Robustin's vain 2017 attempt to improve Coderbus for all

[Oldman Robustin]

This isn't a "HERES WHAT YOU CAN DO TO SERVE ME" post, this is a legitimate thoughtful attempt to cut down on the thousands of man hours that get wasted every year because of a dysfunctional level of communication between decision-makers. Hopefully the benefits are obvious on face, but let me spell it out: Responsibilities are delegated and automated so there's no more screeching at an individual for being a tyrant. The tyrant perception arises when the author/contributors see one maintainer acting unilaterally and assume that individual is overstepping their authority. When that authority is present by default it instantly improves the PR process. There's no more of this horrible peanut gallery where authors are left wondering "WHO HERE SHOULD I LISTEN TO? WHO IS IN CHARGE HERE? WHO IS EVEN A MAINTAINER?". Authors know who their target audience is and now it is infinitely easier for the maintainer to get express their concerns and for the author to listen to those concerns and address them. Both sides can proceed with confidence in knowing exactly where the other side stands and without worrying about some other maintainer intervening or interfering with a review. This process would improve transparency and accountability while minimizing the crippling paralysis and/or bikeshedding that seems to overtake a PR as soon maintainers with differing views get involved in the same PR. For your consideration:

1) Maintainers get assigned assign themselves to PR's, the initial choice to merge a PR is left to their sole discretion (there was no mandatory assignments but the wording was confusing). If the powers that be do not trust a maintainer's sole discretion then that person should either not be a maintainer or they should be assigned to PR's where their judgment is not compromised. Other maintainers can assist with the reviewing and offer their input but they should not demand or threaten the assigned maintainer. The choice to merge will belong to the assigned maintainer alone. Coderbus can decide the basis for assigning PR's but given that the initial reviewer has significantly more weight, the system should allow for 24-48 hours for "supportive" maintainers to assign themselves. Supportive doesnt mean guarantee merge, you're open to the possibility if reasonable conditions are met. If nobody is supportive of the PR within the first volunteer window then a hostile maintainer can intervene and proceed as they see fit.

2) Because the assigned maintainer can exercise judgment freely, they can quickly offer assessments to the PR's author as to what obstacles there are to merging without having to worry about being attacked or overthrown by a more prominent maintainer. There is a process in #4 for another outcome but it is handled quickly, dispassionately, and automatically, not through weeks of bickering and second-guessing what other people think.

3) If the assigned maintainer sees no path forward then that maintainer can close the PR and no other individual may re-open it unless there is a clear error. The closing maintainer should give a serious, brief, and sober assessment on the fatal flaws of the PR.

4) If the closed PR author feels they haven't been given a fair shake they can appeal to an IRC bot responsible for tracking and managing the appeals. The IRC bot will take a short statement from the appellant and automatically attach the closing message/reason from the maintainer. Upon joining IRC all maintainers will get a message linking the PR and the brief statements from both sides. While it would be impossible to removal all politics from this process there should be an understanding that maintainers, coders, whoever should not sit around and try to ambush maintainers as they log in and persuade them to vote a certain way. Any statements that people want to make about the fate of an appealed PR can be made on the PR page, in public, as part of the open record, and let the merits of their input lie there where they can be assessed fairly.

5) Votes will be yes/no/abstain. Yes votes may be appended with a sentence indicating qualifying conditions that may be decisive of their support. Qualifications should be just that smaller issues that do not dominate the content or purpose of the PR, if a maintainer opposes the bulk of a PR and wants to see radical changes, that should be a no vote and they can add their desired changes in a PR comment or PM to the author if they so desire. Defer simply adds your vote count to whoever you defer to. If Coderbus decides that certain individuals should be accorded more weight to their votes, that is fine, but it should be completely transparent and part of the public record. Headcoder can immediately reject or approve any vote but that power should be exercised sparingly. If a maintainer is too busy/doesn't care/whatever they can always abstain.

6) Coderbus can decide if there needs to be a majority vote, supermajority, or even a minority percentage of support necessary to re-open an appealed PR, but that approach must be consistent across all PR appeals. If a vote passes then there is a 48 hour window where any willing maintainer can take the lead on moving the PR forward. If no maintainer volunteers within that window, one of the yes votes will be randomly assigned. The new assigned maintainer will get any of the short statements attached to the other yes votes and should try to take them into consideration when suggesting changes, but it is ultimately left to the new assigned maintainer's discretion. If the new maintainer decides to close or merge a PR, then it is their sole discretion and completely final. If the vote fails then the PR gets a tag/comment for APPEAL FAILED and that PR stays closed.

