[qTox-dev] qTox meeting #4

Sun May 22 11:20:51 UTC 2016

qTox meeting 2016-05-21, #qtox @ freenode

Maintainers present:
* TheSpiritXIII  (a.k.a. DaSpirit)
* tux3
* zetok

Maintainers absent without notification of absence:
* agilob (3rd absence without notification)
* antis81 (3rd absence without notification)
* sudden6 (1st absence without notification)

Short summary:

* users of qTox packages from tox.chat repo should get a notice in
changelog that will point them to packages in the OBS community repo
* most likely at some point those users will be transparently migrated
to OBS repos

Log from the meeting (log time UTC+1):
########################################################################

[18:54:22] <zetok> https://i.imgur.com/UMz944E.jpg :3
[18:55:00] <zetok> [21:08:32] <zetok> tux3: is the travis guard set?
[18:55:00] <zetok> [21:08:42] <tux3> zetok, hell no :)
[18:55:00] <zetok> [21:08:45] <zetok> why?
[18:55:00] <zetok> [21:09:23] <tux3> I want people to think, not be
limited by stupid software restrictions
[18:55:00] <zetok> [21:09:38] <zetok> that's not being limited
[18:55:00] <zetok> [21:09:52] <zetok> they're forced to think on how to
make it work at least on travis
[18:55:00] <zetok> [21:10:19] <zetok> people are lazy, and without some
stimuli they default to not thinking
[18:55:00] <zetok> [21:10:43] <tux3> Merging pull requests is not the
time to be lazy
[18:55:23] <zetok> tux3: and then you went & broke PRs
[18:55:32] <tux3> What
[18:55:42] <zetok> really, you didn't have to try that hard to prove
that I was right
[18:56:09] <tux3> You're the last person to have merged something, what
are you talking about?
[18:56:15] <zetok>
https://github.com/tux3/qTox/commit/8d225a8c97a36838092b325ca830c14546f3a0c6
[18:56:47] <tux3> What about it?
[18:57:00] <zetok> it broke PRs
[18:57:21] <tux3> I have a comment from a guy called "zetok" on this PR
saying that it worked ....
[18:57:22] <tux3> https://github.com/tux3/qTox/pull/3283
[18:57:24] <tux3> Just saying :)
[18:57:47] <zetok> tux3: yes, PR worked
[18:57:54] <zetok> tux3: your merge broke other PRs
[18:58:18] <tux3> What do you mean broke other PRs. The code compiled
and worked, that's what the Travis guard was about.
[18:58:28] <zetok> tux3: you've proved yourself that your workflow is a
bullshit and can't be trusted to do the job
[18:58:46] <zetok> travis also check commit messages
[18:58:49] <zetok> checks*
[18:59:12] <tux3> Will you please explain what you mean by "broke other PRs"
[19:02:19] <zetok> tux3:
https://travis-ci.org/tux3/qTox/jobs/129079369#L168 ← there you go :)
[19:03:49] <tux3> Okay, so the name of the merge commit made your script
fail, is that what you're getting at?
[19:04:38] <zetok> no
[19:05:40] <zetok> I'm pointing out that you yourself went and proved
that your workflow can't be trusted to do the job
[19:06:10] <zetok> normally, I'd say that you should reflect on that
[19:06:34] <zetok> but I think that it's better to get to the point in
this case
[19:06:47] <zetok> So, regarding your workflow: Change it.
[19:07:02] <tux3> You're not making any sense to me. It looks like
you're trying to prove a point
[19:07:17] <zetok> nah, you proved my point quite well :)
[19:07:36] <tux3> I'm not interested in proving points, I'm trying to
understand what the fuck you're getting at
[19:07:37] <tux3> You said "no" it's not the merge commit that's broken,
then what's broken?
[19:07:49] <zetok> since my point was that there will be more mistakes
made when one won't use the merge script :D
[19:08:28] <tux3> Okay so this is about the merge script now, not about
the Travis guard. The travis guard has nothing to do with any of that,
why'd you even bring it up if not for some misguided attempt at proving
a point?
[19:08:42] <zetok> err?
[19:09:08] <zetok> travis guard is a effective way of verifying that
no/less mistakes were made
[19:09:18] <tux3> Travis guard has nothing to do with this, if you look
it hapilly greenlighted my PR.
