new game and it becomes the newest FOTM. There's a mountain of games like palworld that weren't so lucky and died stillborn.
For frick's sake you idiots keep in mind game dev also involves people like envoirnment artists who have 0 programming knowledge (they don't need to) but do have a good taste when it comes to making something like a nice garden. Your grandma would probably excel in decorating a nice large open world but should absolutely be never allowed near git. She will accidentally mess up, work on an obsolete branch etc and potentially piss away weeks of her own work before it's discovered. A surprisingly large people working on video games on the artistic side are near completely tech illiterate. Your solution of version control needs to accomodate those kind of people which git doesn't.
And no, you're not going to hire an expensive programmer just to help decorate a nice video game envoirnment.
Nobody would put their houses next to cliffs like that in mountain range. Not only would all the cold air rush down on them 24/7 but they would also get demolished by avalanches every winter.
Nobody would put their houses next to cliffs like that in mountain range. Not only would all the cold air rush down on them 24/7 but they would also get demolished by avalanches every winter.
uhh anon you know it's a real place right? Look up Lauterbrunnen in Switzerland.
>passion first >credentialism is worthless
The biggest players in the industry in the 80s and 90s already knew this. (before investpoors started buying up successful studios)
These days it is literally considered discrimination to say as a game dev that you'll only hire people who enjoy games. That was one of the things in the big California lawsuit against Blizzard, a woman said that she felt unwelcome because she didn't like games.
Bureaucracy in a company isn't useless, it serves an important function: giving something to do for nepotism hires who are unable to contribute to the actual product
Every autistic male coder in a cubicle can subsidize the existence of 4-10 women who professionally go to meetings and dick around on a computer all day
Depends who and where. Having what's basically a janitor that keeps track of stuff so it doesn't get lost, prods people every once in a while for progress, delegates tasks and shields people in a team from bureaucratic nonsense from above is a good thing.
>real shit >being forced to babysit incompetents even more than I already do through pair programming
I will literally switch a job before I tolerate that shit and I have a relatively high tolerance for bullshit.
Bureaucracy makes EVERYTHING inefficient, whether it be Federal Bureaucracy or Corporate Bureaucracy. Palworld shows that a group of guys all on the same page are way more efficient and profitable than a 200+ horde of programmers, lower-middle managers, middle manager, upper-middle managers, project leads, HR ladies, scrum masters, and bean counters.
>software devs
More like management.
Actual devs rarely have a say in these things, outside of the recent trend of handmade rethoric seeping into the mainstream by calling out the most egregious moronation.
>In software engineering, version control (also known as revision control, source control, or source code management) is a class of systems responsible for managing changes to computer programs, documents, large web sites, or other collections of information. Version control is a component of software configuration management.
They didn't hire any jeets. If you go through the credits all the actual employees are japanese.
They paid some 3D outsourcing work to a jeet studio.
Just like they paid some murricans for localization and voice acting. (which is the real reason why you have Type A and Type B in there)
>OH SHIT A POORLY MADE FOTM OPEN WORLD SURVIVAL GAME WITH DISCOUNT POKEYMANS MADE BANK??? >BECAUSE...IT FOLLOWED THE MOST LUCRATIVE POSSIBLE NICHE IN THE MARKET!? >""""""""""""""FREE MARKET""""""""""""""""" CONFIRMED FOR BROS!!!111!
The most chilling part is that these are the people who don't know left from right and call anyone who's not a temporarily embarassed billionaire a communist and a useful idiot.
Then again >Twitter thread
>seething over bros making money for doing something popular that isn't supplied by the big studios >using terms like "temporarily embarrassed billionaire" and whining about conservatives
Go split your dick you massive homosexual.
>poor and obsessed with guns
should have tried pol
nupol is all federal agents, brown people who dont know what a gun is, weebs who dont know how to drive a car, indian shills and more brown people.
>convenience store
>browns, indians, and more browns
yeah like anon said, they should have tried pol
It's Japanese convinience store.
>t. never went into a konbini
It's full of Indians now
>He doesn't know
Every conbini in Tokyo is staffed by Indians and it's spreading throughout the entire country
>t. Not in Tokyo
>he doesn't actively avoid konbini staffed by non-japanese
Very hard to do in the major metros. Easy to do out here but becoming more difficult every year
It's not that hard to avoid indian ones.
Harder once you start including koreans and chinese.
Starting to see Muslims all over the place even in the fricking inaka now. You look at any social media involving Japan and the comments are Muslims claiming Japan needs Islam, we will take over Japan, etc. everyone thinks this country is some based anti immigrant heaven but thanks to student and work visas it is becoming anything but
if you need gaijin deportation volunteers, I will fly to japan tomorrow
I'm a gaijin myself so I'd be hypocritical if I said let's go, but at least I try to integrate. The Muslims and central Asians don't even fricking bother
integrated gaijin are fine
but gaijin that go to japan and try to make it not-japan, should be all deported
just like this
Good luck, the Japanese are way too passive.
Being hypocritical is fine. That's how countries are saved from this kind of shit.
>b-but muh principles
For 99% of the people pushing for this shit it comes naturally to them to abuse principles like yours.
>He thinks brownies are allowed to enter shops in japan
There's a reason why spanish soccerBlack folk and amerimutts dont go to japan
You've never been to Japan
>no no those degenerates arent white, they are simply larping like they are
well then, where is pol then? soijak party probably
>soijak party
Not even a real site
What's with all the woketard trannies lately?
Anon have you seen the amount of Indian flags on there these days. It's genuinely disgusting hpw many of them are on this site now.
i mean yeah, don't you remember/misc/ meetup from hwndu era? literally not a single white face.
GG era was the last time /misc/ was fun.
After Trump won, the place is a constant shitting street for bots.
That was in israelite York moron
Seems like YOU'RE the one that's obsessed.
Yep. Version control, CI/CD, Agile methodology (or even waterfall), design patterns, solid principles etc. are gigantic wastes of time and money that create multiple bullshit jobs and sap the actual programmers' productivity. Corporations have ruined programming, and AI tools will only make it worse.
The best form of programming is 5-8 guys writing what code they want in one place each and then patching it together at the end.
>Agile methodology
I 100% agree with this post
t. worked on a tech project for an organisation that wanted to try agile for the first time on one of the biggest national reforms to date. Pissed 1.5 billion up the wall and the project got shitcanned before anything even made it to release.
I've had better experiences with agile than waterfall, but these are always going to depend on the scale of the project I feel. The waterfall project was just massive, so it was hard to course correct some things.
It depends on the project. I've seen the goods of it making development very easy and straight forward but also the bads where each call was just a huge, paid timewaste with way too many of them. It really depends on how "well" they execute it.
>blame a methodology for a failure that came from the people in charge of managing a business process transition
lol
Fair point, but I feel that agile is easier to frick up than most because it's so 'freeform' and if you're not tight with actual sprint outcomes and what the focus is, you're bound to piss away time and resources on nonsense.
Also, agile always seems to have a bunch of fricking zealots who preach about how good it is, without articualating the inherent flaws
My entire education and career are those things and I can't lie this post is pretty based
>Product Owner
>Project Manager
>SCRUM Master
How to be useless at a professional level.
>team leader
You can't imagine how much I hate the team "leaders" of the project I'm working in. They are glorified, overpaid seat heaters, because that's all they do. They probably work less than twenty minutes a day.
Don't learn any of this and stay unemployed. More job opportunities in this shit economy for me lol.
I work as a lead dev and would have fired you on the spot for saying that version control, CI/CD, design patterns and especially SOLID are redundant.
Yes, there is circlejerky bullshit that nobody should ever care about, but your job as a programmer is to separate the wheat from the chaff.
I get your point on agile, but the rest of them I won't concede. Your code is useless trash if you don't apply SOLID. Sorry, but it's the truth.
You might be a lead but you're still not a very good programmer if you think SOLID is worth adhering to
You have never touched a single line of game code in your life, and that's a good thing. Stay in your thinly-veiled welfare bullshit job.
>solid defense force
Give it a few years and you finally start to understand what good programming is like. You're not fooling anyone here with your sv web moron opinions
>yeah so what if it takes me 20 years to add any new feature and every time I do it causes a bug? Still doesn't make solid worth following
SOLID can make it take more time to add a feature where you have to break everything into more components than neccessary to satisfy the single use princple and create a bunch of unneccessary interfaces to satisfy the dependency inversion principle
Oh no, I have to create two interfaces to communicate between components. The 5 seconds it takes me to create those is definitely such an amazing loss of efficiency that we shouldn't have interfaces at all
Get fricked, next programmer who wants to figure out what functions already exist to access this component, dig through the whole implementation to find out
Interfaces take time to write, take time to read and make your code run slower
The costs add up
>take time to read
lel
"oh man I really wonder which service BusinessObjectServiceInterface.h uses, I guess I'll have to go check the implementation! stupid interfaces ugh why can't I just write it all in one class"
Try writing code with more than 2 people sometime, or a program you have to support for longer than whenever your college course ends
People can definitely take the principles way too far and get way too anal and abstract about what it means for a class to have a "single responsibility" or what is a "high level/low level module".
