Re: [Twisted-Python] overview: new review queue venue

2016-05-24 Thread Jonathan Vanasco
On May 24, 2016, at 6:49 PM, Glyph wrote: > I KNOW, RIGHT!!! However, protected statuses somewhat reduce the potential > race-condition here. And contributing to a couple dozen Github projects I > have to say that practically this has never been an issue, even though I find > it aestheticall

Re: [Twisted-Python] overview: new review queue venue

2016-05-24 Thread Glyph
> On May 24, 2016, at 2:49 PM, Jonathan Vanasco wrote: > > just a few thoughts: > > The current system as-explained seems to use an "Issue" as a queue item that > is either a "bug report" or a "notice of a pull request, which may also > reference another bug report". In the current system, w

Re: [Twisted-Python] overview: new review queue venue

2016-05-24 Thread Jonathan Vanasco
just a few thoughts: The current system as-explained seems to use an "Issue" as a queue item that is either a "bug report" or a "notice of a pull request, which may also reference another bug report". that is weird. IMHO, a quality contributor will never get turned off by handling the docs,

Re: [Twisted-Python] overview: new review queue venue

2016-05-24 Thread Craig Rodrigues
Hi, Thanks for bringing up these points. See: https://twistedmatrix.com/trac/ticket/8352 Feel free to add any thoughts you may have on how to improve this. -- Craig On Tue, May 24, 2016 at 10:37 AM, Jonathan Vanasco wrote: > FWIW, I just noticed that the CONTRIBUTING message wasn't showing u

Re: [Twisted-Python] overview: new review queue venue

2016-05-24 Thread Jonathan Vanasco
FWIW, I just noticed that the CONTRIBUTING message wasn't showing up when trying to make a PR. github is supposed to make a warning box that shows the contents of /CONTRIBUTING or /CONTRIBUTING.md when opening an issue or PR ___ Twisted-Python mailing

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Glyph
> On May 23, 2016, at 11:56 AM, Clayton Daley wrote: > > Closing PRs will make them less likely to be found by searchers. If every PR > has an issue (common for bug fixes, less common for new features), this is > less of a problem -- is this something the bot would need to verify/fix? The bo

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Clayton Daley
> > I mostly frequent Physics Stack Exchange > . At any time some number of the > questions on the front page are either "closed" or "on hold" for not living > up to site standards in some way. We've asked almost exactly the same > question as is being asked in th

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Daniel Sank
FWIW I thought of another "open source" community which uses a similar idea to closing pull requests if they won't be accepted in their current form: Stack Exchange. I mostly frequent Physics Stack Exchange . At any time some number of the questions on the front

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Glyph
> On May 22, 2016, at 10:51 PM, Tristan Seligmann > wrote: > > Note that even without the bot, I believe you can just create a new PR for > the same branch, so it's not *too* bad, but definitely a little clunky. Many of my comments have had to do why we want this kind of process generally, r

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Glyph
> On May 22, 2016, at 9:39 PM, meejah wrote: > > > Personally, I find closing PRs that aren't going to be merged "soon" > mostly-beneficial. Even if it *might* be perceived as "hostile" by some > contributers, a simple explanation should suffice. (And, if simply > closing a PR with a nice note

Re: [Twisted-Python] overview: new review queue venue

2016-05-23 Thread Glyph
> On May 22, 2016, at 18:12, Clayton Daley > wexternalote: > > "You can't: > require test coverage, > require documentation, > require coding standard compliance, > require people to file a ticket before sending a patch to the mailing list, > The first three of these are *already* norms in all

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Tristan Seligmann
Note that even without the bot, I believe you can just create a new PR for the same branch, so it's not *too* bad, but definitely a little clunky. On Mon, 23 May 2016 at 06:59 Amber "Hawkie" Brown wrote: > Mark has been working on a bot which would reopen it with a comment: > https://github.com/

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Amber "Hawkie" Brown
Mark has been working on a bot which would reopen it with a comment: https://github.com/markrwilliams/txghbot - Amber > On 23 May 2016, at 12:56, Craig Rodrigues wrote: > > I submitted this PR, which is now closed: > https://github.com/twisted/twisted/pull/62 > > I don't want to re-open that