[19:09:26] <tux3> Travis can't catch merge commits before they are made...
[19:09:39] <zetok> tux3: yes, I've corrected that part, and now it would
fail your PR
[19:09:44] <zetok> s/pr/commit/
[19:09:57] <zetok> oh wait, no
[19:10:09] <zetok> https://github.com/tux3/qTox/pull/3312 still not merged
[19:10:48] <zetok> tux3: that's why it's good to push to some other
branch before --ff to master
[19:10:54] <tux3> That changes nothing, since merge commits are not done
through the web interface
[19:11:41] <tux3> Either way, I changed my merge scripts so it follows
your commit naming scheme, so this won't be a problem in the future.
[19:11:50] <zetok> err
[19:12:06] <zetok> that merge was made *after* you've changed your
scripts, AFAIK
[19:13:56] <tux3> It was at the same time I was updating my script, so I
must have used the wrong one at the time. It's definitely updated now.
[19:15:25] <tux3> Now the question I have is why does your script reject
people's PRs if it's another commit that is named incorrectly?
[19:15:41] <tux3> This makes no sense, you should only check the PR's
commits' names when building the PR
[19:16:12] <tux3> It looks like your script checks the whole history
everytime instead
[19:16:23] <zetok> nope
[19:16:32] <zetok> it checks only commits in the supplied range
[19:16:42] <tux3> The supplied range should be only the PR.
[19:16:44] <zetok> now, it is travis that supplies the range..
[19:17:09] <zetok> (and travis can't really supply range only from the PR)
[19:17:21] <tux3> Git should be able to tell you at which point the
branch of the PR diverged, so the script could check it
[19:18:46] <zetok> so you'd want to have another script for that?
[19:19:09] <zetok> shouldn't master simply not break?
[19:19:12] <tux3> Couldn't the script detect when it's on a pull request
branch automatically?
[19:20:07] <zetok> no
[19:20:46] <zetok> note that the scrip can easily be used standalone by
people who would want to verify commits' compliance to the format
[19:24:58] <zetok> tux3: why would you prefer a workaround to the
problem, rather than addressing the problem itself?
[19:25:12] <zetok> >tux3> Now the question I have is why does your
script reject people's PRs if it's another commit that is named incorrectly?
[19:25:19] <zetok> that's not really the best question one could ask
[19:25:27] <tux3> Travis guard would not have stopped me from pushing
that commit, would it?
[19:25:28] <zetok> the better one would be: why it shouldn't do that?
[19:25:47] <zetok> tux3: with the current workflow, no
[19:26:22] * cisc has quit (Quit: Konversation terminated!)
[19:26:44] <linuxmodder> zetok,  fyi for any fedora24 beta users have
them add https://unitedrpms.github.io/unitedrpms.repo for  ffmpeg  bits
[19:29:05] * cisc (~cisc at unaffiliated/cisc) has joined
[19:32:47] * abbat has quit (Ping timeout: 246 seconds)
[19:34:14] * Dragofix has quit (Quit: Don't cry because IRC session is
over, smile because it happened!)
[19:35:19] <nurupo> ><tux3> [...] I'm trying to understand what the fuck
you're getting at <-- you just described every conversation ever i had
with zetok
[19:35:43] <zetok> aw
[19:40:26] * Dragofix (Dragofix at a91-155-20-164.elisa-laajakaista.fi) has
joined
[19:42:05] * abbat (~Anton at 109.188.127.28) has joined
[19:43:21] <zetok> nurupo: btw, are there any plans for pointing qTox
users to the right repo on tox.chat?
[19:43:57] <zetok> (and generally not lying to users, but by now I don't
really expect tox.chat to be that good)
[19:44:05] <zetok> !
[19:44:46] <nurupo> zetok: i would imagine tux3 would open a PR once you
fully move to OBS for linux builds?
[19:45:01] <nurupo> didn't you talk about this with oranges all the time?
[19:45:22] <zetok> did I?
[19:45:46] <tux3> zetok, we could edit the readme to point to OBS, but
for the current tox.chat users there's no plan right now
[19:46:17] <zetok> tux3: done :)
[19:46:28] <tux3> zetok, Okay, good
[19:46:47] <zetok> nurupo: right, from what I remember I've said to
oranges that at some point OBS will likely be used instead