They are broadly useful tools for when the logic of a program starts to become too complex/entangled and you want to isolate it to prevent unnecessary coupling.
>Interfaces take time to write
True
>take time to read
They save significantly more time by hiding implementation details and decoupling isolated parts of the program
>make your code run slower
If interfaces are causing a measurable performance hit to your program you're doing something catastrophically wrong.
Generally the best reason to not use them is when the logic of your program is too simple to warrant it or the area where you're working doesn't change often.
>by hiding implementation details
This is for glue coders.
For actual programmers that's pretty large negative. Particularly once you try to step through it and you have five layers of indirection and can't keep watch windows of out of bound structs.
>SOLID is not redundant
>literal OOP dogma with no metrics to back it up at all
lmfao
It really depends what your programming. If you have a solid vision for the program you don't need any of that. Those aspects of the job are to accomodate constant changes because the people responsible for the functionalities and features can't make up their mind, or have no idea, what they want.
frick your patterns, frick SOLID and frick all other superficial bullshit, every SOLID and pattern worshipping moron I've encountered wrote the shittiest code imaginable and I've worked with hundredths of people over 15 years of programming
What you don't like having Java bullshit in your code anon?
>I have no idea about the actual work I supervise but if a worker with actual hands-on exp disagrees with me I'd rather lose the worker than give up fraction of my israelite-given authority
It's the roman way though
The only person I ever had mention SOLID at my workplace was a system architect, great at his specific job and good at being able to look at the important parts of the problem while being able to adapt to new constraints very easily, but not a programmer, never claimed to be one. His adherence to it was indication that it (like so many similar dogmas) is born out of trying to fit programming into a box. In reality it always depends.
A system architect's "job" is knowing the common patterns at the top level and being able to apply them without missing the forest for the trees. SOLID makes sense to them for this reason; it shouldn't make sense to programmers for the exact same reason. Similar to OOP though OOP's complicated by the fact that its roots in smalltalk describe something completely different to what stroustrupp seems to have taken away from it.
Using CI/CD in the same sentence as SOLID is just bizarre to me. Is it useful having a build, test, and deploy process you can kick off with one command without any fiddling with configs? Hell yeah. Is it useful to have it able to run somewhere other than the dev's hardware? Yeah, a lot of the time. Is it useful having it automatically triggered by various events? Maybe, really depends on what you're building and how complicated your deployments are. Should YOU have that(?) for YOUR specific project? Maybe. How long is a piece of string?
What is even version control? To approach this pragmatically; sticking your code in a git repo is always useful even if you just use it to streamline your backup process. I've seen people with IDE plugins that push code back to the upstream repo every time they save a file. Its like those videos of third worlders using a jackhammer like a chisel without even plugging it in, completely bizarre to see, but even so they get value out of it. I find it hard to imagine why you wouldn't use git. Its not like it costs anything.
What a bizarre world view where architects and programmers have opposed philosophies that clash because I guess you're talking about junior programmers that only wanna write all the code while ignoring patterns to make it more robust in case of future change?
>junior programmers that only wanna write all the code while ignoring patterns to make it more robust in case of future change?
I'm not sure in this sentence whether the "to make it more robust" is being ascribed to the junior ignoring the pattern, or the pattern itself.
Well maybe you should take some English classes then if you think "it" could refer to "patterns" OR "programmers"
what I really hate is when I'm told to take durable robust code and make it fragile, because it's "simpler" or because they dont think the code will be used in other ways
Well I hate having to explain that it'd be faster to just write the 500 lines of code myself than to write 2000 lines of documentation so the crack team in south vietnam can do it a fifth as fast with five times the bugs but we can't have everything, anon.
What were they actually asking for?
Our system architect's job is basically to know how this convoluted system communicates to different vendors and sets up a call.
Ours does some of that, too, actually. You're right, but usually only in relation to new builds.
>I'm a lead dev
>I would have fired you on the spot
woah calm down there middle manager, aren't you overstepping your post a bit there?
version control is a great tool and should even be used when its a one man project
but i agree with the rest
version control is nice to have, but you have to spend a good amount of time to keep your history clean and will never look at it again.
You can just make your application until you have a stable release and only then start using VC
>you have to spend a good amount of time to keep your history clean
no you don't
you add a new feature, its not finished yet but you want to test if its displayed properly in the gui.
You comment the call to handler to make sure the code compiles.
You also notice a minor mistake in another part of the code.
Now you have
>an unfinished feature
>a temporary change
>an unrelated fix
now what do you do,
commit? extract the fix and commit separately, rebase that shit later.
Why would you commit an unfinished feature and some debug shit? You're not supposed to just commit everything at the end of the day.
>commit == push
>push == push into production
Here you go https://en.wikipedia.org/wiki/Branching_(version_control)
they really should have named them more logically
savelocal
savetoserver
much simpler to understand
Don't even start about GIT naming, I'm still mad we aren't allowed to call the main branch "master" anymore
OH YEAH I totally forgot about that
for games, its cool to still have every version of your game that you can build anytime and show others
I really like looking at game prototypes
unfortunately most have been lost because they didn't use source control
>git is github
classic
This.
>savetolocal
>savetoserver
>pushlocaltoserver
>pushservertolocal
>forcepushservertolocal
Things would be massively easier if commands were named after what they did, instead of being "fast" to write.
The cool thing about bash is you can make your own mnemonics that is a miserable word to spell
>I just commit when everythings done
if you use a modern CI/CD pipeline you are forced to commit to start the build pipeline
I only start using VC when there's big changes
or I bring in another coder who may muck things up
it is nice to be able to regenerate an old build
imagine if we had all of ocarina of time's old code and could play the old builds at any point in development cycle, imagine how much we would learn, instead of just the final result and a few snapshots
even a simple stage all -> commit -> occasionally push to remote routine is better than no version control at all, because it lets you backtrack if you or your software ever frick something up
>but you have to spend a good amount of time to keep your history clean and will never look at it again.
???????????????
Patching it together just by eyeballing it? Git provides tools to show differences when debasing in the form of merge conflicts. Catching everything manually is impossible and at the very least would take tediously long, unless it's a toy project with barely any loc
>game should be easier to debug
>on a pay per hour model
>with lazy devs that haven't sleep in months
>with schedules were 1 day of delay is asking too much if the game still runs.
yeah , like that sounds for nintendo and indie sht
but for idsofware or square , thats literally game insdustry cancer.
All git diffing fricking blow compared to good tools like BeyondCompare.
I work in company that deals with natural gas and electricity, for programmer to not know GIT at this level would be a joke, they'd be fired instantly for being a moron.
Look at all the seething corpogays you triggered lmao
based
>some unemployed poorgay "owned" me because he hates version control(because he's not programming)
>continues seething
back to your cubicle, codemonkey lmao
>cubicle
>codemonkey
We have home office because corpos afraid we're going to leave if they don't allow us that, lmao.
Agree with 90%, but version control is an absolute necessity and anyone who says otherwise is unemployed larper
Exactly. Only morons who never worked a single day in their lives as programmers would say Git is bad.
Svn is better than git
how is a versioning system where you cannot keep a local copy of your work and update it better?
oh right we're on Gankereddit where it's hip to be contrarian at all times
git's abilities are fine, just the commands were invented by a crazy person
It's fricking not. I worked almost 5 years with SVN before migrating to Git and it's fricking night and day. I'd never go back and I would question the sanity of anyone suggesting to do so
yep, but when you have 200-500 people working on a project, source control is a must
however git is a convoluted pile of shit that has probably caused more problems with its ambiguous commands which prevent people from using it with 100% confidence
it really depends on the project's design and how flexible and non moronic the devs are
>git is a convoluted pile of shit that has probably caused more problems with its ambiguous commands which prevent people from using it with 100% confidence
Look at this loser, he doesn't fricking know how to use git
I can use git, but you cannot deny that git is poorly designed
https://tom-vykes.medium.com/the-worst-things-about-github-8e8efc60fae3
and "pull requests" has always been a fricking moronic term
especially when you push commits
Gitlab calls them "merge requests", which fits better.
thats what I would call it
in his defense git has kinda bad command bloat where there's too many ways of doing similar (or the same) thing
And most of the time you won't need that, you work on local dev branch, you merge into remote main branch, you pull from remote. For 99% of cases it's all person needs
I don't disagree but it's a valid criticism of git compared to some other VCS
Jira es the worst shit ever invented
anon please update your jira card
please
anon please add story points to your jira
>Oh weird I did but I guess jira's just buggered again I'll put it in later today once I get this working
>>Do nothing and say the same thing tomorrow
They'll get it eventually. Or they wont.