7) This general process has exceptions. Any maintainer can still "leapfrog" an assigned maintainer to quickly merge a PR, but this should limited to fixes and non-controversial subjects. PR's that are reverts or removals should be tagged as such by the assigned maintainer. Removal/Revert PR's will proceed as normal with the exception that any maintainer can immediately call a vote on the PR to subject it to an immediate up/down vote. The vote will simply appear as "[Author] wishes to revert [SUBJECT] under the review of [ASSIGNED MAINTAINER] in [PR#]: Do you support this action?". That vote proceeds as an appeal would except that the PR will immediately close upon a negative vote and the assigned maintainer, regardless of personal preference, should continue with the removal upon a positive vote. Revert PR's cannot be opened within 2 weeks of the original merging unless there the problem relates to a bug or performance issue. Coderbus can break this rule for other occasions but there should be a good reason for it (overwhelming player rejection, completely unbalanced and game-ruining,

8) Each member of the community has access to one non-stacking appeal per 14 days. If they are unsatisfied they should either wait until their next appeal, make a new PR that substantively addresses the reasons for the prior's closure, or you can just use your appeals more sparingly if you care that much about a particular PR.

I'm tired of describing the countless shittery that has taken place in my PR's. I think even my most adament opponents wouldn't disagree that the treatment of my work was deeply flawed. I made a similar thread about the long wait times last year after getting thoroughly fucked through my cult PR's last March/April and while I doubt my suggestions had a direct impact, it appears Coderbus finally stood up and addressed the issue. One of the biggest draws in getting me back to coding was seeing that PR's were getting quick, decisive treatment and its been a huge improvement to have a manageable PR list rather than 4 PR's sitting uncommented for weeks on Page 5. Yet my latest PR has put other dysfunctions with our system front and center. It's clear that despite improvements, there are still gaps in the system that lead to absolute clusterfucks, and even where there aren't clusterfucks there are still many monuments to wasted time and energy because of a complete breakdown in communications between maintainers.

[PKPenguin321]

Can you make this more concise? Just installing git is too scary for most would-be contributors, and a massive manual of DA RULES like this won't help any

[Oldman Robustin]

These arent rules for contributors. This is to help the maintainers get their shit together. It would be a blessing for new contributors because the process is infinitely more intuitive. So and So shows up, makes it clear that they control the fate of the PR, and things instantly get MUCH clearer. Ive been here a while and im still having to guess if certain people are maintainers. As for the appeal, just have any closing comment link the IRC so they can figure out if they want to do next.

The current system is just a fucking circus and nobody benefits from the chaos. I could give a million examples of how ive seen maintainers needlessly waste each others time, but i think even they are smart enough to see the dysfunction for what it is.

[Haevacht]

tl;dr robustin is salty

[kevinz000]

making_friends_with_maintainers.jpg

edit: ok time to have a subsection of fnr titled pr appeals.

[oranges]

This wont' work because we're not paying the maintainers so they're not required to review a pr if they dont' want to, they only review things they want to and they don't have to put any more effort in than they want.

Your idea is dead out of the gate, it's also absurdly stupid in terms of the level of overhead and pointless democracy, but it looks like you already knew that based on your title and just wanted to get something of your chest.

your problem is you show up without building any trust from maintainers, with a giant ass pr, asking questions about code that make it seem like you're barely capable of understanding what you're doing and then whine when nobody shows up to review it even though it's a monstrous amount of code of dubious quality

The right way to do that is to make small changes, build up trust with a maintainer or two, be present in coderbus and then when you have a proven track record you can make big feature pr's and get them through the gate without too much drama.

[oranges]

Also regarding the delay on pr's the only difference there is we hired some newere maintainers who are little more active, so it's nothing to do with anything you said and if nobody likes a pr it still tends to sit there for a long time.

[oranges]

actually that's not quite true, because I'll often close a pr if it's been open for a week or more with no maintainer activity.

[Remie Richards]

>My PR caused issues
>My next PR caused issues
>Clearly the system is wrong and not me!

I'll support one maintainer getting control of a PR's fate only if I can have that dumb Identity PR (and consequently end it immediately, of course)
But in reality: no.

[lntigracy]

Why must your every post be a long drawn out essay

[MisterPerson]

Sometimes you're going to get a situation where 2 maintainers hate an idea and 2 maintainers like an idea. Basically you're just trying to solve that problem. Everything on top of that is just overkill.