[19:46:48] <tux3> Since I can watch for OBS build script changes, it's a
good solution
[19:47:18] <zetok> tux3: what would be good, is to stop pointing users
to the wrong repo
[19:47:28] <nurupo> zetok: pretty sure i saw oranges talking about the
move to OBS with someone in here on several different occasions
[19:47:36] <tux3> zetok, we can do that. For users already on tox.chat
though it's not clear what to do
[19:48:05] <zetok> nurupo: yes, but most(?) of other talks weren't with
me :)
[19:48:19] <nurupo> oh, okay then
[19:48:43] * abbat has quit (Quit: Leaving)
[19:48:49] <zetok> tux3: add a script that will give users notification
to move to new repo
[19:48:58] <zetok> that is, for the tox.chat packages
[19:49:01] <tux3> Like a popup?
[19:49:02] * abbat (~Anton at 109.188.127.28) has joined
[19:49:15] <zetok> um, I'm not quite sure if that qualifies as a pop-up
[19:49:29] <zetok> perhaps pop-up would be better than my idea
[19:50:00] <zetok> I thought about using that thing debian has for
notifying users about ~breaking changes during updates
[19:50:06] <tux3> We can do a changelog with packages that would be
shown by the package managers, but I don't know if people read that
[19:50:27] <tux3> Especially on the GUI update tool, I'm not sure it
shows changelog warnings
[19:50:35] <tux3> (does it?)
[19:50:39] <zetok> they will, be it sooner or later, if you do changelog
for every update :D
[19:50:46] <zetok> dunno about the GUI
[19:50:50] <nurupo> Apper does show changelog
[19:50:51] * zetok never really used it
[19:50:58] <tux3> Okay, that's an idea then
[19:51:07] <nurupo> don't expect anyone to notice it though
[19:51:20] <tux3> Yeah, I'm not sure people will read it
[19:52:05] <zetok> well, as always, the alternative is to include a
post-setup script that will add new repo, key, remove old qtox and
install new qtox from the new repo
[19:52:42] <tux3> Maybe. I'm not fond of it, but I don't see a much
better solution.
[19:53:18] <zetok> perhaps phases should be done
[19:53:23] <zetok> first with only changelog
[19:53:45] <zetok> some time after that, also adding the scrip for the
people who can't be bothered to read "changelog"
[19:54:36] <zetok> (and warn in changelog that new versions will install
new repo)
[19:55:08] <zetok> (that'll teach 'em to read the changelog ;D )
[19:55:13] <ovalseven8> #nsis
[19:55:29] <nurupo> Changelog: [upstream] replace qTox with a potato
(pop-up)
[19:55:30] <nurupo> heh
[20:00:05] <zetok> tux3: do you have time to do the changelog thing now?
[20:00:12] <zetok> ~now
[20:01:52] <zetok> tux3: btw, what about dubslow?
[20:03:30] <tux3> I'll try to do the changelog thing
[20:03:42] <tux3> Dubslow, last I checked, was still kill. Any news?
[20:03:56] <zetok> no, but I haven't tried to contact him
[20:04:25] <zetok> tux3: I take that you too haven't tried to contact him?
[20:05:48] <zetok> tux3: and if so, could you try to do so, and ask what
he thinks about ~stuff and whether he could consider helping with qTox
again?
[20:07:22] <tux3> I think he's gone, if he wants to come back he'll come
back.
[20:07:28] <tux3> But you can write him if you want
[20:07:37] <zetok> :/
[20:07:44] <zetok> so, here's the thing
[20:07:51] <tux3> I mean I can't force him to work on qTox against his
will..
[20:08:01] <zetok> yes, but that's not about it
[20:08:23] <zetok> it's about whether he can considering helping again
[20:08:24] <nurupo> ><tux3> Dubslow, last I checked, was still kill. Any
news? <-- i have seen him active on reddit
[20:08:41] <zetok> tux3: and me asking him about that probably wouldn't
be good
[20:08:53] <nurupo> so he is not kill, maybe just doesn't want to work
on qtox anymore
[20:08:53] <zetok> s/considering/consider/
[20:09:21] <zetok> tux3: instead, you asking most likely would be better
[20:11:45] <tux3> The last I know is that Dubslow quit because he got a
job and was totally burned out on C++ stuff, I can understand that and I
don't want to pressure him to work for us if he doesn't feel like it