Programming died when all the non programmers found out how much moner programmers made. After 10 years of everyone learning to code being mainstream(thanks obama), tech is just another corporated beaurocracy made up more normalgays. Autistic lesser code gods cant even get jobs in current tech because they can’t pass the interview process while the the dead weight extraverts can easily nail HR interview process
This is accurate, my autistic friend couldn’t even get a job and he was the best damn programmer I’ve ever seen
this post needs to be framed. It's infuriating.
Companies that overvalue the interview process aren't worth working for in the first place. If a company is lorded over by HR karens do you really think you're going to have fun working there? They've done you a favour.
This is why I consider the modern environment of game development to be so fricking dumb. The impression I grew up with with people who were making huge changes in the industry came from roots crammed up in a garage / basement working their ass off because the process of coding is fun and playing their game as they go to iron out the bugs.
I agree only to a point. Version control is way too helpful however, if nothing else than to keep track of updates and fixes over time.
for big applications you need this shit to keep your sanity and make sure someone doesn't casually hides a backdoor in your application.
You don't need this for small teams esp. games that are essentially write once.
Small indie studios don't even have 5 programmers.
in what year would you say game engines were still written once and not reused for subsequent products?
>Corporations have ruined programming, and AI tools will only make it worse.
incredibly based
Most of this is valid except for Git. Having backups is way too useful to passup and doesn’t take your own storage space.
git is great though.
i used it for everything related to text documents and not just source code.
honestly i dont want to go back to a time when i didnt know what git was.
i personally like to use lazygit.
>Version control
moron
>CI/CD
moron
>Agile methodology
Partly agree. It's a broadly useful tool but people get way too ideological about it and don't actually check to see if it's making them more productive.
>or even waterfall
That's the alternative to agile moron
>design patterns
You don't need to know them (especially not if you're working on your own) but they are good and recognisable solutions to common problems
They absolutely shouldn't be something you focus your time on as a beginner but for big projects they're generally worth understanding.
>solid principles
moron
So 2/5?
>NOOO SHOW ME THE HECKIN INFOGRAPHS AND METRICS THAT PROVE YOUR POINT
You have corporate brainrot. Just because a tool is useful doesn't mean it's beyond criticism.
Oh you're an ESL who can't read english and didn't even understand that I was agreeing with him
That explains a lot.
I think fancy testing techniques/patterns are bullshit and just exist to give positions to bootlickers who can't program for shit.
Disagree with the first line, agree with the second line.
There's a reason that older games felt more creative/soulful, and it's not simply because there was more uncharted territory. It's because the teams were tight-knit groups where everyone was a generalist instead of a specialist. Games were more cohesive because the developers worked on multiple elements and linked them all together. Whereas now there's hundreds of devs on a team and they all do one specific thing. Nowadays you have to play indie games to get that kind of tight-knit design.
Smaller teams also have a much better sense of what how they can steer the game and what's possible than large teams which lets them be much more dynamic and innovative.
If you get half way into a $300mil game and realise it isn't fun there's basically nothing you can do about it but with a small dev team you can rip it apart completely and only keep what works
Same feeling I got after watching the dev documentaries for Halo 1,2 and 3
In the first two games you really get the feeling it's just a bunch of friends making what they want. They're doing social activities together, everyone is pitching in on everything, programmers have to do voices and storyboards, the composer has to code and help writing the story, the lead dev is adding easter eggs in his spare time.
In halo 3 things are getting a bit more business like, but still relatively casual, but then for Halo 4 it's all corporate, every desk is completely clean, everyone is at their desk typing quietly
A few months ago they fired all the company's scrum masters
I have to do way less work and the projects still aren't deploying on time so it's a net positive for me
AI will make it better. Everybody will be able to make games.
Right is all the corpo buttfrick practices. Left is sane programming with performance in mind.
Left is one guy who cares, right is a bunch of pajeets duct-taping shit together
Left is a C programmer who has pretty harsh opinions on corporate non-standards.
t. not quoted but has read some of his blogposts on arena allocation
Right is how I felt using Unity. It takes 10 minutes to even start up. Glad I moved away from that and am using Monogame now.
>Version control
>ruined programming
version control is just common sense
CI/CD are basically bullshit. but build/deployment automation is good, it just doesn't need to be "continuous". testing is a mixed bag and people can get completely obsessed with it over actually writing code.in the right dose its good.
but yea the rest you can throw into the garbage
Agile is actually a systemic manifestation of demonic forces and corrupts the soul to a near irreparable degree
I am not kidding it legitimately will create evil human beings
>Agile is the devil
What's so bad about dividing the work into chunks, then plan every two weeks what chunks to do right now?
Because in practice the two week chunks are only used for metering/predicting deliveries and you spend more time discussing work than working.
In concept, the original version of Agile is actually good. Because it essential was "isolate the programmers in a room, divide the work into little chunks, and don't let any business people disturb them"
the chunking the work people still do, but the rest is completely absent
in practice, agile teams create lots of devs siloing themselves, not working as a team, and hoarding high impact work for themselves
it also unintentionally incentivizes managers/product managers backchanneling new work into the sprint instead of going through the actual process
to go through the process would require them to admit where they fricked up, so they try to slip shit in discretely, or rely on buddy relationships with particular programmers to add extra features
agile basically guarantees that every team devolves into a malthusian dystopia
Agile makes it so managers don’t have to do any planning (which is supposed to be their job) and make the developers do it instead. It’s supposed to give developers “ownership of their tasks” or something. I had to spend a whole day every week in meeting after meeting doing sprint planning, sprint retrospectives, backlog refinement, etc
>it just doesn't need to be "continuous"
it means testers can play a current build and spot problems sooner instead of waiting for the build at the end of the day or however often
and not have to deal with things that have already been fixed
also to push bugfixes to players, it's still essential you can do that asap without much effort before gamebreaking bugs break everything
I work ln DevOps and I agree. Frick all of this bloated shit.
Someone finally gets it
like everything in the field, it starts by making sense for specific use cases and then incompetent people start copying it without understanding it properly because they think it's some sort of recipe for success
version control and CI/CD is fine, especially if you work on a team. but everything else is bloat
>Yep, Version con-
stopped reading there, way to out yourself
So what this Palworld Phenomenon is is basically what Vidya used to be like 2006 and before where it's was a dozen guys just making their own game and then releasing it to the normalgays instead of making vidya for normalgays
Asset stores weren't available in 2006
>Version Control bad
Make sense why most channers are unemployed
>only reads the first word of the first line and completely misses the context
Ladies and gentlemen, I present you the average corporate OOP dev
Git is based as frick I can can't live without it anymore. But I agree with the rest.
Nobody in their right mind would go against Git. When even the suckless homosexuals use Git you know it's the absolute bare minimum.
Git sucks pretty bad in many respects even for basic stuff and relies on you using tooling to circumvent it.
It's just that so do most of the alternatives.
Personally i'd like something like Fossil but it's also bad with binary files like Git is.
Shit's not scalable dummy and you are a fool if you think they don't do version control now for a game of this scale.
>modern Ganker thinks being inefficient, unskilled and cashing in on normie trends is good now
allowing pokegays back in was a mistake
Contrarian fishing is now terminal there isn't a single sincere post to be found.
>being inefficient, unskilled and cashing in on normie trends
It's ok, Blizzard fired all those people just yesterday.
>Blizzard fired
Microsoft fired, blizzard is not a unique, non-microsoft entity anymore and should never be discussed as if they are on this board. Blizzard makes no longer decisions for themselves, whatever is left will just take orders from the largest tech company on the planet.
this place really is just an outlet for Redditors to be contrarian on
yet they made more cash than you ever see in your lifetime.
Pokegays will buy any slop as long as it's reminiscent of what they were fed as kids.
>people spend money on what they like
no shit sherlock
this larp how git is somehow bad is a weird one
Git has its issues, but I'll rather use it than SVN or a bunch of custom-made version control solutions.
yeah well we arent talking about comparing Git to other version control, the guy in OP post is shit talking version control at all. Like what?
It's just a funny development anecdote. I don't think it's anons necessarily saying version control is a bad idea, just that they are amused by the apparent inexperience Palworld developers had before getting started. The value of version of control is something you learn with experience, rather than being a strict requirement to make anything.
When you start digging, there's always plenty of funny or off-the-wall stories various developers may have. Often times such stories involve source code or various development practices, where in this case it's version control.
Aye.
I program legacy software at an investment company. My big project after starting was converting the project to a new version that uses text files rather than a monolithic binary file.