[Qbopper]

i feel like i may have encouraged this because i was asking what the situation with robustin's pr was and if maintainers were doing something wrong in the other thread

i'm sorry

[lntigracy]

Who decides what happens in that situation?

Does Joan just close the PR?

[Oldman Robustin]

When I pressed Kor on why he shut down my PR 2 days after he told me I needed to find ~one~ maintainer to support it, his answer was basically "It took me a month to track down enough maintainers to get their feedback".

@MrPerson: Its not just that, its also the fact that absolutely nobody felt they had any authority over my PR. oranges was apparently too bashful after being de-maintered to do anything except shit talk me. Kor didn't want to be a tyrant but simultaneously gave me mixed messages of support and doubt. Joan's only comment was about hating the rewards yet she ended up closing the system's PR and trying to merge only the rewards. However the shadow those three cast over the PR within the first 30 minutes of it being open meant that nobody else was willing to give me a firm opinion on it. The only other maintainer that gave their input was Jordie a few days later who didn't like the rewards. Hearing "rewards" in 3/4 of the maintainer's comments led me to spend a month on rewards only to get meme'd and have maintainers try to merge my rewards but completely killed my system with no serious discussion and no warning. Maybe its because only the most powerful maintainers even try to code big features anymore (cause or effect?) that this isn't a more noticeable problem, but then again its only been the addition of big features that have kept this game alive and relatively fresh after all these years.

oranges: The assignment doesn't have to be against anyone's will. 48 hours for supportive maintainers to take the reins, after that the vultures can swoop in and either kill it or demand fundamental changes.

You claim there's "pointless overhead" but the biggest overhead of all is the endless fucking paralysis by maintainers who don't want to overstep into someone's elses territory. Not just with my engineering PR, but the most common response I get for discussing merger of ANY of my PR's is "What does Kor think? Does Kor approve of this?" Yet Kor, for better or worse, is absolutely loathe to direct the maintainers. It seems like everyone would be happier if they could manage the review and merger of a PR without the shadow of some "high echelon" coder dropping in and closing it while bullying any maintainer who supported it. You say its too democratic but you can't even explain what the current system is, nobody can, its an amorphous mess. I've heard that Kor is the ultimate decider of new features, yet Kor himself doesn't want that kind of responsibility and defers to other maintainers in a way that resembles a democratic process. I saw no less than 3 occasions since I've started coding again where you ask a maintainer what they think, they ask what Kor thinks, Kor asks what the maintainers think, and then there's a subsequent clusterfuck of opinions that can tip one way or the other just depending on who happens to be logged on at the moment. This system turns that shitshow into an occasional vote where the maintainer's can quickly resolve disputes without the endless fucking game of telephone.

The massive advantage here is there's clearly defined expectations for all parties involved. It will open up channels of communication between maintainers and authors and people can pitch ideas before a PR and move forward with confidence if they find a sponsor - rather than both the author and the sponsor living under the shadow of a more hostile maintainer and everyone dithers for weeks because of the fear that they will get in trouble for merging something Kor doesn't like.

Not only do I get a quick resolution on where I should be spending my time, but I also have very clear directions on what I can improve if an idea flounders.

There's really no downside here, maintainers can reach agreements on certain things, like Kor having domain over all Planetstation PR's. There's still more than enough soft power in play for molding the system to particular needs or desires, but this spares authors and maintainers alike from the paralysis and wasted time that occurs when I talk to a maintainer about my PR, they point to Kor, Kor points at a piece of abtract art written in a lost language, the art has a caption that says "talk to Cheridan", Cheridan links me to some indescribable fetish site which has an abundance of ads for Florida Oranges. I ask for Oranges, coleslaw tells me to kill myself and that its "oranges", then Shadowlight steps in and closes the PR, I talk to Shadowlight and do everything he asks only to find out he's not even a maintainer and has no authority to even advance my ideas, and then I die of alcohol poisoning because there's not enough booze in the world to make this process tolerable.

[Oldman Robustin]

>My PR caused issues
>My next PR caused issues
>Clearly the system is wrong and not me!

I'll support one maintainer getting control of a PR's fate only if I can have that dumb Identity PR (and consequently end it immediately, of course)
But in reality: no.

If you think the Cult PR was anything but blatant Coderbus politics then I don't know how to bring you back to reality. For my engineering PR, even the people who wanted it dead the most wouldn't doubt that it was handled in just about the worst way possible.

