[03:01:20] <Krinkle>	 Wow! A google for "fallback createTreeWalker" brings up a 6 month old pastie.org dump as #1 result
[03:01:24] <Krinkle>	 ... one I wrote last year
[03:01:38] <Krinkle>	 https://www.google.com/search?q=fallback+createTreeWalker
[03:02:21] <Krinkle>	 must've indexed irc logs somehow, and even got to #1 hit
[03:02:22] <Krinkle>	 sweet
[09:28:56] <gerrit-wm>	 New patchset: Hashar; "run tests for some whitelisted users" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/39310
[09:29:17] <gerrit-wm>	 New review: Hashar; "PS5: escape dots in aude email" [integration/zuul-config] (master); V: 2 C: 2;  - https://gerrit.wikimedia.org/r/39310
[09:29:17] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/39310
[09:41:01] <gerrit-wm>	 New patchset: Hashar; "update workflow with Verified+2 enhancement" [integration/doc] (master) - https://gerrit.wikimedia.org/r/39085
[09:41:50] <gerrit-wm>	 New review: Hashar; "Removing CR-2 since the Verified+2 has been deployed" [integration/doc] (master); V: 0 C: 0;  - https://gerrit.wikimedia.org/r/39085
[11:14:43] <ori-l>	 hashar: what's with jenkins running PEP8 against puppet manifests?
[11:15:55] <hashar>	 ori-l: bug ? :-)
[11:15:59] <hashar>	 should probably remove that lint job
[11:16:10] <ori-l>	 would you like me to file one?
[11:16:41] <ori-l>	 a bug, i mean
[11:17:00] <ori-l>	 also, result of experiment: puppet manifests do not conform to the python style guide :)
[12:11:24] <hashar>	 ori-l: sorry, yeah please fill in a bug :-)
[12:11:33] <hashar>	 ori-l: will not take care of it before january first
[12:11:38] <hashar>	 first / though
[12:12:22] <Danwe>	 Krinkle: Do you know whether ResourceLoaderModule::getMessages works how it should? I return a bunch of messages as array there but the messages won't be found
[12:15:03] <gerrit-wm>	 New patchset: Ori.livneh; "Update comment" [analytics/glass] (master) - https://gerrit.wikimedia.org/r/39375
[12:15:20] <gerrit-wm>	 Change merged: Ori.livneh; [analytics/glass] (master) - https://gerrit.wikimedia.org/r/39375
[12:21:35] <Danwe>	 Krinkle: nevermind, it seems to work now, seemed like some heavy caching issue.
[15:32:13] <hashar>	 ^demon: hey :)  I have added a bunch of submodules in mediawiki/extensions , would you be able to review it please ? https://gerrit.wikimedia.org/r/#/c/39374/
[15:33:19] <^demon>	 merged.
[15:34:30] <hashar>	 \O/
[15:34:58] <hashar>	 ^demon: I also wrote a script to compare the list of extensions in Gerrit versus the currently registered module : https://gerrit.wikimedia.org/r/#/c/39373/ :-D
[15:35:03] <hashar>	 not that urgent though
[15:35:14] <^demon>	 Well, the script update-extensions does that.
[15:35:55] <^demon>	 https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions.git;a=blob;f=update-extensions.sh;h=93a00ed30fdc153c9f0375260bfc5b6a1f1f2a5e;hb=HEAD
[15:38:11] <hashar>	 ohh
[15:38:15] <hashar>	 a simpler version :-D
[15:38:47] <^demon>	 If you want to add features like maybe a --dry-run to just print the results, we can add to the existing extension.
[15:38:52] <^demon>	 s/extension/script/
[15:38:57] <^demon>	 Anyway, let's abandon 39373.
[15:39:16] <hashar>	 will do
[15:45:22] <gerrit-wm>	 New patchset: Hashar; "basic inline documentation" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/38520
[15:46:08] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/38520
[16:06:47] <gerrit-wm>	 New review: Hashar; "Thanks for the change." [integration/jenkins-job-builder-config] (master); V: 0 C: -1;  - https://gerrit.wikimedia.org/r/38555
[17:47:05] <^demon|lunch>	 anomie: Sooo, I was trying to build luasandbox...but I think I might be missing some library. LUAI_MAXCALLS is undefined. Was there something other than libluajit you installed?