Since that finished, I've slowly been teaching the 85 year old guy on our team how to use git, he's almost got it
Good gravy, those banking and investment projects are always something else.
It's a bit too common of a phrase to universally claim as such, don't you think? Visits to a doctor would get weird.
Tell me the flashdrive story again grampa, it was a rollicker. I liked the part where you used flash drives but most people don't use flash drives. that was the craziest part.
Just people who don't know anything about real programming who think it's a buzzword like Agile
yeah well turns out that they did use Git but transitionted to SVN because senior dev didnt know Git
Fair enough. Nothing wrong with SVN.
>Nothing wrong
Anyone who says this is from pol
That sounds wrong.
It's literally a dogwhistle
Go back in time to Tumblr.
Nothing wrong with using "nothing wrong".
>didnt know Git
Even if they were CLI purists and refused to use something like GH Desktop, they'd only be using the same handful of commands most of the time unless something went wrong. How hard is it to learn just that?
You'll often find that people tend to prefer to use what they know works, if they don't have reason to look into an alternative.
According to the senior dev, he strongly prefers Perforce because of its Unreal engine integration, but Perforce was too expensive for an indie company. So he suggested using SVN instead of Git, which the CEO considered to be outdated and a downgrade from Git. Still the CEO felt he had to take a blind dive and trust the senior dev as it is hard to find someone on his level interested in working for his indie company.
You have to realize these people have at most written fizbuzz and edited some ricing configs so naturally version control seems like a meme to them.
You only need git for large products with multiple teams that work relatively independently from each other.
Also nice projection with your fizbuzz, newbie.
you should have version control for anything where changes might need to be undone or just having undelete protection. or conversely if you know you're going to do something experimental and don't want to waste time commenting and uncommenting code to control which parts are going to run between the production code and the experiment. you can just make a throwaway branch and easily merge updates into it, or not.
you've just never programmed anything worthwhile on your own.
>you've just never programmed anything worthwhile on your own.
I have, and I literally only use git to transfer code between my computers. You're projecting hard
>still hasn't named anything
yeah, didn't think so.
>larp
That's exactly what it is. The interviews are all bullshit
>oh actually we're dumb as frick lol
Like no you're not. You hired a bunch of Nintendo Interns. You know EXACTLY what you are doing.
The problem with people like you is that you make up wild stories in your head and treat them as if they're true.
Maybe read the interview yourself?
Git hasn't been good ever since the Sneedacity drama(look it up)
Git is not github you contrarian homosexual
using a command line tool in 2024 just because it's a command line tool is peak hipster soidev
there were better graphical alternatives 30 years ago
What's stopping you from using a visualizer for your git repo?
I use the inbuilt visual studio one, it's breddy gud
there's a bunch of git frontends, which is absolutely neccesary for comparing files, but it doesn't change the fact that the commands are not totally clear, and the frontends still use the language of git commands instead of clearer english
criticizing git always triggers the drone ants
I am begging you to look up what LARP stands for.
I guess it's time to try game devving, too bad got no Asian jeans to be actuall good.
Gitgays malding is always a joy to see.
>malding
That's not a real word, ESL-kun.
This moron saw that stupid meme discord screenshot huh? Yes, they did use git but switched over to SVN
>To be honest, I was a little hesitant to migrate the engine in the first place because companies that use svn these days have a legacy image. Compared to that, anything like a version control system is fine. Fully trusting his words, I also migrated my version control system from git to svn.
https://note-com.translate.goog/pocketpair/n/n54f674cccc40?_x_tr_sl=ja&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=wapp&_x_tr_hist=true#33883a37-8fdb-4f0d-ab06-853c8aa8be4a
Fricking mouthbreathing morons i swear to god. See an image on the internet and they think it's 100% authentic. Fricking monkeys.
>SVN
Just like Blow and other game devs.
I'm not entirely sure but I think it's because like Perforce it can handle art assets which Git more or less implodes on.
>The market wins again!
By promoting a game that looks like pokemon with Fortnite made by that man Nintendo should hire?
Yes.
But enough talking about Stardew Valley
Isn't that the guy who cried after some rapper made fun of his mario "subversion".
>Lunatics
making games the way god intended
big corporations with hundreds of useless drone jobs are a blight on gaming
Yep, I'm calling based on this one.
Truly exemplary behavior and I'm not kidding.
They hired a random japanese /k/ommando?
they hired some store clerk who had an interest in weapons
sounds like someone I'd find posting on /k/
>Xwitter cap
kys.
I don't get why they need to be buying new flash drives constantly.
that pic isnt even correct because the holocaust started midway through 1941 and ended in january 1945 when the camps were liberated. So 3 and a half years, 4698 israelites per day.
Israel is killing hundreds per day in the most in efficient way possible.
4k per day sounds extremely reasonable.
*inefficient, I meant.
are they turning the bodies to ash, including the bones?
They're destroying bodies, desecrating graves, and blowing up records of birth certificates etc. to facilitate genocide denial.
So I guess you're a fan of that kind of thing?
The fricking recettear whoring math keeps coming into my mind.
>shitto, fuccu, we needo programmeru, Shinji-san!
>donto worryu, Kaoruo-dono, i foundu sales clerku, he fold programmingu code over 9000 timesu
>game sells millions
This shit is beyond comedy. What are these western fricks doing if tatami-san over there on the island has this type of performance with hiring hobbyists?
People forgot that Lethal Company came out a few months ago and it was made by one Roblox furry. Those giant companies are useless.
did they seriously not use version control
HOW
it can be done if your devs are not moronic
But the only reason you'd ever do it is if your devs ARE moronic...
>yakuza substories.png
Not using Git is amateurish as frick. Only people who never worked developing software before would thinks it's a good idea not using it. Forget Linux; Git IS Linus Torvald's greatest accomplishment.
gamedevs dont use git because it's shit for handling assets
there's git lfs for that, which is also a total piece of shit to setup
git is supposed to be only for text files though
they use other version control systems that aren't git
Because it's not designed to handle binary files, only source code.
>ITT: "programmers" astonished that you can do shit in other ways than the corporate told them to
you love to see it
they did use version control though and whats worse they used the more corporate one
Whatever floats their boat.
>the market wins again
friendly reminder that most software projects written in the 80s and 90s were done without source control, and they still end up being some of the best things ever made for computers
nowadays all the tools get in the way of coding
Try testing program to work both on 10 Android versions and Iphones while running code on PC without Virtual Machines. In 80s and 90s barely anyone used tech and there was no need for it
thats the right way to use VMs
and I guess to segregate your dev environment from the rest of the system, cause then you don't have to uninstall a bunch of stuff and can easily restore things from snapshots if updating a tool broke things
but it is just one of many things in modern development that makes things slower
FRICK MOBILE DEVELOPMENT
ITT people don't understand the difference between half a dozen obsessed creatives banging out a game and a corporate made large scale projects
Now you could argue video games shouldn't be corporate made large scale projects and you'd be right but that's a different argument
While you've been arguing about version control. Lil' Taro has finished his new character controllers in UE4 blueprints, sent the files to his coworker who merged them all in the current build of the game.
They won in that it sold a lot of copies. As a game it's a barely stable janky pile of doodoo with moronic balance and some of the worst AI pathing and behavior I have ever seen. The amount of times an enemy has just stared ahead at me without moving or attacking is ludicrous, it's a complete failure of programming that is only excused because early access.
>you can download your games whenever you want as long as you want as long as you dont say Black person Black person Black person
This upsets Ganker because of course it does.
What do you mean they don't understand git? Like how initially setup one? Because I only commit, push and pull stuff.
everyone knows that
but do you know how to rebase a repo? switch branches, revert changes to files or change a repo location
WITHOUT GOOGLE?
and because the syntax is so convoluted, it's just an invitation for mistakes to happen that you won't realise for a while
also many git ui programs are still bad and it's not clear if git is actually doing what you want to do
there's just so much noise and it's so verbose
git is such a good way to filter out morons
they out themselves. with pride
git is more complicated than it needs to be
the complications are there if you need them
you probably don't
keep it simple
it just works
>keep it simple
Literally impossible with git.
Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
Couldn't you just do an interactive rebase (to the hash of the first commit), choose edit on the first commit, then change the message to what you want it, finish the rebase, then force push?
Unfortunately not. Editing the initial commit message is only possible through commands.
It gets even worse when trying to edit the author.
BZZZZZZT!
See pic related.
Googled it, you're right, my first idea wouldn't work, I'd have to use git rebase -i -root instead.
Still only had to check one stack overflow thread for that one though.
>edit
that's what "reword" is for.
git commit —amend
No, the first commit, not the latest commit.