I'm the longest consecutive maintainer still working, and I have no more power than any of the rest of them.
This is how it's always been, and it's fine.

Maybe you've been a maintainer, and one that avoids major balance/features discussion, for so long you've forgotten what a shitshow it can be for the rest of us.

Your last big feature was Hand of God, and both the PR and your single attempt to balance it got merged within 48 hours of you finishing the PR. You've done great things for this game but I don't think you have a fair perspective on what it takes to merge big features when you're not already a maintainer. When you did a your first (and only) HoG balance pass you got 2 comments before it was merged. I brought cult back from the dead but when I tried to give it a fresh coat of paint I had Joan open up a competing PR within 24 hours and it took almost a month, several forum threads, and 100 commits to get merged. It even ends in typical Coderbus fashion:

phil235 commented on Apr 28, 2016 • edited

97 commits

Whoever merges this, please don't forget to do squash&merge.

KorPhaeron commented on Apr 28, 2016

@phil235

Are you saying it's okay to merge

@optimumtact
Member
optimumtact commented on Apr 28, 2016

are you kidding kor, this is one of the most reviewed pr's I've seen thanks to a certain maintainers Bias

@Razharas
Contributor
Razharas commented on Apr 29, 2016

Ok its time to clean pr list so this gets in



There was also this beauty that finally killed my will to code new features:

https://github.com/tgstation/tgstation/pull/17180

Kor comments on the 2nd day of the PR with >Buffing Hunters. I undertake 2 weeks of work, widely approved, addressing all sorts of concerns. Then Kor strolls in on Day 15 and shuts it down on the same reasoning. Kor had the opportunity to give an ultimatum on Day 2: I think xenos are already very strong and I will close this unless you nerf hunters. Instead I get meme arrows and 2 weeks of wasted work because, just like today, I don't get a definitive answer until the PR is done, suddenly closed, and someone reveals that it was never getting merged in the first place. Would it really have been that hard for Kor to say "This won't get merged until you nerf hunters, if you can't make that commitment then I will close this PR"? No, instead we both lose. Xenos continued down the slide of non-maintenance into a completely broken state and I wasted weeks of work because I only got an ultimatum once the PR was already dead.

[Xhuis]

Everything you said in the post above is true. I don't agree with some of what you've done but the conduct of Coderbus and the repository as a whole has been laughable at best for a while now. There seems to be a clear divide where maintainer​s either abstain from doing anything whatsoever, even commenting, thus forsaking their entire purpose, or pass judgement on a whim, although with some exception they neither play the game nor do they make notable contributions to it other than balance changes and speedmerged reverts.

It really is saddening to see it in such a state and it's one of the reasons I'm no longer coding. What is the point of having someone as a maintainer if they shirk their duties to the "design lead" who doesn't even want to do it?

[captain sawrge]

[bandit]

OK, so, I'm going to be blunt with you. The following does not represent my opinion of you, just what is going on:

When people don't like someone but have no real way to do anything about it, then they will put that person through endless hoops and bureaucracy and stalling and misdirection just to fuck with them for laughs. This is what is going on here with you. The reason you can't make sense of what you were told about the maintainer system is because it isn't supposed to make sense. The reason you're not getting anywhere with your argument is because everyone has treated these two threads as shitposts upon arrival, both in the thread and in IRC.

Whether this is good or dysfunctional operating practice is left as an exercise to the reader.

[Xhuis]

If a person's well meaning contribution is stalled or outright halted because the maintainers have a personal bias against its author, I think we need new rules or new maintainers.

[onleavedontatme]

then they will put that person through endless hoops and bureaucracy and stalling and misdirection just to fuck with them for laughs.

The only stalling anyone did was giving a hard no to avoid being at the center of this stupid drama that has now spanned 10 PRs and 2 forum threads and OOC and IRC for days.

Every "hoop" he was asked to jump through has been clearly outlined for months if not years in so we're playing a pretty fucking long game if asking him to do the exact same things we ask everyone else to do (no copypaste, atomic PRs, etc) was all to "fuck with him for laughs" (We actually ask people to do those things to prevent the codebase from collapsing into garbage like so many downstreams do before they give up and rebase from us)

https://github.com/tgstation/tgstation/ ... IBUTING.md

[captain sawrge]

If a person's well meaning contribution is stalled or outright halted because the maintainers have a personal bias against its author, I think we need new rules or new maintainers.

[bandit]

Except this stupid drama would have been avoided a lot faster by saying no up front, closing the PR, no further work asked. The band-aid is ripped off, instead of plunged into an acid bath.

