Paradise brings on a couple more maintainers/Maintainer work day

[florodude]

As of right now there are currently 538 reported issues, and 86 pull requests.  For those not in the coding world, that means that there are 86 fixes, tweaks, features, etc. that have already been coded by somebody in our community and are in limbo.  My suggestion is that Paradise brings on a couple more maintainers/changes their practices on approving PRs/has a work day.  Looking at some of our PRs, many of them are very simple fixes or bugchanges that should have to wait 3-4 months in order to be processed.

 

I propose that a couple more maintainers that have regularly been proposing PRS are added to the maintainers team.  In addition, I think the team should take a look at their policy and practice and re-evaluate practices regarding how long things can take.  Finally, I propose the maintainers find a day where they can take a couple hours to get on voice chat of discord, and hammer things out together.  Looking at the PRs proposed, I think the team could get through quite a few of them in a few hours.

 

Finally, because I work in a full-time director position, I understand the frustration of people suggesting things but not actually doing anything to be a part of the solution. Therefore, I would volunteer to help maintain in any way the crew would need me.  I'm a novice coder, but learning.  If the crew doesn't need me, that's also fine, but I wanted to at least volunteer to be a part of the solution.  Thanks. 

[necaladun]

Surprisingly, more maints etc doesn't actually mean faster merges. More people to discuss things in general slows things down, as more people need to be consulted and agreed with. A lot of these would be much easier with less people.

The 538 reported issues isn't really something more maints would fix. That requires people to actually code fixes to, and then go through them. Anyone is able to do that, they don't need special roles to do so. A lot of them aren't even relevant anymore, and could easily be marked for closing if people just put the time in.

4 hours ago, florodude said:

Finally, I propose the maintainers find a day where they can take a couple hours to get on voice chat of discord

We prefer in general to use text, as that is logged and can be referred to more easily. We actually do have  days where we spend a couple of hours discussing these things, every week. 

Recently things have been a bit slower than we'd like, IRL has gotten in the way, and IRL>spessmen.

If you wish to help, I suggest going through the issues on the github, and/or helping others clean up their code and review it.

[florodude]

What about a task force under the maintenance team?  One that would look through the PRs, filter/categorize them to help things run faster.  Find which ones aren't really applicable anymore, which ones are just easy fixes, find which ones they would actually discuss.  That sort of thing.  Would that be helpful? 

[necaladun]

A task force isn't needed - anyone can comment on PRs, issues, etc. I encourage anyone who feels like doing so to do so.

[florodude]

Well, just trying to be helpful and be a part of the solution

[Ziiro]

/tg/, in an effort to clear up their issues, gave out an antag token for every issue that got fixed to the person that fixed it. Cleaned up their backlog pretty quick.

Options for non-monetary incentives are limited in this space, if there's a better one I don't know what it is.

Edited February 16, 2018 by Ziiro

[Bxil]

2 hours ago, Ziiro said:

/tg/, in an effort to clear up their issues, gave out an antag token for every issue that got fixed to the person that fixed it.

That's... That's actually genious.

[florodude]

Agree. I really like that. 

[Tayswift]

1 hour ago, Bxil said:

That's... That's actually genious.

I don't know about that. Treating antagonist tokens like currency isn't necessarily a good idea. It incentivizes gaming for antag tokens and is unfair to those who can't/don't have the time to learn how to code. And it leads to systems like this where antag tokens become like microtransaction rewards you get for playing certain roles. The most fair way to treat antagonist picking is to use the random number generator. As soon as antag tokens start being handed out for other reasons it turns into a currency that people will want.

[florodude]

I mean yeah some people don't have time to code but that is also kinda on where they're prioritizing their time. This way gives incentives to helping the community, which is a good thing. 

[Bxil]

1 hour ago, Tayswift said:

I don't know about that. Treating antagonist tokens like currency isn't necessarily a good idea. It incentivizes gaming for antag tokens and is unfair to those who can't/don't have the time to learn how to code. And it leads to systems like this where antag tokens become like microtransaction rewards you get for playing certain roles. The most fair way to treat antagonist picking is to use the random number generator. As soon as antag tokens start being handed out for other reasons it turns into a currency that people will want.