>Try to amend the message of the first commit in any repository without checking five different stackoverflow threads. I repeat: literally impossible.
just make a new repo lmao
Unironically your best option if you don't have five years worth of experience in git.
if you have a five year old repo you have no reason to change the first commit message
it's a moronic scenario you made up in your head, literally just make a new repo if you made a typo in the first commit
>if you have a five year old repo you have no reason to change the first commit message
The email address I used when authoring that message is no longer in use.
I wish to change it to my new email address. Why is this not a valid use case according to you?
you're trying to revise history, the commit is five years old
i know your old email address has your deadname in it but you're just gonna have to live with it
>but you're just gonna have to live with it
Not a valid argument. Thank you for conceding.
I was going to say it's not that difficult, that one quick google search gave me the solution, but then I remembered that I have five years worth of experience in git, so that's not really a counterargument.
Have morons not heard of
git rebase --root
I swear Dunning-Krugers here...
Of course I haven't heard of git rebase --root. How often do you think I need to change the initial commit a repo?
Not often I imagine, but there's this thing called - manual, worth reading.
I'm not going to read through the git manual for fun.
The manual is for looking up specific thingsI don't know how to do.
that's not the only thing it's good for, it also tells what you can do and what workflows it applies to.
multiple times to maintain consistency
just study?
>just study?
If I want to do a single small thing, I am not going to read an entire book on the subject.
Instead, I will use something else. It's really as simple as that.
>git rebase --root
How the frick is anyone new to git supposed to know what any of that means?
Seriously, why can't we just do:
>git edit commit [commit id] -author "[email protected]" -message "Initial commit"
on any commit?
>worth reading
lmao, absolutely not.
You remind me of those unixtards who create thousands of manual pages and then get angry that no one wants to read them to learn how to do the simplest thing.
>unixtards
I actually am unixtard, but I don't get angry and I don't write manuals.
Didn't mean to imply all unixgays are the same unixtards who write manuals for fun.
I could've worded it better, my bad.
I know what you mean, considering nowadays most manuals are either complete shit just tells you visit this and this site instead. But generally manuals are useful and when use new software or library and need to look something up first thing I do is check whether there's a manual for it. In unix systems usually there's.
My point being, that if you're software developer - manual should be your best friend.
>Seriously, why can't we just do:
>git edit commit [commit id] -author "[email protected]" -message "Initial commit"
>on any commit?
git hashes depend on all preceding commits in the tree
if you amend a five year old commit, all the following commits are going to get new hashes, which will frick up any script and external link that point to a specific commit
git is trying to prevent you from doing something really fricking stupid
>git hashes depend on all preceding commits in the tree
What? Why? What could possibly be the justification for that moronic decision?
The hash is supposed to be a unique identifier of the commit. If it didn't depend on the history, then two commits with the same content but different histories would have equal hashes.
integrity of the source tree, the current HEAD is a product of every commit in the tree
if this wasn't the case, your link to 4e203fc today could point to different contents tomorrow
I don't see why you'd just not add a hidden identifier to the hash that can't be amended instead like a counter or a date added. (and treat merges as new ones)
As is the most obvious thing is prevented by design. I guess it makes sneaky bullshit harder in the particular case of a huge open source platform project like Linux but for most everyone else it seems deeply silly.
Don't forget that git was developed specifically for linux kernel source control, it was just later on adopted by normies.
And in kernel development history and tree integrity is extremely important.
>git notes add `git log --all --pretty=%h | tail -1` -m "message"
git uses directed acyclic graphs.
Of course you can't just change commits without breaking the whole structure.
It is indeed possible though either through rebasing (will frick over your whole revision)
or by doing a hack by creating a commit and using git replace
when there's a cheat sheet to basically translate git into something every english speaker understands, it makes you wonder why didn't they name the commands better
>set up a script in my ide to instantly commit to master whenever I press Ctrl+S
pshhht Nothing personnel "agile" software dev
nobody knows it without google at first.
you just use google until you internalize it but its not like you ever sit in a black box with no internet access to look shit up.
what im saying is even people who love git often dont know how to use git without googling
when you have 2000 employees all using git, people are going to end up making mistakes with git that end up eating time
it's easy to make mistakes with git
nta, but I heard about a game jam that took place on a cross-country train ride where the Internet kept going out, and so all the indie devs were running around panicked asking each other basic programming questions.
It honestly sounded kind of cool, having to do without google for a couple days.
>google how to revert in git
>simplest is do x
>careful when doing x. it fricked up mine.
>the correct way is doing y
>y is a convoluted chain of commands
Jesus, feels like being a git manager is a full time job.
they should probably have 1 git expert who handles all the tricky commands and controls merging, and just have the codemonkeys do push/pull/commit
>WITHOUT GOOGLE
What a weird and pointless criteria do anons program in a cave? You have access even there these days.
if you dont know git commands off the top of your head, you don't understand it intimately enough to know if you're making a mistake with it
https://slashdot.org/story/15/06/30/0058243
>But I really have not been involved. People like Greg and the actual graphics driver guys have been in much more direct contact with Valve. I think it's great to see gaming on Linux, but at the same time, I'm personally not really much of a gamer.
>I'm personally not really much of a gamer.
>I'm personally not really much of a gamer.
>I'm personally not really much of a gamer.
why the frick doesn't linus torvalds play videogames?
what does he think computers are for?
>what does he think computers are for?
Something to run the kernel on, everything else is a side effect.
Git is filled with troons anyway
Not using or understanding version control will slow you down massively. Even for projects you're working on individually it's still better than nothing because it makes regression much easier and bugs much faster to resolve.
Blow also doesn't wank off programming as an art moron, he's just as stupid as you and thinks "programming should solve real problems" which to him means use literally 0 abstraction ever until you're forced to because he's too stupid to understand when it's valuable and when it isn't
I just copy and paste folders and files
Why use version control
Because if your computer shits itself you won’t have to worry about losing your files.
automated backups
And if your hard drive corrupts?
Point is it’s just safer to store it on a server vs having it locally.
Development is doable without it but it’s a nice tool to have.
automated backups to dropbox
>VC == backup
lmao
Do you test after every single feature?
I'm guessing you're not that forward thinking or else you'd already use version control
So now imagine you implement a batch of features and then you find out that one of them introduces a critical bug.
But oops because you're a fricking moron and your development versioning is a straight line you can't revert that change without also reverting all changes which came after it.
So now you either roll back 2 weeks worth of work or spend half a week trying to find and fix everything that you changed.
Oh but frick now your hard drive with the 3 months of work between august and november is corrupted and there's no way to ever roll back to that patch.
And there's a community patch available to fix the long load times but you have literally no idea what it changed because your tooling of "just copy the files bro" can't identify the diff
There's literally no reason not to use version control. It is just a better and easier alternative to copy and pasting files that guarantees you always know where your history is and what each patch changed.
If I do a major refactor then I make a copy of the old code incase I need to revert by copy and pasting the files into a new folder
Why bother using a version control system
>refactor
I hate this word and I hate people that are obsessed with refactoring
Code needs to be refactored, it's just a fact of life
not all the time
you may need to tidy things up after your first draft so you can easily make changes later, but some people seem to get dopamine rushes from mindlessly editing code that already works well and can be easily modified without breaking anything
homosexual did you not read a single thing that I wrote?
Copying files doesn't allow you to revert individual commits you fricking moron, it only lets you revert back to a specific point in the history of a branch.
You also have to manually track what changed in each commit whereas a version control system would do that for you.
Copying files allows me to do anything I want. What's a commit? What's a branch? If I'm adding something to the program which could break it, I just make a copy
homosexual I will write this out for you in the most grade 3 hin-glish for your stupid ESL ass if I have to:
>You add feature 1 to file x.cpp
>You then add features 5, 7, 8, 15 etc. to file x.cpp
>Oh frick, feature 1 introduced a critical security vulnerability
How do I revert feature 1 without VC?
just remove feature 1?
lmao what is this braindeath
>just remove feature 1?
How?
You didn't even track what lines changed when you added feature 1 because you just copy pasted the files
This is compared to:
>Right click the commit number
>revert changes from this revision
With VC history
>You didn't even track what lines changed
If it's merely a matter of lines then should be able to remember, or look at the code and see what you changed and revert it
>should be able to remember
lol
>or look at the code and see what you changed
that's version control...
I don't understand what you're getting at here
If you're making a potentially codebreaking change, make a backup copy of the files, it's that simple
If you're making a potentially codebreaking change, just don't make it moron
>doesn't understand the point when he said "just check the difference between the two versions manually lol"
>which is LITERALLY what a version control system does automatically
You don't need to check the difference, just revert back to the copy you made
If it's a minor change you want to revert, try Ctrl+Z
That would also undo all of the changes you made after that point moron
So just go to the particular file you want to revert instead of the whole thing
That would still revert all of the changes to that file that happened later dumbass
Yes, it would
If all you need to change is just one fricking line in a file go and do it manually
And what if you need to revert the changes in 120 files dumbass?
