Add a meritocracy #377

PlasmaPower · 2017-05-29T23:52:55Z

I decided to not go with a clickbait title this time 😄

Look at #361 for more details, but I tried to give the meritocracy minimal power. Democracy still decides the outcome >90% of the time. This is just to prevent against sneaky take over attempts.

Basis of how it works:

At least one positive non-author meritocracy review on the most recent commit of a PR is required for that PR to be accepted.

Changes from #361:

Autogenerate meritocracy list (currently top 10 contributors + top 10 voters)
Don't count reviews as votes (so members of the meritocracy can allow a PR without voting for it)

mark-i-m · 2017-05-30T00:06:41Z

For anyone else looking through this PR, it's a lot easier to read if you append ?w=1 to the URL, which gets rid of whitespace changes (e.g. indentation changes).

PlasmaPower · 2017-05-30T00:08:16Z

@mark-i-m yes definitely do that

hopes nobody notices my trojan horse coded in Whitespace

phil-r · 2017-05-30T00:26:59Z

👍 What can possibly go wrong?

mark-i-m · 2017-05-30T00:14:07Z

github_api/voting.py

-        if vote and vote_owner != pr_owner:
-            votes[vote_owner] = vote
+    for vote_owner, is_current, vote in get_pr_review_votes(api, urn, pr):
+        if (vote > 0 and is_current and vote_owner != pr_owner


Should this be vote_owner.lower() != pr_owner.lower()? i.e. are we sure the cases match?

That's how it was, and it works since they both come from the GitHub API.

Ok, that's what I thought

PlasmaPower · 2017-05-30T00:36:08Z

@phil-r your comment keeps duplicating client-side:

No idea why. It seems to be a GitHub bug. The bug survives a refresh, although the duplicated comment versions don't.

Edit: posting another comment seems to have fixed it.

phil-r · 2017-05-30T00:39:45Z

@PlasmaPower I've also noticed, refresh helped, maybe I've triggered some easter egg? 🐰 🥚

PlasmaPower · 2017-05-30T00:40:53Z

@phil-r I think it's actually a bug with reviews. It only seems to happen when the comment is next to a review. Overall, reviews are poorly implemented IMO.

PlasmaPower · 2017-05-30T02:22:51Z

When you fix a minor mistake but it resets the voting window 😢

I guess #378 wins the merge conflict race.

mark-i-m · 2017-05-30T02:24:18Z

Hey! 🎉

mark-i-m · 2017-05-30T02:25:48Z

What was the fix?

PlasmaPower · 2017-05-30T02:27:03Z

@mark-i-m In patch.py I cached gh_api.prs.get_contributors not gh_api.repos.get_contributors.

mark-i-m · 2017-05-30T02:29:54Z

Btw @PlasmaPower you down voted your PR by crying in the post above...

PlasmaPower · 2017-05-30T02:30:17Z

@mark-i-m lol oops. 👍 I think that fixes it?

Edit: yep

PlasmaPower · 2017-05-30T02:31:40Z

Reminds me of when reactions to comments counted for the PR as a whole.

PlasmaPower · 2017-05-30T02:35:44Z

Had to do another slight fix: .iteritems() was removed from Python 3. Changed to just .items()

mark-i-m · 2017-05-30T02:38:45Z

@PlasmaPower Are you squashing the commits in your own fork?

PlasmaPower · 2017-05-30T02:41:15Z

@mark-i-m Oh yeah, I probably shouldn't now that chaosbot squashes them for us. Doesn't make much of a difference though, and there's no added security - I could always amend a malicious change to a previous commit.

mark-i-m · 2017-05-30T02:43:03Z

I have my 👁️ on you

PlasmaPower · 2017-05-30T02:44:11Z

waits until @mark-i-m goes to sleep /s

PlasmaPower · 2017-05-30T02:44:26Z

Or do you sleep with that eye open?

mark-i-m · 2017-05-30T02:45:27Z

You don't know if I sleep (at the same time as you)

PlasmaPower · 2017-05-30T04:49:57Z

I just realized I forgot to pass reverse=True to sort, so it was an opposite meritocracy... oops! Will be fixed soon, just double checking a few things.

davidak · 2017-05-30T04:52:19Z

@PlasmaPower this solution looks very good!

So at least one member of this group has to do a review seperately from voting. So people needs to know if they are in this group and that it's helpful if they review. Also that they have responsibility for the project doing it.