Is it only me who sees no problem with that? It creates no problems, if anything it encourages people to learn how to code, which I find a good thing. If people help you out, why wouldn't "pay" them for it?

[Calecute]

How much is an antag token worth in the antag token black market? 

Just kidding. 

I see no problem in paying people in antag tokens. "It incentivizes gaming for antag tokens" - this is exactly what we want by using antag tokens as payment. 

Also I'm not sure how antag tokens works, but I think even if it reduces chances of other people becoming antags depending on how many pull requests are accepted per unit of time the effect will be very small. And people are paying in reduced chance of antag for bug fixes and features, it can be worth it. 

Also I see absolutely no reason why it would lead to the proposed tg reward system, nor why this system is bad. 

But in my experience people are quite unreasonable about payment and fairness, we could use more worthless prizes, like Karma, or special snowflake loadout options. 

[Tayswift]

58 minutes ago, Calecute said:

I see no problem in paying people in antag tokens. "It incentivizes gaming for antag tokens" - this is exactly what we want by using antag tokens as payment. 

This is terrible in a number of ways. I can easily file a bunch of minor issues for every single spelling/grammar error I see in the game and then fix them one by one with a PR and rack up antag tokens. Hard to tackle, important issues will be ignored in favor of the best time-to-antag token ratio issues.

You say "learn to code" but you fail to realize that a lot of people are very busy with stuff that doesn't involve this game. Not everyone is willing to invest the time to learn how to use git and pick up coding.

This also rewards people who enjoy playing antag. Not everybody enjoys antag and I'm not sure it's a good idea to put antagerry on a pedestal as a reward people are scrambling over each other for.

Once an in-game effect is treated as a currency or commodity, it opens the doors to microtransaction-esque things like the PR I linked where you get rewarded for playing AI and punished for playing cargo tech.

[Bxil]

25 minutes ago, Tayswift said:

This is terrible in a number of ways. I can easily file a bunch of minor issues for every single spelling/grammar error I see in the game and then fix them one by one with a PR and rack up antag tokens. Hard to tackle, important issues will be ignored in favor of the best time-to-antag token ratio issues.

Don't hand out tokens for grammar fixes then.

26 minutes ago, Tayswift said:

You say "learn to code" but you fail to realize that a lot of people are very busy with stuff that doesn't involve this game. Not everyone is willing to invest the time to learn how to use git and pick up coding.

I see no problem with this.

26 minutes ago, Tayswift said:

This also rewards people who enjoy playing antag. Not everybody enjoys antag and I'm not sure it's a good idea to put antagerry on a pedestal as a reward people are scrambling over each other for.

I don't think having 5 fixes for the same issue is that much of a problem. Give the antag token to the first/best fix, then embrace the salt which affects no-one.

27 minutes ago, Tayswift said:

Once an in-game effect is treated as a currency or commodity, it opens the doors to microtransaction-esque things like the PR I linked where you get rewarded for playing AI and punished for playing cargo tech.

You just told me not everyone likes playing antag, thus this is not really a punishment to anyone. And no-one says we need that PR, nor that PR will ever get merged you showed.

[EvadableMoxie]

For people who don't like playing antag, nothing really changes. They aren't trying to get antag anyway, and getting an antag token they won't use is no different than right now, where they just get nothing.

For people who like antags but can't code, the odds they'll be an antag go down slightly since they are now competing with people with tokens.  However, they get to benefit from all those fixes that wouldn't otherwise have been done.  So things here are a wash at worst.

For people who like antags but can code, the system is obviously great for them.

 

Seems like at worst people break even here. Who exactly is hurt by this?

[Tayswift]

2 minutes ago, Bxil said:

Don't hand out tokens for grammar fixes then.