[Xhuis]

haHAA im relevant guys haHAA

snip snop pudding pop

Code quality is of course a different story. I can vomit out a gamemode in a day (I should know, I've done it) but no matter how good it is it's still shitty if the code doesn't work.

The endless stalling is another road entirely. Goof is an example of this. Because of his consistently terrible behavior, his actual PRs are often stalled or mocked because "it's goof." That seems unacceptable and petty for a codebase as good as ours.

[onleavedontatme]

Because of his consistently terrible behavior

If someone makes themselves unpleasant to work with it seems exceedingly unfair that they get to cry bias when nobody wants to work with them.

Code quality is of course a different story.

Robustin considered atomizing his PRs and not copypasting code to be jumping through hoops, or at least that is the impression I got from his comments in IRC and github over the past month or whatever.

[onleavedontatme]

Except this stupid drama would have been avoided a lot faster by saying no up front, closing the PR, no further work asked. The band-aid is ripped off, instead of plunged into an acid bath.

Yes I acknowledged we should have done that and apologized for it about half a dozen times now.

[Xhuis]

If code quality is hoops, than he's climbing a cliff. As for individual bias, I think that it should be put aside when a common goal is finally pursued. While I may personally dislike you, I volunteered for planet station because I consider it a good cause and my own disagreements with you as a person aren't relevant.

[bman]

if i was robustin i'd be pissed off too i was led around in circles to avoid being told no and then bickering maintainers who just do the same circles thing then in the end he gets a closed pr and 6 "sorrys"

just saying

[Oldman Robustin]

I had no issue with the concept of atomizing, my issue was only being told that after the PR was finished. Its not complicated.

>Open very simple, narrowly focused PR purely for testing

>Nah this is shit, add decent rewards or you can fuck right off!

>Slowly add rewards, like one every 3 days pursuant to the design thread I had attached in the PR header

>Get code reviews but no calls for atomizing, atomizing goes unmentioned

>PR finally done, ready for testing, the only thing I wanted in the first place: "Ok btw u need to atomize"

>Are you sure? I mean cant we test it first?

>Yea lets go ahead and test it at least

>Shadowlight out of fucking nowhere: "CLOSED, ATOMIZE"

The hoops weren't just code quality, outside of the occasional nitpicking I will follow every code review comment that I can.

You say you're sorry but even now you can't explain why Joan was allowed to close my PR's based on an outright falsehood, or why you lied to me on Friday, or why sometimes you go into PR's like clock cult and say:

"This is gonna get merged obviously, I just think a complicated team role that players get to try maybe once-twice a month is gonna suffer, and blaming players for being stupid wont fix that."

That was a PR that not only got tested, but testing had revealed huge fundamental flaws in the mode. Yet my PR was cockblocked before testing even began and you, on rare occasions, talk to me and say things like "I think this will improve the game" and "I think players will enjoy this" while expressing vague reservations about its impact on gameplay design. When you opened that "FUTURE OF MEDBAY" meme PR I explicitly reached out to you via PM because I wanted to know then and there if I was wasting my time. I couldn't have been more straightforward, I wanted to know, from your own mouth, whether there was any hope for a PR that you saw fit to mock so openly and oranges was so set on destroying. It was that evening that you reassured me, you thought my idea had merit, my PR wasn't hopeless. I really wish I had saved those comments but I trusted you, just like I trusted you again on Friday. I reached out to you time and time again trying to understand my position. Who do I need to persuade? What is the criteria for this getting added? Silence. I wasn't worth your time.

When are you wearing your """"design lead"""" hat so that you can comment on one PR saying that you think a game mode is going to suck (and you were quite prescient there) but you're not even gong to try and stop it, but then you surf into my Engineering PR and make conflicting statements of support and doubt only to spring out at the very end and go "JUST KIDDING, not only do I oppose this PR, I'm gong to make sure no maintainer ever helps you merge it!" Even Joan's only explanation was "Kor changed his mind, Razharas doesn't have enough influence to get your PR merged". I found out within 15 minutes of closure that Dunc, who didn't even know what my PR consisted was, was also willing to help me. But even now I'm in this absurd position of "Well if I go listen to Raz and Dunc and make some changes that they support, is that enough to get it re-opened? Do I have to go to Kor or Cheridan first and get a commitment that if they re-open it that you won't immediately re-close it?"