It would be good if you add that to the readme and we need some kind of news section to inform the community of such important changes.

davidak · 2017-05-30T04:57:03Z

@PlasmaPower If this get's merged you can mention everyone currently in the list so they know that they now have this opportunity.

PlasmaPower · 2017-05-30T05:08:47Z

Current meritocracy: @chaosbot, @muchzer, @pchauncey, @amoffat, @hongaar, @rhengles, @rudehn, @xyproto, @eukaryote31, @ad-m, @mark-i-m, @Smittyvb, @reddraggone9, @PlasmaPower, @phil-r

I've updated the README, pushing the changes in a second. The meritocracy will be logged every time chaosbot checks for new PRs.

(yes I know chaosbot is in the list, but hardcoding chaosbot is both unnecessary and probably a bad idea. chaosbot will also fade from the list as merge commits give way to squashes)

PlasmaPower · 2017-05-30T05:15:29Z

Okay, both of those changes have been pushed. Ready to merge! 🎉

Other slight updates can wait for another PR, I'll only add on important fixes.

(now to wait 9 hours because the extended voting window is broken... 😴)

PlasmaPower · 2017-05-30T12:24:41Z

Rebased - fixed settings.py conflicts.

PlasmaPower · 2017-05-30T12:29:36Z

The Travis failure appears to be from master.

Edit: yep, fixed in #390

andrewda · 2017-05-30T13:51:18Z

@PlasmaPower

@andrewda I agree, the reason I didn't implement it now is because it'd be the first dynamic part of the server.

It doesn't have to be dynamic. How I currently have it written is to update a meritocracy.json every time it's generated, just like how the voters.json works. Both approaches are a little bit gross, but we can work on refactoring that and/or using Django in another PR.

tarunbatra · 2017-05-30T14:42:37Z

This is something nice, @PlasmaPower. Though it chooses the same set of ppl, more or less, but look at the response. 😄

PlasmaPower · 2017-05-30T15:20:06Z

@tarunbatra It's also nice that it automatically updates itself. I might adjust the config later though.

PlasmaPower · 2017-05-30T16:25:30Z

How I currently have it written is to update a meritocracy.json every time it's generated, just like how the voters.json works.

The difference there is that voters.json needs to survive restarts, but meritocracy.json doesn't (it'd be essentially write-only). I think a dynamic server is the better solution.

andrewda · 2017-05-30T16:26:24Z

I completely agree

chaosbot · 2017-05-30T16:32:06Z

🙆‍♀️ PR passed with a vote of 30 for and 1 against, with a weighted total of 27.5 and a threshold of 6.5.

See merge-commit 3e6d99c for more details.

hongaar · 2017-05-30T16:35:33Z

well well well... another regime change by @PlasmaPower!

andrewda · 2017-05-30T16:36:42Z

For some reason, this duplicated the object in voters.json and now it's invalid. This breaks the voters page.

{ "amoffat": 12468, "g19fanatic": 1442, ... }{ "amoffat": 12468, "g19fanatic": 1442, ... }

PlasmaPower · 2017-05-30T16:38:21Z

That might also remove the top 10 voters part of the meritocracy. Anybody know what's causing it?

andrewda · 2017-05-30T16:38:54Z

It's also crashing repeatedly http://chaosthebot.com:8081/chaos-stderr.log

PlasmaPower · 2017-05-30T16:39:16Z

Oh I think that's caused by two PRs being merged at the same time. I'll get a fix ready.

commit 9f4af54 Author: Lee Bousfield <ljbousfield@gmail.com> Date: Tue May 30 10:45:49 2017 -0600 Fix voters.json logic (broken by #377)

rudehn · 2017-05-30T21:46:10Z

So what happens when somebody who is in the meritocracy reviews the PR, but falls out of the meritocracy before the PR is merged? Does the review no longer count?

andrewda · 2017-05-30T21:47:23Z

I believe that the next time the PR gets polled, the review will no longer count.

rudehn · 2017-05-30T21:49:23Z

Is that working as intended, or a bug?

PlasmaPower · 2017-05-30T21:50:21Z

That's how it works, and there's really no way to fix it unless we get the database up and running (even then it'd be a bit of a pain).

mdcfe · 2017-05-30T21:50:47Z

@PlasmaPower This is supposed to be chaos - who knows?