[20:11:51] <nurupo> it's kind of scary that someone who has been missing
for over a year still has push access into master
[20:12:20] <zetok> tux3: yes, but the point is not to pressure
[20:12:46] <tux3> nurupo, yes maybe I should remove him. I still trust
him, but it makes sense
[20:14:09] <zetok> hmm
[20:14:25] * zetok wonders how to word that without the pressure
[20:15:27] <tux3> Feel free to ask him what's up if you want, but I'm
sure he hasn't forgotten about us, if he wants to do something else
that's his choice.
[20:16:04] <zetok> there are reasons why me asking most likely wouldn't
be good
[20:16:35] <tux3> Then me asking on your behalf wouldn't be honest,
because it's definitely your idea.
[20:18:45] <zetok> sorry, but for me: honesty < qTox
[20:19:41] <zetok> "Been a while, would you want to help with qTox
again?" ← that should do :|
[20:20:23] <zetok> no pressure™
[20:23:12] <tux3> "No". I'll be doing the changelog update.
[20:23:21] <zetok> tux3: btw, have you subscribed to the ML
[20:23:23] <zetok> ?
[20:23:45] <tux3> The qTox ML was created?
[20:23:48] <zetok> yep
[20:23:53] <tux3> Neat, link pls
[20:24:08] <zetok> https://lists.tox.chat/listinfo/qtox-dev
[20:24:52] <zetok> you would have known if you read the mail about the
previous meeting :(
[20:25:26] <tux3> Hm, I might have missed it. I sent a subscription request
[20:26:07] <linuxmodder> zetok,  is it  solely  devs or  anyone  wishing
to help
[20:26:16] <zetok> linuxmodder: anyone :)
[20:27:10] <linuxmodder> not that I need another noisy list  but  as I
am trying to do more with wtox  why not then :)
[20:27:34] <linuxmodder> already  have  76 filters for MLs :)
[20:30:29] <zetok> nurupo: so.. instructions on the tox.chat?
[20:32:00] <zetok> tux3:
https://lists.tox.chat/pipermail/general/2016-May/000074.html to read
later :)
[20:32:17] <nurupo> zetok: ><nurupo> zetok: i would imagine tux3 would
open a PR once you fully move to OBS for linux builds?
[20:32:37] <tux3> zetok, I definitely didn't receive the mail about that
meeting
[20:32:48] <zetok> nurupo: imagination is a powerful thing
[20:32:58] <zetok> nurupo: yet alone, it doesn't make things happen
[20:33:35] <zetok> tux3: are you subscribed to `General` ML?
[20:34:10] <tux3> zetok, I imagine not, so I'm about to
[20:36:49] <zetok> nurupo: meanwhile, tox.chat is still misinforming
users, and I heard that you're maintaining the stuff now
[20:40:05] <linuxmodder> OBS is  Osuse build system  yes
[20:40:34] <zetok> open build system, instance of which openSUSE is
running :)
[20:40:45] <linuxmodder> ah
[20:41:24] <nurupo> zetok: you are free to open PR if you have time as i
don't have it
[20:42:30] <zetok> nurupo: ah, it was proven to me time ago, over and
over again that I'd only waste my time if I've tried to improve things
governed by Tox committee
[20:42:49] <zetok> so thanks, but no thanks, I prefer to not waste my time
[20:43:42] <nurupo> so, you care enough to complain but not enough to fix?
[20:44:14] <zetok> I cared enough to fix, but that has been taken care
of, now I'm just pointing out that you're misinforming users
[20:46:25] <nurupo> zetok: what exactly is misinforming users, the
download of qTox?
[20:47:28] <zetok> among stuff
[20:47:45] <tux3> I pushed the changelog entry that tells to go to
https://software.opensuse.org/download.html?project=home%3Aantonbatenev%3Atox&package=qtox
[20:47:59] <tux3> tox.chat updates every day no? So we'll know if it
worked tomorrow
[20:48:33] <zetok> nurupo: you might want to ponder on how I've
complained about qTox issues not having labels until tux3 added me to
collaborators, in which position I couldn't complain about that any
longer, since I was the one responsible for fixing it :)
[20:48:39] <zetok> tux3: thanks :)
-- 

Kind regards,
Zetok Zalbavar
----
My Tox ID:
29AE62F95C56063D833024B1CB5C2140DC4AEB94A80FF4596CACC460D7BAA062E0A92C3424A0

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.tox.chat/pipermail/qtox-dev/attachments/20160522/81f89859/attachment.sig>