Which is why I've turned to oranges of all people because despite him being the most hard-line about any new job features, at least he cares enough to give me a straight answer about what my PR needs to change. All of this just reinforces that there needs to be change, I don't care who holds the power, but there desperately needs to be some clarity about how things are run because it stifles contributors and it paralyzes maintainers when nobody knows who's calling the shots.

[Xhuis]

While maintainers like to give the illusion of coherently discussing things and decisions as a group, as far as I can tell (emphasis on that) it's entirely self-sufficient. A PR with 200 comments can be closed by someone who just saw it and didn't bother to read any of them before deciding that he didn't like the premise. Coderbus (the IRC) is a complete joke. People just spew memes constantly and deflect actual concerns with insults or abstinence. Kor is maybe the only person who actually cares about it.

[Remie Richards]

<rant>

your single attempt to balance it ... your first (and only) HoG balance pass

I don't know if you're doing this intentionally, but you're making out this was a bad thing.
The first round of feedback, I listened to, and I did the balance.
The rest of the feedback was that the mere CONCEPT was rotten to the core, so instead of limping the gamemode on for a year or so, I took it out back and shot it, the literal best course of action there.
The majority no longer cared for the idea after seeing it in practice.

Also don't talk to me about up hill battles and fighting against maintainers.
My first 2-3 major features were all done prior to being a maintainer, having to put up with some pretty vague reviews including Aran's infamous "no"'s with no explanation.
Just like you I started with no experience, and 5 years on I'm a software dev for a living, so it's not like we didn't start on even footing either.

You may not like to hear it, but merge rates and maintainers actually being willing to teach and help you are actually higher than they've ever been.

As was said before, start smaller.
My first PR was some fireaxe balance, I changed one number.
The next adding "midbullet3" to the game, to buff the damage of the stechkin without buffing c20rs in the process.
A couple of fixes here and there, welding tanks and turrets not deleting properly.
Not long after that I did my first major feature, augmentation.

If you start small, and build rapport with maintainers before jumping on big things like reviving entire gamemodes and "fixing" entire departments.
The dance machine alone would have been a decent start.

</rant>

[Xhuis]

How is that a rant? You didn't even call him the scum of the earth and that he should kill himself. You must be new here.

Jokes aside, I think the point he's trying to press is that you shouldn't need to build rapport as you put it. In an ideal world, your work would be judged on its merit, not the merit of the creator.

And unfortunately, with how long it's taken us to get here it may be a long time coming, or reliant on individual maintainers. I respect you for it because you don't pull any punches. But you're one of the only ones. That's a problem and it needs to be addressed.

[Wyzack]

In the real world people are paid actual money to deal with problems on the scale of these and therefore is not even remotely applicable to a system of volunteers

[Xhuis]

Au contraire, that's a larger reason why we need change. If you volunteer you are choosing to do it because you care about it. That's the definition of a volunteer. I'd be absolutely okay with becoming a maintainer under the conditions I outlined. I've coded for three years, so experience isn't an issue. Asking people to behave like adults is not an unreasonable request.

[Oldman Robustin]

Wyzack I have plenty of issues with certain maintainer personalities, but I made this thread with a solution that was completely agnostic to what I thought of certain people. I proposed a system thats a win-win for everyone if its executed properly. I believe Remie when they say that things have improved, but that's not an excuse for accepting the current state of affairs for the rest of time either. There's volunteers on both sides of the equation and this is an improvement for both groups.

@Remie

I can at least appreciate your willingness to recognize a mistake and put it down quickly. My comparison was only that in spite of very clear warning signs about HOG, you faced 0 resistance in getting it tested and 0 resistance in getting it merged. Nobody pulled the line about "well I don't want to test this because the core concept won't change regardless" (Kor from my last PR) to well we can't risk adding this on a trial basis because it "cant easily be removed" (oranges from my last PR). The level of scrutiny and skepticism was incredibly low despite the fact that not a single early test (to my recollection) went well. However, despite the fact that I think HOG was pretty awful, I'm still glad we tested it and saw exactly what did and didn't work, what the flaws were, what was fun about it, and try to learn from the experience. That's also why I'm still fuming that Kor, despite his uncertainty about how my mode would play out, felt that it wasn't even worth testing. I've seen Kor make the same comment about "Well endlessly theorycrafting is silly, how about we merge it and find out?" countless times. Kor didn't even doubt that my PR would be popular during testing but was perfectly content to let theorycrafting trump any need to actually test to see if negative side-effects would come to fruition.