@Ayyko

#392: Update README in anticipation of #377 Description: :ok_woman: PR passed with a vote of 20 for and 0 against, a weighted total of 19.5 and a threshold of 6.5, and a current meritocracy review. Vote record: @Ayyko: 1 @Leigende: 1 @MUCHZER: 1 @Mursaat: 1 @Smittyvb: 1 @SylvainThrd: 1 @andrewda: 1 @cthulhuely: 1 @davidak: 1 @eukaryote31: 1 @flibustier: 1 @hongaar: 1 @ike709: 1 @md678685: 1 @mikesmiffy128: 1 @phil-r: 1 @reddraggone9: 1 @rhengles: 1 @rudehn: 1 @viktorsec: 1

PlasmaPower force-pushed the meritocracy branch from f6b7885 to 402256f Compare May 29, 2017 23:57

PlasmaPower mentioned this pull request May 29, 2017

Get rid of the annoying extra whitespace in the middle of the message #378

Closed

mark-i-m reviewed May 30, 2017

View reviewed changes

PlasmaPower force-pushed the meritocracy branch from 402256f to 38be09b Compare May 30, 2017 02:21

PlasmaPower force-pushed the meritocracy branch from 38be09b to 20fabfc Compare May 30, 2017 02:35

README: reviews no longer count as votes

fcbaf91

PlasmaPower force-pushed the meritocracy branch from 67de294 to fcbaf91 Compare May 30, 2017 12:24

chaosbot removed the conflicts label May 30, 2017

Early break on meritocracy satisfied

d2323e2

PlasmaPower mentioned this pull request May 30, 2017

Fix flake8 (and therefore Travis) #390

Merged

hongaar added a commit to hongaar/chaos that referenced this pull request May 30, 2017

Update README in anticipation of Chaosthebot#377.

eac39b0

PlasmaPower mentioned this pull request May 30, 2017

[WIP] Refactor voters persistence #391

Closed

3 tasks

chaosbot merged commit 3e6d99c into Chaosthebot:master May 30, 2017

chaosbot added the accepted label May 30, 2017

PlasmaPower added a commit to PlasmaPower/chaos that referenced this pull request May 30, 2017

Fix voters.json logic (broken by Chaosthebot#377)

9f4af54

chaosbot added a commit that referenced this pull request May 30, 2017

Squashed commit of the following:

51e16a8

commit 9f4af54 Author: Lee Bousfield <ljbousfield@gmail.com> Date: Tue May 30 10:45:49 2017 -0600 Fix voters.json logic (broken by #377)

Add a meritocracy #377

Add a meritocracy #377

Conversation

PlasmaPower commented May 29, 2017 • edited Loading

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017

phil-r commented May 30, 2017

mark-i-m May 30, 2017

Choose a reason for hiding this comment

PlasmaPower May 30, 2017

Choose a reason for hiding this comment

mark-i-m May 30, 2017

Choose a reason for hiding this comment

PlasmaPower commented May 30, 2017 • edited Loading

phil-r commented May 30, 2017

PlasmaPower commented May 30, 2017

PlasmaPower commented May 30, 2017

mark-i-m commented May 30, 2017

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017 • edited Loading

PlasmaPower commented May 30, 2017 • edited Loading

PlasmaPower commented May 30, 2017

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017

PlasmaPower commented May 30, 2017

mark-i-m commented May 30, 2017

PlasmaPower commented May 30, 2017

davidak commented May 30, 2017

davidak commented May 30, 2017

PlasmaPower commented May 30, 2017 • edited Loading

PlasmaPower commented May 30, 2017 • edited Loading

PlasmaPower commented May 30, 2017

PlasmaPower commented May 30, 2017 • edited Loading

andrewda commented May 30, 2017

tarunbatra commented May 30, 2017

PlasmaPower commented May 30, 2017

PlasmaPower commented May 30, 2017

andrewda commented May 30, 2017

chaosbot commented May 30, 2017

hongaar commented May 30, 2017

andrewda commented May 30, 2017

PlasmaPower commented May 30, 2017

andrewda commented May 30, 2017

PlasmaPower commented May 30, 2017

rudehn commented May 30, 2017

andrewda commented May 30, 2017

rudehn commented May 30, 2017

PlasmaPower commented May 30, 2017

mdcfe commented May 30, 2017

PlasmaPower commented May 29, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading

PlasmaPower commented May 30, 2017 •

edited

Loading