Then go grab your previous backup you made and revert to it
That would revert all subsequent changes in all of those 120 files you fricking moron
Isn't that the point? You want to revert the changes you made and go back to a previous state from a certain point in time?
Yes that was my entire point.
That is why not having version control is bad.
Because with version control, you can just revert that commit and you keep all of the non-breaking changes which came after that point you fricking dumbass
So you're telling me you make changes to 120 files and then after that you don't actually test to see if it works or not?
NTA but this is the most amazing stupidity I've ever seen. Are you 14?
I have a suite of automated tests which run both before and after I push commits to trunk which don't require any input from me because I use VC.
I don't do manual full coverage tests after every single commit because they take too much time and if they fail I can just revert the commit that caused them in a single click because I use vc.
I can't test for every security vulnerability and neither can you because we don't know about every security vulnerability that exists moron. Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them and check if your dependencies were safe.
I don't need to test for security vunerabilities because I program video games not web servers
After I make something I test it to see if it works or not, that's the process, doing partial rollbacks on features like you're talking about happens so rarely
And the rare time it happens you save a massive amount of time.
No I don't
I copy and paste a file over, or I go into a file and copy and paste a function over
This ideal circumstance when you go and change features A, B, C, D, E, then you can just rollback A and it just works again even though you've changed all those other features sounds like a pipe dream. Usually the other changes you've made will be in a similar area so if you just roll back A it's probably not going to work fine with the new features you've introduced
>I copy and paste a file over, or I go into a file and copy and paste a function over
So you waste a bunch of time?
it takes seconds
>You do literally 0 regression testing? Holy fricking moron
What I mentioned is regression testing moron
You've failed to make substanial argument here
>after I make something I test to see if it works or not
You think that's regression testing?
Hallo sar
This is very funny coming from the guy who thinks buffer overflows are an obscure C bug only discovered recently
You're a little baby programmer
Not what I wrote
Doesn't change that you're a moron and completely wrong
You wrote this
> Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them
This is a hilariously wrong and naive statement, like factually wrong, not like the opinions we've been discussing here
You're a newbie
Anon I know your entire job is copy pasting replies from chatgpt but when you misquote other people's arguments against me you just expose yourself as a brainless ESL codemonkey.
And misunderstanding what I wrote to present it as an error doesn't change the fact that you think that integration testing is regression testing because you're a fricking moron
You don't know because the fact that they exist is widely publicised and modern programmers have access to good, easily searchable documentation and reference material from the internet which was not true historically. You would either have to read the docs yourself or hope whatever printed reference you were using mentioned the pitfall and that you understood it.
It is a very easy mistake to make, which is my entire point, but it's much less easy to make these days. Most IDEs and compilers will even warn you when you're making it these days.
I posted what you said verbatim, you don't know how C works at all buddy
regression testing is making sure the changes you made didn't break anything that already worked
>I posted what you said verbatim, you don't know how C works at all buddy
And that quote about people not knowing about buffer overflows is 100% accurate when they're likely the single most common C security vulnerability in production code.
>regression testing is making sure the changes you made didn't break anything that already worked
Huh weird that's literally the opposite of what you said earlier
>after I make something I test to see if it works or not
I guess chat gpt must have written that one too
>And that quote about people not knowing about buffer overflows is 100% accurate
No, it's not. Everyone knows what a buffer overflow is. You would know if you took a single class on C at school
>Huh weird that's literally the opposite of what you said earlier
You misunderstood what I said
You're an arrogant little webdev who thinks the entire world adheres to your extremely inefficient development practices
>Everyone knows what a buffer overflow is
Anon how old do you think C is? Did you take your first class in C in fricking 1973?
>You misunderstood what I said
HAHAHHAHAHAHHAHAHAAH fricking homosexual lmao
I didn't take any classes on C personally, I'm just saying, if YOU took a class on it, you would know what a buffer overflow is from semester one. It's not some super technical thing, you're just writing to memory somewhere you shouldn't be - which is extremely easy to do by mistake in C, which you would know if you had actually used the language
Every single C programmer in existence knew what a buffer overflow was moron, it's just an easy mistake to make. It's like off by one errors, you don't need to read the fricking standard to understand them, they just happen because they are easy to make.
Most people that started coding in C came from coding in fricking assembly for whatever architecture they were hired for back then, you think they didn't know what overwriting memory was?
People who came to C from assembly obviously knew what the underlying code was doing.
People who started with C didn't inherently and frequently fricked up because it's an error which happens silently and not something that most programmers write intentionally.
I feel like you homosexuals have literally never spoken to a single other programmer to understand how people actually learn. Most people just mash keys into the compiler until something works and then they stop testing.
>you would know what a buffer overflow is from semester one
Which I clearly do given that we're talking about it
But for some reason your esl ass thinks I don't because I'm saying it's easy to make and overlook?
Ok moron.
>Which I clearly do
You don't. You think you have to read the standard to know what it is. You have absolutely no idea
You can't bullshit your way out of this one bro
>You think you have to read the standard to know what it is
>You don't know because the fact that they exist is widely publicised and modern programmers have access to good, easily searchable documentation and reference material from the internet which was not true historically. You would either have to read the docs yourself or hope whatever printed reference you were using mentioned the pitfall and that you understood it.
Uh oh ESL bro
Looks like chat gpt missed out on that little bit of context
Historically everyone always fricking knew about buffer overflow attacks because it's incredibly fricking obvious to anyone who knows how to use C
Like I said, you can't bullshit your way out of this, just be a man and admit you didn't know what you were talking about and made something up
>Historically everyone always fricking knew about buffer overflow attacks because it's incredibly fricking obvious to anyone who knows how to use C
>source: 13 year old child who openly stated that he has never used C
lmao
I use C all the time, I'm 38
>I didn't take any classes on C
If you think really hard about it you should be able to figure out the difference between "I didn't take any classes on C" and "I use C"
>I didn't take any classes on C
>but the only way you could have possibly used C is by taking a class on it
>but I've also been using it since 1972
Sure bud
You can use C without taking a class on it
Yet somehow in this reply:
You seemed to think the only way to learn C was through classes.
C doesn't even have classes you fricking moron
Now you're just lying about what was said
Concession accepted
are you trolling or esl
>he didn't take dicky class
>easy mistake to make
Meh just turn off compiler options like shadow stack if it's really a big issue for you.
Not quoted.
It doesn't not take seconds to find a single function. Unless you want to pretend you have 100% of your code memorized at all times. You remember the extact name of every function in every file? Or are all the games you work on so simple you never change anything you worked on months ago? If you don't use VC, you enjoy wasting time
It takes seconds to copy some files over
Actually going into an old file and digging out some old functions you want to regress to happens pretty rarely, and when it does it doesn't take a long time to do (outside any updates you need to make to it, but VC doesn't save you from that either)
>After I make something I test it to see if it works or not
you can at least be 100% sure something that used to work isn't broken now if you have automated tests
playthroughs are fine for integration testing that automated tests cannot easily cover
and you will thank yourself if you port a game to another system, where many things can break
automated tests are kind of shit for game development
porting things to other systems really only affects about 1% of the codebase
for making sure the ui still works and doesn't get stuck in weird states, especially if features are added, then it's still useful there
though TDD is awful and should never be used for a project where the design evolves rapidly and requires tests to be rewritten constantly
the tests should just be to lock in the correct behaviours
this talk is interesting though
>porting things to other systems really only affects about 1% of the codebase
depends, what if the system you're porting to uses different endian, and you have bit shifting functions and comparisons in your code?
suddenly all those will break
but porting is much simpler now because everything runs on engines and you dont have to store floats as ints and other tricks on certain hardware
>what if the system you're porting to uses different endian
we're talking about things that actually happen in reality
it used to be a real thing that came up
now hardware and software is more standard, so porting is easier
The last time I remember anyone giving a shit about endians was in the 90s
Security vulnerabilities were just an example moron. It could be segfaults, memory leaks and other crashes could be introduced in this way.
>After I make something I test it to see if it works or not
You do literally 0 regression testing? Holy fricking moron
>doing partial rollbacks on features like you're talking about happens so rarely
And when it does happen it saves you a frick ton of time
It's also not the only or even biggest advantage of VC, but for some reason because you're a moron with a shovel you can't stop yourself from digging
>I don't need to test for security vunerabilities because I program video games not web servers
Hope you never make any multiplayer games.
The surface area for security vunerabilities in multiplayer games is very small
How's the pay at amazon, anon? I head the working conditions can be pretty crap.
That's completely contextual.
Not really, it applies to every multiplayer game, any code that can actually cause security problems is very small and very contained
You're making a hasty generalization with no basis. It completely depends on the implementation. There's countless stories of ACE, even from veteran developers like Valve in CS2.
