fryguybob has quit [Read error: Connection reset by peer]
fryguybob has joined #pypy
msjyoo has quit [Ping timeout: 264 seconds]
phlebas_ has joined #pypy
string_ has joined #pypy
cadr__ has joined #pypy
wallet42_ has joined #pypy
njs has quit [Ping timeout: 246 seconds]
Graypup_ has quit [Ping timeout: 246 seconds]
EWDurbin_ has joined #pypy
kandinski has quit [Ping timeout: 244 seconds]
graingert_ has joined #pypy
_habnabit has quit [Ping timeout: 246 seconds]
JStoker has quit [Ping timeout: 276 seconds]
kandinski has joined #pypy
dpn` has quit [Ping timeout: 276 seconds]
ulope_ has joined #pypy
kbtr has quit [Remote host closed the connection]
Graypup_ has joined #pypy
simpson has quit [Ping timeout: 252 seconds]
phlebas has quit [Ping timeout: 252 seconds]
petronny has quit [Ping timeout: 252 seconds]
string has quit [Ping timeout: 252 seconds]
mvantellingen has quit [Ping timeout: 252 seconds]
EWDurbin has quit [Ping timeout: 252 seconds]
cadr_ has quit [Ping timeout: 252 seconds]
graingert has quit [Ping timeout: 252 seconds]
wallet42 has quit [Ping timeout: 252 seconds]
ulope has quit [Ping timeout: 252 seconds]
trfl has quit [Read error: Connection reset by peer]
blueyed has quit [Ping timeout: 252 seconds]
igitoor has quit [Ping timeout: 252 seconds]
dpn` has joined #pypy
mvantellingen has joined #pypy
phlebas_ is now known as phlebas
wallet42_ is now known as wallet42
ulope_ is now known as ulope
kbtr has joined #pypy
string_ is now known as string
EWDurbin_ is now known as EWDurbin
njs has joined #pypy
graingert_ is now known as graingert
igitoor_ has joined #pypy
JStoker has joined #pypy
trfl has joined #pypy
petronny has joined #pypy
simpson has joined #pypy
_habnabit has joined #pypy
msjyoo has joined #pypy
igitoor_ has quit [Changing host]
igitoor_ has joined #pypy
igitoor_ is now known as igitoor
<mattip>
mjacob: is there a way to manage pull requests: mention PRs from the issues and issues from the PRs, get visual cues when they are closed?
blueyed has joined #pypy
<kenaan>
stevie_92 cpyext-gc-cycle 0b73323d80cc /: Fixed some issues with tracking/deallocating cpyext gc objects Added rudimentary support for lists (gc ...
Zaab1t has joined #pypy
Zaab1t has quit [Client Quit]
marky1991 has joined #pypy
Rhy0lite has joined #pypy
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
lritter has joined #pypy
<mjacob>
mattip: the review system is not based on pull requests
<mattip>
mjacob: I have gotten used to conveniently marking comments from issues in pull requests and visa-versa on github
<mjacob>
mattip: i think it should be possible
<mjacob>
the other question is whether we like sourcehuts review system, which internally is based on emails
<mjacob>
what will we use it for? internal reviews or just external contributions?
<mattip>
what do you mean "review system"? The todo link above or something else?
<mattip>
"You need to log in to submit ticket" - if it is email based, can we set up a mail account for anonymous issues where we need to approve them before they appear on the tracker?
dddddd has quit [Ping timeout: 245 seconds]
<mjacob>
mattip: no, we're talking past each other
<mjacob>
sourcehut has different concepts
<mjacob>
sorry, typo
<mjacob>
sourcehut has different components
<mjacob>
the ticket tracker ("todo") is one, the repository hosting is another
<mjacob>
and then there's yet another component for doing reviews, and it's based on emails
<mjacob>
the idea behind sourcehut's review component is that people can do "old-fashioned" reviews on the mailing list, but to make it more user-friendly there's a web UI for renders the review mail thread in a nice way
<mjacob>
mattip: i asked about commit-ticket-and-vice-versa references and it turns out that it's not yet implemented
<antocuni>
I think we should aim to be as mainstream as possible: using what everybody else uses brings tons of advantages
<ronan>
+1
<antocuni>
that's why I'd like github, if only Armin didn't veto it :)
moei has joined #pypy
<mjacob>
antocuni: different people have different opinions about what "other people" use ;)
<mjacob>
the first question is: what's our goal?