[17:48:37] <anomie>	 ^demon|lunch- On my system, I had to install the development files for regular Lua in addition to LuaJIT.
[17:51:53] <^demon|lunch>	 I'm doing this on an ubuntu vm. Looks like I needed liblua5-1-dev
[17:51:59] <^demon|lunch>	 Seems to have worked :)
[17:55:57] <hashar>	 anomie: hi :) I forgot to ping you this afternoon
[17:56:17] <anomie>	 hashar- Oh, did you deploy those config changes?
[17:56:33] <hashar>	 anomie: I will deploy your multi datacenter changes tomorrow morning (europe), aka starting 8am utc
[17:56:47] <hashar>	 anomie: I did deploy the $realm -> $wmfRealm one
[17:56:51] <hashar>	 a few hours ago
[17:56:58] <hashar>	 been testing the rest in beta
[17:57:03] <hashar>	 will test again tonight probably
[17:57:36] <hashar>	 anomie: also I have split the MWRealm class in its own change : https://gerrit.wikimedia.org/r/#/c/39191/ but you reviewed it already
[17:57:51] <hashar>	 anomie: had to do that to write the unit tests and I thought it would be easier to amend / review this way
[17:58:06] <hashar>	 anomie: maybe you can get someone else to triple check it and then that can be merged tonight.
[17:58:12] <hashar>	 anomie: that change is harmless (for now)
[17:58:57] <anomie>	 hashar- I saw that. 32167 should somehow be rebased on top of that and 39060, except I can't figure out how to do that.
[17:59:51] <hashar>	 anomie: trick with git rebase :-D
[17:59:53] <hashar>	 not a big deal
[18:00:50] <hashar>	 I have marked the first change CR-2 to prevent submission till https://gerrit.wikimedia.org/r/#/c/39057/ (MWRealm) is merged
[18:03:05] <kaldari>	 It seems the job queue is currently not running automatically for mediawiki.org. Is there any way we can get that running?
[18:03:15] <kaldari>	 AaronSchulz: ^
[18:03:57] <AaronSchulz>	 there is nothing special about mw.org
[18:06:52] <kaldari>	 guess I just don't understand how the job queue works then
[18:07:57] <kaldari>	 AaronSchulz: What's the average delay on job queue items being processed?
[18:08:42] <AaronSchulz>	 hard to say
[18:08:51] <kaldari>	 i.e. a second, an hour, a day?
[18:09:42] <AaronSchulz>	 that could also depend how many things are in the queue
[18:10:15] <AaronSchulz>	 if there is nothing in a queue and something is added, there could be a 5min delay before the queue is listed as available
[18:10:52] <AaronSchulz>	 wikis with available queues are selected randomly and jobs are run from them
[18:11:08] <AaronSchulz>	 there is a jobs-loop.sh script in puppet for this
[18:11:41] <kaldari>	 OK, maybe I just didn't wait long enough, I
[18:11:42] <kaldari>	 '
[18:11:47] <kaldari>	 I'll try it again
[18:11:50] <kaldari>	 thanks for the info1
[18:12:48] <AaronSchulz>	 anyway, a day would be pretty long for mw.org
[18:12:57] <AaronSchulz>	 it would be nice to have specific averages I guess
[18:15:11] <MatmaRex>	 hey, maybe i can request someone to look at https://bugzilla.wikimedia.org/show_bug.cgi?id=40124 here? (this is the arbitrary prefs via API thing)
[18:15:28] <MatmaRex>	 Krinkle said he's gonna reply, but then didn't :(
[18:15:55] <MatmaRex>	 and now there's christmas and all that and i'll be lucky if i get any input sometime in february at best :/
[18:17:16] <Reedy>	 http://ganglia.wikimedia.org/latest/graph.php?c=Miscellaneous%20pmtpa&h=spence.wikimedia.org&v=2617299&m=enwiki_JobQueue_length&r=hour&z=small&jr=&js=&st=1355941020
[18:33:36] <Krinkle>	 MatmaRex: I think I've given my opinion on the arbitrary prefs thing already on various ocasions.
[18:34:25] <MatmaRex>	 Krinkle: you keep claiming there are security vulnerabilities in this, and i have yet to hear any details.