At this point, you may as well put a price in antag tokens on each issue then. Grammar issue: 0 antag tokens. Greys being able to wash their hands: 3.99 antag tokens. It's absurd and will be gamified. We should not be treating an in-game effect as a currency. Being antag should not be upheld as some sort of reward. We have enough antag fishing on the server as it is.

6 minutes ago, Bxil said:

You just told me not everyone likes playing antag, thus this is not really a punishment to anyone.

If you don't like playing antag, you can just turn off antag preferences. If you like playing antag, you will have to play AI/captain/whatever roles have the highest number of tickets to keep up with everyone else. It is a punishment for those who like to play antag.

[florodude]

Quote

We should not be treating an in-game effect as a currency.

Currency implies payment or trading.  Antag tokens shouldn't be able to do either. 

 

Quote

Greys being able to wash their hands: 3.99 antag tokens. It's absurd and will be gamified

That's taking the argument to a point of absurdity.  Each issue solved = 1 antag token, unless it's grammer/typo fix.  boom.  Legislation solved.   As for it being gamified, this is a game we're playing.  Who cares?

 

I think you are vastly misjudging the effect this will have on the game.  Each round is 2ish hours on average.  So let's say there's 12 round a day.By the end of the game there's at least 5 antags, but I'd guess the average is closer to about 8.  But let's stick with that minimum.  There's 60 antags a day.  How many antags will realistically get solved each day?  I'd say at absolute maximum 3. (In case you want to argue the point of maximum three, look at the other side of things.  If you really think more than 3 tokens a day would be redeemed, that means over 3 issues with our server are being solved a day, that's worth some praise!)  That means 57/60 antags every day are normal, and 3/60 are from solving coding issues.  That comes to about 5%.  The question I'd bring you all is, is 5% maximum of antags worth our backlog of issues being fixed?  In my opinion, heck yes.

[Shadeykins]

5 hours ago, EvadableMoxie said:

For people who don't like playing antag, nothing really changes. They aren't trying to get antag anyway, and getting an antag token they won't use is no different than right now, where they just get nothing.

For people who like antags but can't code, the odds they'll be an antag go down slightly since they are now competing with people with tokens.  However, they get to benefit from all those fixes that wouldn't otherwise have been done.  So things here are a wash at worst.

For people who like antags but can code, the system is obviously great for them.

 

Seems like at worst people break even here. Who exactly is hurt by this?

This is not how antag tokens work.

I would honestly be opposed to this. It requires admins or a maint to constantly note people for antag-tokens, and then it requires administrators to constantly hand out antagonist (which artificially raises the number of antagonists in a round).

Yes, this is how antag tokens actually function--it's literally a note on your account, you ahelp and ask for antag (traitor/vamp/changeling) and the administrator makes a judgement call in awarding it.

If we want to increase code contribution, we need to get serious about treating our contributors in a more even-handed fashion and actually hiring people into the coding team. There have been plenty of people over the ages who have contributed a tremendous amount to the codebase who were never made coders. IK3I, Purpose, Tzo, Flattest, Ziiro and Birdtalon all come to mind (regardless of how you feel about any of these people, they were all substantial contributors or competent coders). Mentoring new coders is something we should be doing as well, we don't really even have a good wiki page as an introduction to BYOND code (and we really should, but I sure as heck can't make it since I don't know BYOND coding).

A little thank-you goes a long way. For instance, wiki contributions nearly quintupled when contributors were awarded the praise they had long deserved via a reddit/forum/discord post and the wiki contributor tag was introduced.

As Neca said, the root of the problem is in the number of contributors - not with the number of maints.

Edited February 17, 2018 by Shadeykins

[necaladun]

Rather than on a per-PR basis, I think a reward for the top X number of contributors in Y amount of time could work. Make them compete for it!

[florodude]

That could be cool. But I think at least most of us are in agreement that some sort of reward should be offered for contributing on a regular basis? 

[Thamuel]

3 hours ago, necaladun said:

Rather than on a per-PR basis, I think a reward for the top X number of contributors in Y amount of time could work. Make them compete for it!