As for your 2nd comment. This is was my 100th PR. My first PR was "Reduced Splash Volume of Thrown Reagant Containers", since I didn't know how to code ALL I could do was "little stuff" for months. "Slows down nutrition loss in Slime People", "Gives Pest Spray an item state", "Reduces shadowling burn damage by 3". I only opened my first big PR because I found out that cult had become newcult, which had (exactly as I had predicted) turned out to be the least popular mode in our history and was removed from rotation. I kept asking for someone to bring oldcult back and eventually Kor or some other coder said "well why don't you do it yourself then?". From there I got a little more ambitious, I took Sleeping Carp out of Gang and gave it to traitors... I still did silly stuff like the Staff of Honk but I grew increasingly confident that for what I lacked in coding ability I possessed in thousands and thousands of hours of game knowledge and my views on whats fun and not fun are highly in tune with the playerbase.

When I came back this year I immediately noticed game-breaking stuff like minebots doing 40 damage indoors or Xenos effectively doubling their health vs. lasers, stuff that was adversely affecting rounds and made those my first priorities for fixing. Last month I saw 3 issues on the rise: The absolute shitshow that DA had devolved into, the utter dominance of the ebow, and the fact that Clock Cult meta had settled in a particularly bad place. I have consistently been able to spot flaws in our game (last month: ebows, DA, clock cult) before the playerbase forms a consensus about it. I look through my PR history and I see a 2016-2017 Hall of Fame for cancer removal. There's not a single PR in there that I have regrets about. My previous biggest PRs, the Cult reintroduction, the Cult simplifcation, and the Cult runed structures has done something that has eluded the entire coderbus for years: Control/create a game mode and leave it in a better condition than you found it. HOG, Shadowlings, Gang War, Blob, Newcult, Malf, Clock Cult, Wizard - I see these names and all I can recall is a history of frustrating changes that only underscored how out of touch the developers felt to me. I'm not asking for praise or trying to promote myself, but as a newcomer who barely understood code, I took a stance that was the opposite of what other team modes were taking. I streamlined and simplified while other team modes were charging down the path of deep, complex mechanics that required incredible teamwork and coordination to pull off. I know there's a few cult-haters on the maintainer team but I doubt even they can argue that simplification wasn't the correct approach. In January the cult win rate was exactly 50% and in February it was 50.9%, when I came back in January and began my second overhaul one of the first comments in my feedback thread was "NO DONT TOUCH IT ITS PERFECT". I had hoped that I would had gained credibility, that after a year of making unquestionably positive changes with virtually no controversy about those changes after merging, that I would have some confidence, some leeway, something closer to what the Coderbus regulars and maintainers have that gives them some deference and some "well lets give it a chance" attitudes from those in charge.

No, that didn't happen. I've been around long enough that people have a hardened view of who I am and anything and everything I do will be filtered through a cognitive bias that will let people see whatever they want to see in me. After pouring my heart into this PR I routinely see comments that the system will go unmaintained... like I'm just going to cast this thing off as soon as its merged and wash my hands of it. If after 100 PR's I haven't seen a difference, I doubt I ever will. I'd rather just see some reformation so that we can all do our work better, smarter, and nobody has to go through the hell I've experienced ever again.

[PKPenguin321]

Security's representation on github 2: Electrical engineering boogaloo

[Qbopper]

my views on whats fun and not fun are highly in tune with the playerbase.

look, no offense, but if at any point in your life, with anything, you feel that you're totally in tune with a group of people and you know what they want, you're in too deep

your views on what you believe is fun/not fun, or even your views on what you believe others think is fun/not fun will never hold up to the majority in reality, that is how people work

just because you claim to have called X Y and Z as being issues early on doesn't mean a whole lot as a defense to me

unrelated: dude look I know you get harassed for this all the time but did you not have an English teacher that taught you how to be concise? So many of your points can be distilled to much shorter paragraphs and sentences, it's resulting in people not reading your posts and dismissing you because lol essays ecks dee

none of this is really on topic though, so on topic: like Kor has said, it's been realized that not making things clear early on was a massive mistake, what else needs to be addressed here

[Yoshmaster]

Better catch up on all this new reading material

[Haevacht]

Atomise out the nuke, get rid of it already.

[TheColdTurtle]

Do I have to cite this thread if I use it in an essay?

[oranges]

unrelated: dude look I know you get harassed for this all the time but did you not have an English teacher that taught you how to be concise? So many of your points can be distilled to much shorter paragraphs and sentences, it's resulting in people not reading your posts and dismissing you because lol essays ecks dee

He's a lawyer they can't help it they have to be explicit in legal contracts, every lawyer I've met types like that, I honestly suspect fwoosh or oxford are too.