[18:34:42] <Krinkle>	 that's just a small part of the issue
[18:35:01] <MatmaRex>	 please do tell what the security vulnerabilities are in the implementation i'm proposing
[18:35:32] <marktraceur>	 Hrm, where's a csteipp when you need one.
[18:35:36] <Krinkle>	 the security issues would be in the usage inside the gadget, not in the core implementation. as long as it uses the API to export variables, the transport will be fine.
[18:35:53] <Krinkle>	 there is no security factor in this, that's fine.
[18:36:23] <MatmaRex>	 your primary arguments against it was security :/ see Tyler's comment: https://bugzilla.wikimedia.org/show_bug.cgi?id=40124#c30
[18:36:35] <MatmaRex>	 your FUD tactic is working out, nobody wants t otouch this iwth a stick now
[18:36:54] <MatmaRex>	 because of imaginary security issues
[18:37:19] <Krinkle>	 My primary argument is not security.
[18:37:34] <MatmaRex>	 i'd appreciate if you noted this on the bug.
[18:37:48] <Krinkle>	 Someone started talking about it on the bug and I clarified some things.
[18:38:25] <Krinkle>	 There are however concerns that imho should eventually wontfix this bug.
[18:39:04] <MatmaRex>	 there is no way for plain old user script to save non-public non-volatile preferences now.
[18:39:11] <MatmaRex>	 this is bad. i'm trying to solve the issue.
[18:39:25] <Krinkle>	 There is also no way to create unicorns and deliver them to your doorstrep with one click.
[18:39:31] <MatmaRex>	 ...
[18:39:42] <Krinkle>	 I know the use case and as I said before, the use case is valid.
[18:39:42] <MatmaRex>	 i'm also increasinly under impression that you're not reading my replies :/
[18:39:44] * AaronSchulz  wants a pony unicorn

[18:39:46] <MatmaRex>	 D:
[18:39:48] <Krinkle>	 But this is not ht way to implement it.
[18:39:59] <MatmaRex>	 Gadgets 2.0 is for gadgets.
[18:40:12] <MatmaRex>	 there are not gadgets in core MW.
[18:40:14] <MatmaRex>	 no*
[18:40:15] <Krinkle>	 MatmaRex: The fundamental issue is that user scripts (not gadgets) do not exist.
[18:40:29] <Krinkle>	 They're like illegal immigrants.
[18:40:39] <AaronSchulz>	 bz quip?
[18:40:43] <MatmaRex>	 what.
[18:40:54] <Krinkle>	 There is no registry, they are loaded with <script>, (no delivery system, arbitrary execution in global scope)
[18:40:56] <MatmaRex>	 i think that metaphor flew over my head.
[18:41:32] <MatmaRex>	 to keep with your metaphors, is MW core going to be a totalitarian government with a registry for everything?
[18:41:34] <marktraceur>	 But they're mighty helpful and do great stuff. (let's keep 'er positive, folks)
[18:41:49] <Krinkle>	 MatmaRex: No, not in this context any way.
[18:42:27] <marktraceur>	 Krinkle: It sounds more like you want to build a way for user scripts to be more integrated, to be registered and delivered and so on like other script resources.
[18:42:46] <Krinkle>	 MatmaRex: My main concerns are conflicts, user interface and maintainability.
[18:42:48] <MatmaRex>	 that's called Gadgets' it's also missing essential features.
[18:42:51] <MatmaRex>	 eh
[18:43:01] <MatmaRex>	 i've already linked a script that solves the first two issues
[18:43:12] <MatmaRex>	 and i don't see the third
[18:43:14] <Krinkle>	 No, it doesn't "solve" anything.
[18:43:20] <kaldari>	 Reedy: The current job queue on mediawikiwiki seems to have 122 items dating back to November. Why would it have items that old in it?
[18:43:23] <Krinkle>	 It is a perfect example of the problem.
[18:43:34] <MatmaRex>	 should i integrate it into core?
[18:44:12] <Krinkle>	 No, but lets imagine for now that we did so that we can make a point.
[18:44:33] <Krinkle>	 How would one trigger it and register the preferences from a user script with that?
[18:44:58] <Krinkle>	 the script is loaded at a practically random point in time with importScript() (or mw.loader.load(url)).