I have to agree with this idea. It would ensure that there is an award for contributing, but still being able to control the amount of antag tokens so that it doesn’t start becoming unfair. I support the general idea that an in-game award system would be put in place as long as it does not sacrifice the fairness of playing antagonist. Back to this idea in particular, it would also ensure that the coding will not be sub-par for the sake of collecting the tokens. A problem I can spot in this idea is who decides what piece of code is better than another. As someone who usually works with code, I can say that the quality of the code can be based of how many times the computer interacts with the RAM (in simple terms). The problem here is that it requires someone with coding knowledge to understand which code is better as it is not simply a measure of which code has less lines. Continuing the disadvantages, the second measure of how good code is, is how it is formatted and how easy it is to understand by other programmers, the problem here being that this is a subjective measure.

Also let me point out (correct me if i’m wrong) that antag tokens do not actually affect the antag rolls as they are used aftter the round has already started

Edited February 17, 2018 by Thamuel
Grammar mistake

[Calecute]

17 hours ago, Tayswift said:

This is terrible in a number of ways. I can easily file a bunch of minor issues for every single spelling/grammar error I see in the game and then fix them one by one with a PR and rack up antag tokens. Hard to tackle, important issues will be ignored in favor of the best time-to-antag token ratio issues.

 

I see, you think there'll be an incentive alignment issue. This will probably happen, but I'm not sure how prevalent it'll be. I mean, antag tokens are not that valuable. There are work arounds, but they will make the system much more complex as you point out in your other comment.

But I think antag tokens raising the number of antags and being admin time intensive to handle makes this idea much worse. 

Limiting the total amount of antag tokens given really solve this thou, but at the cost of paying less for the work done. But on the other hand, competitions sometimes are more effective at extracting work for little pay. I think we should at least consider given karma or some other small token of appreciation for every pull request instead of a competition, but I'm not sure what would be better. 

 

 

8 hours ago, Thamuel said:

A problem I can spot in this idea is who decides what piece of code is better than another. As someone who usually works with code, I can say that the quality of the code can be based of how many times the computer interacts with the RAM (in simple terms). The problem here is that it requires someone with coding knowledge to understand which code is better as it is not simply a measure of which code has less lines. Continuing the disadvantages, the second measure of how good code is, is how it is formatted and how easy it is to understand by other programmers, the problem here being that this is a subjective measure.

Well, the maintainers work is exactly to evaluate and approve code, so I don't think we'll have a problem with that. 

[IK3I]

So I think there is a deeper issue at work here in terms of the review process. As it stands, if dev is bottlenecked by a maintainer not having time to review code for an extended period of time, there should be failsafes in place such as authorized reviewers that can condense it into something they can meaningfully review or a temporary stand in that can act on their behalf for the time they take their leave.

While this is at its core a hobby we all indulge in, having no means of production for more than a few days at a time should not be an acceptable state. I don't know what process works best, but as it stands, its been a week since anything was merged even with simple bugfixes and features on the table for significantly longer that that time. Active development is the lifeblood of SS13 and it would be best to ensure it can continue even when the normal team is having issues.

[Allfd]

18 hours ago, IK3I said:

Active development is the lifeblood of SS13 and it would be best to ensure it can continue even when the normal team is having issues.

Rewarding fixes won't make a difference if they are never reviewed and merged.

[IK3I]

1 hour ago, Allfd said:

Rewarding fixes won't make a difference if they are never reviewed and merged.

My point was that more devs isn't the solution, but rather a more robust review process. As it stands, if a maintainer is having issues IRL, then the review and merge process dies until they can resume their duties. This is the least robust form of QC and creates situations like this where we have multiweek stretches of zero relevant changes.

 

Rewards should only ever be given out on a case by case basis and should never affect anything IC in a balance altering fashion. I'm simply suggesting that the admins and maintainers get together and assign trusted individuals to act in their stead should they encounter a situation IRL that would prevent them doing their duty for para.