I hope you don't expect your multiplayer games to be truly secure. Most of them aren't.
>Years ago you needed an extremely in depth understanding of the C/C++ standard to even know that buffer overflow attacks existed, let alone how to avoid them
Buffer overflows have been known about since C was invented, they're a very simple bug to understand
>Buffer overflows have been known about since C was invented
Sure, but you are significantly overestimating the average C and C++ programmer here if you think "known" means "widely known" or "widely understood".
Very few people have ever actually sat down and read the standard or the full definition of the methods they call.
>For actual programmers that's pretty large negative. Particularly once you try to step through it and you have five layers of indirection and can't keep watch windows of out of bound structs.
If state crosses the boundaries of structs you've fricked up the segregation of your components and they should be more strongly coupled.
Segregation like this should let you ignore what's going on above that method call and isolate errors to individual components that cause them. This is much better than having to keep track of the literally hundreds of attributes that a single class can have which potentially get mutated and affect how the program runs.
You do not need to read the standard to understand buffer overflows
It's a very simple mistake to make when you're programming in C
>You don't need to check the difference
>or look at the code and see what you changed
You were talking about manually checking the difference between two versions
You cannot walk back that post, no matter how hard you try. You accidentally got tangled in your own web of shitposting
Thanks for conceding.
>You cannot walk back that post
I have no idea what you're trying to get across. I very rarely cross-reference code to manually revery something, I was trying to understand this hypothetical situation being presented to me
is LITERALLY what a version control system does automatically
Ah yes, let me just check the difference between now and last week
>git -select-version $92ha4 -execute jkaq -rebase-into-now -into $2nc90a -timeframe-in-ms 298127398129
Oh shit, you accidentally deleted three random commits!
Don't worry, you can fix this by executing three other esoteric commands, but you will lose your latest changes... sorry!
>I'm too moronic to use a hammer without breaking my fingers, therefore hammers are bad
git log
git diff <commit from a week ago> HEAD
so hard
>git log
What the frick does this do? Why am I making a log when I want to compare things?
>git diff <commit from a week ago> HEAD
Why the frick do I have to write an extremely long string of random numbers and letters? What the frick does HEAD mean?
I can just open my project folder, open the folder of last week, open the file I want to compare and compare it with the already open file. Simple as.
Use SVN then you stupid monkey Black person, it's basically the "backup your file once a day" version control for homosexuals who are too stupid to copypaste a commit ID
Why would I when I can just make a copy of my code?
You gui plugin/program if you don't want to type commit ids, you moronic Black.
Ah so I need a tool to fix my tool so I can use a tool?
No thanks, I'll stick to making daily local backups.
Because between the change that broke things and your current version there might be 50 different changes you've done in the meantime. Version control will tell you exactly what was in the risky change, while your backup will show you all 51 changes forcing you to recall and manually reconstruct what you did. Meanwhile VC saves you time and headspace by doing it for you.
that is basically what VC does, except instead of making a 1GB copy of the project for each backup, it just tracks the 3kb diff of the code that actually changed, with 100% accuracy and not relying on your memory of what you did or didn't modify.
this thread has pretty much proven that people like you are just abject morons. it's not so much that you just lack an understanding of how VC is useful, you are just aggressively and willfully stubborn about it because you're too dumb to figure out how to use it.
Why the frick would you have 1GB of source code? It's text
If you make changes to code, don't test it, then just forget what you fricking wrote, you are a moron
>then should be able to remember
So how many lines of code are in this project you're managing without vc btw? 5?
>or look at the code and see what you changed and revert it
You're doing this manually? In 500+ LOC files?
Just remove feature 1.
lmao
If you introduced a bug, fix the bug
If you're making an addition / change that you might want to revert, make a backup copy
It's that easy
Give the folders descriptive names like "BeforeFeatureX"
how do you know which usb drive has the file/code you want?
Blow doesn't use Git either and tells people to frick off when they ask him why.
I use github desktop.
tell us the true man
is it sht ?
there's nothing wrong with it, and it's great to easily view diffs and select what you want to commit
don't care, still not buying your goyslop
um, those are illegal practices though?
ITT: people who think github is git.
that's "old ID software" level of based, no wonder the game is actually good
Programmers lack creativity, they will never make a good game.
Self taught doesn't mean not knowing real programming. Someone who has actually shipped hobby shit can be worth 10 CS graduates who have only done their course work. But neither is being a hobbyist guarantee of skill.
Here's my version control
>Click Myfolder (1)
>CTRL+C
>CTRL+V
>MyFolder (1) (1)
still not figured out how to tell windows to increment to (2)
>My project re01
>My project Final
>My project Final (1)
It just works.
>ragtag bunch of jap autists band together to make a janky crafting survival parody game because they felt like it
>twitter immediately collapses into a singularity from sheer seethe and goes out of their way to slander it
>actual AAA devs are actively shitting on it because "it feels cheated"
>same people who spent years complaining about Pokemon's lack of quality are now furiously bootlicking nintendo
>people donwright refusing to accept it is successful and make up lies about THE EVIL AI GENERATION
>same twitter that claims to support the "little guy" and hates big corporations
I don't know you guys, but this whole thing has been quite the eye opener
bunch of jap autists band together to make a janky crafting survival parody game because they felt like it
the same studio made a moderately successful craftopia 4 years earlier and palworld was advertised multiple times at major game events like tgs, you fell for a meme
>the same studio made a moderately successful craftopia 4 years earlier
>a game so successful no one talked about it
I see
Subversion bros….. we are so fricking back…….
Gitroons need a man in the middle program to do CTRL+C and CTRL+V in their 1 man project
git sucks, but it still works
honestly a lot of computing needs to be redesigned and written from scratch
there's simply too much shit
jonathan blow is right
Hi, Ganker.
What the frick is Git?
That is all, have a nice day.
https://git-scm.com/book/en/v2/Getting-Started-What-is-Git%3F
It's a tool to measure your productivity.
The more they see your name in Git, the more they think you're a cool programmer :^)
that's not how it should be used, but unfortunately it is often used that way
also most code shouldnt have been written in the first place
if your project has code in it, it's going to break
and the more code, the sooner it will
Thanks, that's exactly what I wanted to know.
>mouthbrrather is so dumb he introduces CVEs in his code because he pulled a third party library to compute if a number is even and log it with colors
>"YOU NEED VERSHUN c**tRUL!!"
>palworld instantly sells millions of copies
>meanwhile western devs are getting laid off by the thousands
curious
Nooo! my soulful nip devs- ACK
sirs, I have finished tracing the electabuzz
lmao this moron doesn't know how outsourcing works
>ignored that literally everyone on the core team is japanese
The only foreigners credited are some english VAs, some localizers and some outsourcing studio for 3D.
Any principle/strategy/design/whatever that has a catchy acronym as a name is worthless bullshit. There might be exceptions, but it's 100% true for anything *management related.
Your job turns 100 lines of code into 1000. You are ruining society and civilization.
I get the impression that this entire game hinges on them lucking out and finding the one really experienced dude who knew Unreal Engine inside out and basically became project lead in all but name.
it's dumb to use solid as a solo dev, no one else will use your code, you can modify everything whenever you want and you don't need a dozen unnecessary abstactions
must be OOP
>yeah you don't know anything, sorry
NTA but solid is explicitly OOP
>Oh but muh modules!
No
I'm gonna guess you're secure enough in your job that a bunch of college kids who don't know shit don't faze you, but just in case: You're 100% right
I work at a company with 500 devs, delivering a suite of software for over 30 years at this point
I have seen the old code that no one ever had time to refactor. Takes fricking 3 days of reading to understand a tiny part of it, whereas new code following proper patterns is easy to maintain
>how do I remove a feature
Find where you defined it.
Change the name to_getridofthisshit() and compile it.
Clean up all errors that occur.
Remove definition.
>Also has aggregator commenter brainrot
It's inoperable, doc
good morning sirs
This seems like the thread to ask: does anyone know a good recipe for rogan josh? Thank you sirs
do not be racist my bastard
>rogan josh
Damn, that looks good.
Thanks for letting me know this exists, anon.
it just looks like curry to me
Anon, you need to find yourself a good Indian restaurant.
Because once you taste real authentic curry, you will never want to go back to mashed potatoes.
>CHINKS !
>PAJEETAS !
>VATNIKS !
>PABLOS !
>YUROPOORS !
i wonder which skin colour typed this...
Well this comment was definitely written by a mutt mulatto
>comment
newBlack person
Well this snap was definitely tweeted by a homosexual
Nobody has called them "leafs" in decades.
Lurk moar
>pretending
whiter than you Trayvon
only a Black person can understand Black folkpeak so...
my condolences... I'm sorry you were born like that...