[18:45:19] <MatmaRex>	 it's all there in the docs i linked as well.... mw.loader.using('the gadget', ...)
[18:45:41] <MatmaRex>	 and then gConfig.register('lipsum', 'Lorem ipsum gadget', [settings go here])
[18:46:11] <Krinkle>	 Right, so you implement a module that acts as a registry
[18:46:13] <MatmaRex>	 if you don't have RL to handle the loading for you, you can just register the setting on document load.
[18:46:24] <MatmaRex>	 the settings*
[18:46:34] <Krinkle>	 That's exactly what Salvador has already implemented in his branch, in a very integrated, solid, translated way.
[18:46:38] <kaldari>	 AaronSchulz: http://ganglia.wikimedia.org/latest/graph.php?c=Miscellaneous%20pmtpa&h=spence.wikimedia.org&v=2617299&m=mediawikiwiki_JobQueue_length&r=hour&z=small&jr=&js=&st=1355941020 but the delay is 40 minutes and counting
[18:46:55] <MatmaRex>	 Krinkle: when am i going to see this beatiful solution on WMF wikis? 2015?
[18:47:22] <Krinkle>	 MatmaRex: Oh so you're playing that card, I see.
[18:47:25] <MatmaRex>	 i've been hearing that gadgets 2.0 is in the works for a few years now
[18:47:33] <Krinkle>	 11 months, not more.
[18:47:42] <Krinkle>	 Before that it term wasn't used yet.
[18:47:43] <MatmaRex>	 well, i have an issue now. i'd like it solved reasonably soon. :/
[18:47:48] <Krinkle>	 (I know, since I started it with Roan)
[18:48:00] <MatmaRex>	 "gadgets next version", then, whatever
[18:48:00] <Krinkle>	 MatmaRex: This is a pretty elaborate framework we're talking about.
[18:48:13] <MatmaRex>	 precisely. that's why i want something simple.
[18:48:19] <Krinkle>	 You can't have it right now.
[18:48:27] <MatmaRex>	 so that it can be reviewed and deployed within a reasonable timeframe.
[18:49:02] <MatmaRex>	 even more, it used to work, and it was implemented intentionally
[18:49:13] <MatmaRex>	 (per the last comment on the bug)
[18:49:31] <MatmaRex>	 and now it got broken with a "security fix" that caused about 4 different regression related to that API
[18:49:38] <Krinkle>	 No, that's not true. It was never intended to allow arbitrary and unlimited storage in user prefs.
[18:49:40] <MatmaRex>	 (including this one)
[18:49:54] <MatmaRex>	 https://bugzilla.wikimedia.org/show_bug.cgi?id=40124#c32
[18:49:59] <MatmaRex>	 "When I was implementing the change that introduces options module for API my
[18:50:00] <MatmaRex>	 original plan was to use it for storing gadget preferences. It can be seen in
[18:50:00] <MatmaRex>	 unit tests I wrote.
[18:50:00] <MatmaRex>	 "
[18:50:16] <MatmaRex>	 (Beau is the author of this part of API)
[18:50:28] <MatmaRex>	 so yes, it quite was
[18:50:43] <Krinkle>	 Well, that use case was never considered with the team afaik. Not discussed and not communicated.
[18:50:46] <kaldari>	 Reedy, AaronSchulz: We're supposed to be doing the launch of Echo on mediawiki.org today, but the job queue is not behaving as expected. Who would be best to discuss this with?
[18:50:53] <MatmaRex>	 my proposed solution doesn't even allow entirely arbitrary prefs
[18:50:55] <Krinkle>	 I'd say he sneaked it in through the back door.
[18:51:12] <Krinkle>	 MatmaRex: I know
[18:51:58] <MatmaRex>	 well, it was merged, i'd assume "the team" knows what's happening to their code
[18:52:32] <Krinkle>	 MatmaRex: So at this point the dilemma is that we either go and review an "ideal" solution (Salvador's Gadgets' fork), that has integration with mw.user, mw.msg, Special:Preferences etc. (the preferences can be read without loading the gadget).
[18:53:04] <Krinkle>	 or we have people we don't have work on bringing in and reviewing another solution in mediawiki core that will be used by user scripts.
[18:53:20] <MatmaRex>	 we could also do both.