<mjacob>
the second question is: how do we get there?
<mjacob>
if we primarily want to increase the number of drive-by contributions: then yes maybe, let's switch to git and github
<mjacob>
but before that we should set up a gihub mirror accepting pull requests and see whether people actually contribute
<marky1991>
do you really expect drive-by contributions for pypy? seems too hard for that imo
<antocuni>
I'd like to have a better development workflow, with pull requests, code review, a better issue tracker, a continuous integration which actually works, etc.
<antocuni>
github is the de facto standard and proven to work
<antocuni>
gitlab is also good and used by many people
<antocuni>
bitbucket is admittedly not very good
<mjacob>
i won't disagree about bitbucket
<mjacob>
i tend to disagree with the rest
<mjacob>
i don't think that by switching to github, we'll start doing "proper" code reviews
<mjacob>
as in, i don't think that we don't do code reviews because of lack of tooling. it's more a cultural thing.
<antocuni>
true
<antocuni>
it's also true that as of now, we can't even try to change our culture because we don't have the tools for doing that
<mjacob>
also, i observed that people which want to use powerful issue tracking and code review support tend to use something else
<mjacob>
sure, it's a chicken-and-egg problem
<mjacob>
although i don't think that bitbucket's review support is so bad that it blocks us from doing more reviews
marky1991 has quit [Remote host closed the connection]
<antocuni>
to start with, it's hard to integrate with CI; I'd like to be able to open a PR, ask for review, and have a reassuring "all tests are passing" before pressing "merge"
<antocuni>
I mean, it is probably possible to achieve the same with bitbucket. It's just harder because it's a 2nd class citizen
cadr__ has left #pypy [#pypy]
<mjacob>
not sure about this. did anyone ever try whether triggering CI in case of a pull request is possible?
<mjacob>
(i'm thinking about the common case, where the to-be-reviewed branch is already in the repository)
<mjacob>
so it's more a merge request than a pull requst actually ;)
<antocuni>
that's exactly what I am talking about; with github and gitlab it's a feature which is already there and working. With other systems, you need effort
<mjacob>
what do you mean by "2nd class citizen"?
<mjacob>
github doesn't have included CI support (bitbucket has it, BTW)
<antocuni>
ok, it seems you are arguing for the sake of argue. Sorry, I don't feel like
<mjacob>
i'm sorry to hear that you think this
<antocuni>
well, if you open a random github project, you see that it is straightforward to setup a CI workflow. Yes, it is technically not offered by github itself, but that's not my point
<mjacob>
sure, i agree that it's easier to set up
marky1991 has joined #pypy
marky1991 has quit [Ping timeout: 272 seconds]
antocuni has quit [Ping timeout: 245 seconds]
kipras has joined #pypy
NemeXis has quit [Remote host closed the connection]
Rhy0lite has quit [Quit: Leaving]
<kenaan>
mattip win32-encoding b5fb07545a94 /pypy/: merge py3.6 into branch
<njs>
bitbucket's CI is extremely not compelling from a quick look. the free-for-open-source tier is useless for anything except tiny projects, and it only supports linux.
kipras has quit [Ping timeout: 268 seconds]
<mjacob>
njs: i wasn't suggesting to use it. it was more a side remark.
agronholm has quit [Ping timeout: 252 seconds]
_whitelogger has joined #pypy
agronholm has joined #pypy
forgottenone has quit [Quit: Konversation terminated!]
marky1991 has joined #pypy
marky1991 has quit [Remote host closed the connection]