Back on topic: anyone who thinks getting stuff merged isn't a social game is going to eat shit all day long, you do have to suck up to the maintainers, you do have to listen to what they say and do it, you do have to get them to back you up, if a maintainer doesn't like you they can make your day very hard and if you tick them off enough will be very very very hard to undo the mess you end up in, especially if you alienate all of them.

[oranges]

Also robustin the size of your ego is also preventing you getting anywhere.

[lntigracy]

unrelated: dude look I know you get harassed for this all the time but did you not have an English teacher that taught you how to be concise? So many of your points can be distilled to much shorter paragraphs and sentences, it's resulting in people not reading your posts and dismissing you because lol essays ecks dee

He's a lawyer they can't help it they have to be explicit in legal contracts, every lawyer I've met types like that, I honestly suspect fwoosh or oxford are ttoo

Also see dezzmont as another example.

Also /vg/'s code works the same but in reverse, we don't assume you're incompetent until you show repeatedly that you're incompetent, everyone is treated the same from the get-go until they do something that makes their future stuff have to be more thoroughly reviewed.

Haha goofball.

[oranges]

free goof

[kevinz000]

pretty sure i'm on maintainer's list of people to delay prs just so everyone can make sure it doesn't break shit.
hmmmm...

[oranges]

well if you do make pr's that have a history of breaking things you will get black marks

[danno]

It makes me smile to think of the time you put into this nonsense when you could have done literally ANYTHING else

[cedarbridge]

unrelated: dude look I know you get harassed for this all the time but did you not have an English teacher that taught you how to be concise? So many of your points can be distilled to much shorter paragraphs and sentences, it's resulting in people not reading your posts and dismissing you because lol essays ecks dee

He's a lawyer they can't help it they have to be explicit in legal contracts, every lawyer I've met types like that, I honestly suspect fwoosh or oxford are too.

Back on topic: anyone who thinks getting stuff merged isn't a social game is going to eat shit all day long, you do have to suck up to the maintainers, you do have to listen to what they say and do it, you do have to get them to back you up, if a maintainer doesn't like you they can make your day very hard and if you tick them off enough will be very very very hard to undo the mess you end up in, especially if you alienate all of them.

I'm in the same profession. Completeness is not the opposite of being concise.

[Haevacht]

It makes me smile to think of the time you put into this nonsense when you could have done literally ANYTHING else

Like remove the construction nuke.

[Qbopper]

unrelated: dude look I know you get harassed for this all the time but did you not have an English teacher that taught you how to be concise? So many of your points can be distilled to much shorter paragraphs and sentences, it's resulting in people not reading your posts and dismissing you because lol essays ecks dee

He's a lawyer they can't help it they have to be explicit in legal contracts, every lawyer I've met types like that, I honestly suspect fwoosh or oxford are too.

Back on topic: anyone who thinks getting stuff merged isn't a social game is going to eat shit all day long, you do have to suck up to the maintainers, you do have to listen to what they say and do it, you do have to get them to back you up, if a maintainer doesn't like you they can make your day very hard and if you tick them off enough will be very very very hard to undo the mess you end up in, especially if you alienate all of them.

I'm in the same profession. Completeness is not the opposite of being concise.

Cedar makes my point for me, just because legal documents might often feature longwinded text doesn't mean you can't be concise otherwise

This is mostly off topic though and I doubt robustin is willing to listen to any feedback on it because of the meme status it's reached where everyone posts about ecks dee essays, so we should probably not have this conversation here

[Oldman Robustin]

I type at length because my adversarial stance means people will fill in any blank I leave with some strawman to derail things with.

Also because I wanted this to be a sincere suggestion and not just a "fuck the system" thread, so details are helpful.

Alternatively I could've just proposed that all Kor comments on non-maintainer PR's get a massive oversized tag that appears whenever the comment doesn't start with "I like this" or "This is a good idea" and the tag reads:

ATTENTION: KOR HAS EXPRESSED AMBIVALENCE AND/OR NEGATIVITY TO YOUR IDEA. YOUR PR WILL NOW STALL AND BE CLOSED WITH NO WARNING IN 2 WEEKS. CONTACT YOUR NEAREST KOR IMMEDIATELY OR CLOSE THIS PR TO SPARE YOURSELF.

Cheridan is the only person who can remove the tag.