[18:53:23] <Krinkle>	 I don't make these kind of calls,, but I know I'm not available for it.
[18:53:36] <Krinkle>	 (because of other projects I have to work on)
[18:53:45] <MatmaRex>	 and later migrate to the "better" solution on wikis with gadgets enabled
[18:54:26] <Krinkle>	 Sure, that sounds great. But realistically speaking I don't think Platform prioritises this, that's too bad.
[18:54:41] <Krinkle>	 I only have 2 arms and 2 legs.
[18:54:46] <MatmaRex>	 well, you have very strong opinions on the matter and -2'd the change on gerrit. so it looks like this is a project you make these kind of calls.
[18:54:49] <MatmaRex>	 :/
[18:54:59] <MatmaRex>	 so it looks like you make these kind of calls*
[18:55:08] <AaronSchulz>	 2012-12-19 18:54:25 srv195 tawikisource: Memcached error for key "tawikisource:shorturls:title:திருவிவிலியம்/இணைத்_திருமுறை_நூல்கள்/பாரூக்கு_(எரேமியாவின்_மடல்)/அதிகாரங்கள்_5_முதல்_6_வரை" on server ":": A BAD KEY WAS PROVIDED/CHARACTERS OUT OF RANGE
[18:55:11] <AaronSchulz>	 lol
[18:55:22] <AaronSchulz>	 sha1 that stuff ;)
[18:55:40] <Krinkle>	 MatmaRex: There's a lot of changes I'm cc-ed on, some I look into in my 20% time, some I don't. I can give my opinion, sure.
[18:56:12] <Krinkle>	 MatmaRex: Also, I think as-is it wouldn't be merged. But if say gConfig would be generalised for user scrips, It stands a good chance (given someone can review it).
[18:56:19] <Krinkle>	 I wouldn't -2 it.
[18:56:32] <MatmaRex>	 well, it works perfectly well for user scripts
[18:57:20] <MatmaRex>	 also, -2 is IMO more than opinion, it is a very strong statement, especially made by one of the WMF people
[18:57:52] <Krinkle>	 It is code review, not intention review.
[18:58:09] <MatmaRex>	 your -2 was for the first draft. i rewrote most of the code since.
[18:58:10] <Krinkle>	 regarding the implementation as of that version of that patch.
[18:58:16] <MatmaRex>	 please re-review or drop the -2.
[18:58:32] <MatmaRex>	 that's one of the things with -2, that it sticks...
[18:59:11] <MatmaRex>	 and therefore applies for all versions, not only "that version"
[18:59:19] <Krinkle>	 That was the idea, since in most cases different patch sets only make small ammendments so that it isn't implied that by default it is 0, but the previous concerns still apply, unless re-reviewed
[18:59:33] <MatmaRex>	 your -2 was on PS3, current patchset is PS8
[18:59:44] <Krinkle>	 I know, I'm looking at it now.
[18:59:46] <MatmaRex>	 am i supposed to make a new changeset? that would be immensely stupid
[18:59:52] <Krinkle>	 no
[19:00:34] <MatmaRex>	 you didn't even leave any precise comment, only "Do not merge pending discussion on bug 40124 regarding whether this feature is wanted in mediawiki core." and a suggestion not to use boolean parameters
[19:00:39] <MatmaRex>	 the discussion has since stalled
[19:00:46] <MatmaRex>	 (waiting for you to reply, mostly)
[19:02:51] <MatmaRex>	 basically, dropping a -2 bomb should IMO be considered a promise of making a timely followup, should anything change :/
[19:04:37] <Krinkle>	 and I intend to do so, it is on my dashboard
[19:04:49] <Krinkle>	 but there is a lot on there
[19:04:56] <Krinkle>	 (and more things that aren't on there)
[19:05:18] <MatmaRex>	 i understand this, but please understand me as well :/
[19:06:04] <Krinkle>	 I do, but there's what I can do, and what I have time for, and things I don't know anything about and should stay away from.
[19:06:42] <Krinkle>	 Parent5446 also raised some concerns.
[19:06:53] <MatmaRex>	 they were addressed.
[19:13:33] <Krinkle>	 MatmaRex: and you have some code issues you replied yourself, those aren't addressed, right?