YIKES! imagine being poo poo coloured LMFAO
>all the hackernews senior software dev trannies post on Ganker
makes perfect sense
why do you think there's gamedev articles on HN like every other day
it's clear, all programmers learnt coding to make videogames
Where is the best place to build a Mediterranean castle?
in the mediterranean
I thought I was in /pal/...
Near the Mediterranean
wet dry world from super mario 64
hello sirs
c++ is a frick
oop is not compatible with gamedev
software design patterns do not apply to games
that is all
>he doesn't even ECS
ngmi
ECS is just a poor man's database. I use sqlite for my game engine.
NOOOOOO DONT CRITICIZE GIT VCS! I COULDNT DO MY HECKIN REMOTE CORPORATERINO JOB WITHOUT IT!!!
There's argument against VC unless you enjoy wasting time
I could actually, cause I just RDP onto the developer PC that's standing underneath my desk at the office
But without version control it gets kinda hard to coordinate with hundreds of other devs
If you need hundreds of devs you're doing something wrong.
Oblivion was made with about 40 programmers.
Half Life 2 had about 50.
It took about 50 developers 10 months to make Halo 2.
Rockstar needed 130 developers for GTA4.
Now dev studios have 600 employees and getting nowhere
>muh vidya
Don't care, only shit-tier devs work in vidya
free market is always better than corporativism pushed by political agendas like wokeness and DEI, funded by billionaires
nasdaq.com/solutions/corporate-esg-solutions
no act against leftists is immoral
I will always blame myself for obeying my moronic boomers and teachers instead of trying to learn myself
If I were a zoomer I would just learn to program and learn some useful skill
What kind of moron doesn't want version control? It's great when working alone, it's vital when working in a team.
I like it but I'm not going to pretend like it's necessary. ID didn't have a problem with multiple people working on the source code at once for Doom.
Git isn't the problem as much as working on a huge project with a large team is fundamentally awful
Git is often not ideal for a game project because the majority of the work usually involves art assets nowadays rather than code work. Having full version history for projects where the full WIP filesize can be easily in the TB range (high quality sculpts for bakes can be 100 mill + vertices) is legitimately awful.
I want to say the same.
I work in a big company as well though it's not mainly developing software but still it has a huge IT department with many applications.
Ignoring software development standards can be fine if you are just doing a one time gig that no one will care about updating in the future. Maybe that's what the palworld team had in mind as well (or it was just straight up cluelessness or manager's greed). But even they will notice if they are planning on keeping this game updated for some years, that technical debt is real and that they could have made their lives way easier (read 'develop cheaper') if they adhered to software development standards.
Sure, you are not supposed to read them like the Gospel and you need to tailor them to your needs but just straight up ignoring them will only hurt you in the long run.
This is also something that all the college kids in here will learn as soon as they have their first junior position because everyone hates fixing bad code that doesn't adhere to the necessary standards. So guess who gets all of those grunt tickets. Correct, the junior.
Oh, and not using Git (or svn for the boomers here) and a (at least basic) cicd system for software projects with any kind of relevance is just 1) torturing urself unnecessarily and 2) has a high chance of simply blocking your whole development operation at many occasions for stuff that could have been easily avoided by using version control and a cicd system.
>Jonathan blow
WHOOOOOOOOOP
The game is objectively fricking garbage. It's really annoying seeing twats pretend like Palworld is some sort of incredible success story instead of just a random fluke. Every other week streamers decide to shill a new game and it becomes the newest FOTM. There's a mountain of games like palworld that weren't so lucky and died stillborn.
Seethe, nintentroon
For frick's sake you idiots keep in mind game dev also involves people like envoirnment artists who have 0 programming knowledge (they don't need to) but do have a good taste when it comes to making something like a nice garden. Your grandma would probably excel in decorating a nice large open world but should absolutely be never allowed near git. She will accidentally mess up, work on an obsolete branch etc and potentially piss away weeks of her own work before it's discovered. A surprisingly large people working on video games on the artistic side are near completely tech illiterate. Your solution of version control needs to accomodate those kind of people which git doesn't.
And no, you're not going to hire an expensive programmer just to help decorate a nice video game envoirnment.
This looks AI made
Because it is.
Nobody would put their houses next to cliffs like that in mountain range. Not only would all the cold air rush down on them 24/7 but they would also get demolished by avalanches every winter.
uhh anon you know it's a real place right? Look up Lauterbrunnen in Switzerland.
Switzerland is AI made.
>passion first
>credentialism is worthless
The biggest players in the industry in the 80s and 90s already knew this. (before investpoors started buying up successful studios)
These days it is literally considered discrimination to say as a game dev that you'll only hire people who enjoy games. That was one of the things in the big California lawsuit against Blizzard, a woman said that she felt unwelcome because she didn't like games.
lmao thankfully i'm not american where they have reverse burden of proof over that shit as soon as you have more than 25 employees.
Just more proof you shouldn't open up an office in a non-corporate friendly state.
>Techro says made-up shit because plebs like stories of success about an underdog
SIR YOUR STAND-UP MEETING
SIR YOUR SCRUM HOW CAN YOU MAKE A GOOD GAME WITHOUT SCRUMS
SIR WHERE ARE YOUR SHREW MEETINGS ABOUT MEETINGS
Bureaucracy in a company isn't useless, it serves an important function: giving something to do for nepotism hires who are unable to contribute to the actual product
Every autistic male coder in a cubicle can subsidize the existence of 4-10 women who professionally go to meetings and dick around on a computer all day
Depends who and where. Having what's basically a janitor that keeps track of stuff so it doesn't get lost, prods people every once in a while for progress, delegates tasks and shields people in a team from bureaucratic nonsense from above is a good thing.
Programmers are morons
Why did Scrum and Kanban become synonymous with Agile instead of real shit like extreme programming?
>real shit
>being forced to babysit incompetents even more than I already do through pair programming
I will literally switch a job before I tolerate that shit and I have a relatively high tolerance for bullshit.
Bureaucracy makes EVERYTHING inefficient, whether it be Federal Bureaucracy or Corporate Bureaucracy. Palworld shows that a group of guys all on the same page are way more efficient and profitable than a 200+ horde of programmers, lower-middle managers, middle manager, upper-middle managers, project leads, HR ladies, scrum masters, and bean counters.
Did I fall into Ganker? Palworld????
Never ask a
>woman, her age
>man, his salary
>scrum master, what they do
git commit -m “xyz”
git push origin -f master
git merge -ours
git commit -m "stuff"
Frick you software devs for ruining plcs.
>software devs
More like management.
Actual devs rarely have a say in these things, outside of the recent trend of handmade rethoric seeping into the mainstream by calling out the most egregious moronation.
>MBAs saved vidja gaeming by suicide bombing all the big studios
What the frick is this?
Why all Dunning-Krugers of this board flock to this thread?
yes, it is. it's just extremely inefficient.
if it works it's not stupid
I didn't say it's stupid, I said it's inefficient.
>a bucket of usbs is version control
Then so is just maintaining a single working copy of the code
by definition it is not.
How is it not?
Your versioning is fully controlled, there is just one version
>In software engineering, version control (also known as revision control, source control, or source code management) is a class of systems responsible for managing changes to computer programs, documents, large web sites, or other collections of information. Version control is a component of software configuration management.
It is, definitionally, a system for managing changes
You delete your old branch and replace it
Lets agree to disagree.
>0 is not a number, it is the lack of number
NTA, but yes, that description is exactly what the digit 0 represents.
t. mathlet
Is that why they hired a shit load of jeets? How did these jeets manage to send their products if everything is stored locally?
They didn't hire any jeets. If you go through the credits all the actual employees are japanese.
They paid some 3D outsourcing work to a jeet studio.
Just like they paid some murricans for localization and voice acting. (which is the real reason why you have Type A and Type B in there)
>OH SHIT A POORLY MADE FOTM OPEN WORLD SURVIVAL GAME WITH DISCOUNT POKEYMANS MADE BANK???
>BECAUSE...IT FOLLOWED THE MOST LUCRATIVE POSSIBLE NICHE IN THE MARKET!?
>""""""""""""""FREE MARKET""""""""""""""""" CONFIRMED FOR BROS!!!111!
The most chilling part is that these are the people who don't know left from right and call anyone who's not a temporarily embarassed billionaire a communist and a useful idiot.
Then again
>Twitter thread
>seething over bros making money for doing something popular that isn't supplied by the big studios
>using terms like "temporarily embarrassed billionaire" and whining about conservatives
Go split your dick you massive homosexual.
>git rebase
more like, git recringe
Anyone complaining about using git or version control does not code. Pure and simple.
Or is jeet.
Just thought this needed clarification, anon.
I mean look at yanderedev.
The game is shit in general BUT he made a game and his code looks like he is an Indian studying CS in california.