/tg/Station 13

Currently, the method for appointing new headcoders is that a retiring head chooses his or her replacement. There has been no restriction on this, meaning that tomorrow I could retire and appoint Tonto, or Twobeard, or Hitler's ghost.
Moving forward, appointments must be approved by simple majority of the maintainership.

Secondly, headcoders and maintainers should meet at least a minimum level of activity. I believe there is a need for a concrete measurement of activity, so that an otherwise inactive head cannot hide behind statements of "But I'm on irc/forums/twitter all the time!"
Right now, this level has been set at a bare minimum of 5 contributions per month, as tracked by the user's Github page.


If you have any comments or suggested changes, feel free to reply -- these rules, especially in regards to voting and contribution minimums are still subject to change.

>implying hg isn't hitler's ghost

The ectoplasmic residue test was inconclusive!

Gunna latefag here. Contributions of actual code are a poor metric of activity as it would be easy to abuse.

The historic problem with headcoders is that they just afk in coderbus, never actually show leadership, and just contribute code like a coder.

A better metric would be somehow tracking if they are giving the project enough direction and leadership, as that is what they are ultimately for and is what separates them from just a regular coder.

I'd like to chime in personally that so far during my entire stay here I have never at any point seen headcoders actually lead the game in any direction or do anything at all. The absolute most I've ever seen done is calling for a feature freeze and that really is it.

There no yardstick to measure leadership. As long as a head is active enough to detect problems and engage them proactively that is enough. The historic problem with headcoders is not that they 'just contribute code', it was that they didn't do anything at all ever. If you have contributions on the github tracker, it means you are at least giving something to the project. Again, the problem we're trying to address here is heads who hide behind statements of "But I'm on irc all the time (because I have my client set to startup with windows and autojoin)!"

As far as leading the game's direction, I do not see that as my responsibility. This is an open-source volunteer project, and I think contributors should be free to explore whatever possibilities they want within reason. I'm a representative, not a dictator. Still, I'm very often asked for my opinion on design decisions and for assistance with coding, which I'm always happy to provide. I do push for things, and sometimes something will come out of it. The most recent example is splitting the map's z-levels into separate map files, to reduce map edit conflicts and necessary maintenance.
I'm also always monitoring the activity of code contributors to see who would make good maintainers. Currently we have an issue of having too many PRs for our maintainership to handle, so they're tending to stay open longer than they should and not getting reviewed well enough to catch bugs. It's difficult to address, because it's not a problem I can just throw people at. They have to have experience with code standards, git wizardry, and work well in a team or else it just makes more problems to deal with.

To be told that you've never seen me "do anything at all" is frankly the most insulted I've ever felt in my time here. Thanks.

In support of Cheridan I'd like to point out that group discussions often do result in things been changed at our request, in a way that isn't quantified anywhere. Paprika did us all a favor and redid the map with full windows more or less at my behest, for example. Tkdrg put in a lot of great work on turfs, even an attempt at making walls objects at my suggestion. All said and done, however, we don't have the luxury of a team required to do 40 hours a week each on whatever we mandate. I'd like to think I've done a lot to make the game feel more realistic and dangerous, and I think that lead has been followed some by our contributors. Cheridan is right that maintaining the repo itself is a lot of work, especially given that we also want to contribute ourselves.

I just feel that your metric makes no real sense, both from the perspective of viewing the head coders as leaders or as maintainers of code quality. It is hardly a step up from the old requirement of "have a pulse" and is an empty gesture.

And your right, there is no yardstick for leadership. It is hard to measure in a way that allows a leader to make unpopular decisions. But it is still worth noting that "make 5 changes to the code" is really as meaningless as "I am on all the time." You should try to figure out a better metric or, worst case scenario, figure out a way for people to agree that the head coder is not active enough as a head coder somehow, because while some previous headcoders went dark both in code and leadership, others just borderline abandoned their position while still contributing code.

WJohnston wrote:I'd like to chime in personally that so far during my entire stay here I have never at any point seen headcoders actually lead the game in any direction or do anything at all. The absolute most I've ever seen done is calling for a feature freeze and that really is it.

Feature freezes are arguably the most important thing a headcoder can do, because it is the only thing that has historically been able to force the project back onto the rails. It is their ultimate weapon as maintainers and quality control, which is arguably their most important job.

I also agree that feature freezes are the only way to keep us on track.

On the other hand, cracking down on people using the issue report section (which coders use to measure the success of feature freezes) as a soapbox/feature request center/balance opinion lobby is also pretty important

In accordance to the policies outlined in the OP, Hornygranny has been removed from his position of headcoder due to inactivity.

HG has died of old age, this is truly the end of an era.

Does this mean HG gets deadminned, too? Does this mean we'll have ERP back?

Miauw wrote:>implying hg isn't hitler's ghost

YOU'RE HITLER'S GHOST!

Or at least Himmler's Ghost. Like a discount Hitler.

This does not have any impact upon his adminship. Admins do not have a written inactivity policy.

ERP never ever.

Cheridan wrote:This does not have any impact upon his adminship. Admins do not have a written inactivity policy.

ERP never ever.

Its not written but inactive admins get removed from time to time.

It tends to happen in waves where inactive admins just get removed all at once. Its somewhat uncommon but when it does rarely happen it can be interesting to see such a large amount of people being tossed out.

E: The last great inactive admin purge happened over a month ago actually.