[19:13:35] <Krinkle>	 https://gerrit.wikimedia.org/r/#/c/37503/8/includes/User.php
[19:13:40] <Krinkle>	 "(It should be $this instead of $user here.) "
[19:14:18] <MatmaRex>	 Krinkle: yes, but i don't want to submit another patchset, because i will have to copy all other inline comments to it (again)
[19:14:43] <MatmaRex>	 i'm not a PHP coder, and i don't know the codebase too well, so i have questions
[19:14:50] <Krinkle>	 please don't. nobody does that. Gerrit is annoying, but I'm not sure duckpunching makes that better.
[19:14:51] <MatmaRex>	 e.g. on https://gerrit.wikimedia.org/r/#/c/37503/8/includes/api/ApiOptions.php
[19:15:11] <Krinkle>	 copying is an interesting approach though
[19:15:21] <MatmaRex>	 well, what else can i do
[19:15:28] <MatmaRex>	 nobody is going to look at older patchsets
[19:15:57] <MatmaRex>	 i implemented it and it kinda works
[19:16:01] <Krinkle>	 well, you get e-mail notifications for each comment. Those addressed/answered, I delete from my inbox.
[19:16:08] <MatmaRex>	 (it'd help if jenkins ran the full tests against the change, though)
[19:16:35] <MatmaRex>	 can you or somebody force it to just once? i saw that a whitelist was just implemented
[21:49:06] <AaronSchulz>	 kaldari: we really need a "request new captcha" ajax thing
[21:49:32] <kaldari>	 AaronSchulz: yes
[21:50:17] <kaldari>	 which extension is currently serving the captcha on the cluster?
[21:50:33] <AaronSchulz>	 ConfirmEdit...the FancyCaptcha.class stuff
[21:50:48] * AaronSchulz  is playing with different fonts locally

[21:51:13] <AaronSchulz>	 FreeMonoBoldOblique.ttf seems decent
[21:51:31] <AaronSchulz>	 I may nuke all the ones in swift and make a new set for test2wiki
[21:57:59] <Reedy>	 AaronSchulz: use windings or something
[21:58:34] <AaronSchulz>	 Reedy: hmm?
[21:58:49] <Reedy>	 for the captchas
[21:58:56] <AaronSchulz>	 windings?
[21:59:48] <MatmaRex>	 wingdings* ~
[22:04:10] <Reedy>	 That too
[22:08:15] <marktraceur>	 ^demon: Mind if I bend your ear about a repository request? :)
[22:08:16] <AaronSchulz>	 someone is back
[22:09:52] <^demon>	 marktraceur: Ping.
[22:10:03] <^demon>	 marktraceur: Or pong, I suppose.
[22:10:30] <marktraceur>	 ^demon: http://www.mediawiki.org/wiki/Git/New_repositories/Requests last one is Git2Pages, I was hoping to have it set up soonish
[22:12:38] <^demon>	 This is a mw extension?
[22:13:25] <^demon>	 Also, who is terrrydactyl? I have no such ldap user
[22:13:38] <^demon>	 Oh there he is.
[22:13:45] <marktraceur>	 ^demon: Yes, a new extension, and she just created a labsconsole account
[22:15:00] <^demon>	 Repo done, https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/extensions/Git2Pages
[22:15:05] <marktraceur>	 Sweet.
[22:15:07] <^demon>	 Just needs a .gitreview and .gitignore.
[22:15:28] <marktraceur>	 ^demon: Is that your next task or should I attempt to fashion them?
[22:15:53] <^demon>	 Reedy has a script that does it automagically. Maybe we could ping him ;-)
[22:16:04] <^demon>	 I used to have a script but I got a new laptop.
[22:16:49] <marktraceur>	 Reedy: We could use some brilliant magic script! :)
[22:17:29] <marktraceur>	 I'd imagine it's not terribly difficult to mimick, though...
[22:21:49] <marktraceur>	 Evidently I can't do that.
[22:23:32] <marktraceur>	 ^demon: Is there a way I can make the .gitignore and .gitreview? It seems the normal methods of submitting changes won't work.
[22:23:47] <^demon>	 `git push origin HEAD:refs/for/master`
[22:26:26] <marktraceur>	 I think I fixed it....yup
[22:32:46] <Reedy>	 heh