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Craig Rodrigues
I submitted this PR, which is now closed: https://github.com/twisted/twisted/pull/62 I don't want to re-open that PR, but I am using that as an example As an example, if I wanted to re-open that, how would I go about doing it? I am not an administrator of the Twisted GitHub project, so on that we

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread meejah
Personally, I find closing PRs that aren't going to be merged "soon" mostly-beneficial. Even if it *might* be perceived as "hostile" by some contributers, a simple explanation should suffice. (And, if simply closing a PR with a nice note explaining, "please re-open when X is fixed/changed" scares

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread meejah
Glyph writes: > This is a very useful data point. I do not have any concrete > experience with it and I was kind of wondering about this. FWIW, Tahoe-LAFS *just* started using AppVeyor too, and I also find it horrifically slow. That said, the Tahoe tests run pretty slowly on a VirtualBox window

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Donald Stufft
Twisted has been enforcing these rules since before they were considered part of the norm and I believe that Glyph was referencing is that back then people said that Twisted was going to fail or w/e because of requiring those things. Sent from my iPhone > On May 22, 2016, at 9:12 PM, Clayton D

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Clayton Daley
> > "You can't: > >- require test coverage, >- require documentation, >- require coding standard compliance, >- require people to file a ticket before sending a patch to the >mailing list, > > The first three of these are *already* norms in all of the OSS projects of this calibe

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 22, 2016, at 6:18 AM, Clayton Daley wrote: > > The thing is, if you perceive it as "hostile" that a project closes a PR - > i.e. "says that they're not going to do more work on it" - that is a cultural > problem; it suggests a certain implicit level of passive aggression in > opening

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Clayton Daley
> > The thing is, if you perceive it as "hostile" that a project closes a PR - > i.e. "says that they're not going to do more work on it" - that is a > cultural problem; it suggests a certain implicit level of passive > aggression in opening a PR which I don't want to assume. It's sort of like > h

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Tristan Seligmann
On Sun, 22 May 2016 at 10:12 Ralph Meijer wrote: > On May 22, 2016 9:36:28 AM GMT+02:00, Glyph > wrote: > >[..] > >(Please nobody try to do the clever thing where you configure buildbot > >to automatically pull all PRs, that would effectively negate any > >security protections...) > > > >I've be

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Ralph Meijer
On May 22, 2016 9:36:28 AM GMT+02:00, Glyph wrote: >[..] >(Please nobody try to do the clever thing where you configure buildbot >to automatically pull all PRs, that would effectively negate any >security protections...) > >I've been assuming that in the worst-case scenario, we'd do what >Cryptogr

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 22, 2016, at 12:24 AM, Craig Rodrigues wrote: > > On Sat, May 21, 2016 at 6:04 PM, Glyph > wrote: > > Right now, we need to manually vet each change before sending it to > buildbots, because they are shared mutable environments that we can't afford >

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 15:23, Glyph wrote: > > >> On May 22, 2016, at 12:15 AM, Hynek Schlawack wrote: >> >> Ah finally a fine bike shedding thread that gets everyone involved. ;) > > OKAY NOW THAT I'VE GOT YOU ALL HERE LET'S TALK ABOUT > https://twistedmatrix.com/trac/ticket/288 > > *slams

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 22, 2016, at 12:15 AM, Hynek Schlawack wrote: > > Ah finally a fine bike shedding thread that gets everyone involved. ;) OKAY NOW THAT I'VE GOT YOU ALL HERE LET'S TALK ABOUT https://twistedmatrix.com/trac/ticket/288 *slams a metal grating shut over the only exit from the mailing list

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Craig Rodrigues
On Sat, May 21, 2016 at 6:04 PM, Glyph wrote: > > Right now, we need to manually vet each change before sending it to > buildbots, because they are shared mutable environments that we can't > afford to have running untrusted code automatically. > This is quite useful actually. We would need a t

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 22, 2016, at 12:15 AM, Hynek Schlawack wrote: > > Ah finally a fine bike shedding thread that gets everyone involved. ;) > >> Right now, we need to manually vet each change before sending it to >> buildbots, because they are shared mutable environments that we can't afford >> to have

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 22, 2016, at 12:09 AM, Hynek Schlawack wrote: > > A lot of projects do follow this workflow, and maybe it will be fine for us. The real question is; is FreeBSD support really worth it for the cost to contributors, since that's the only platform we currently support bu

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Hynek Schlawack
Ah finally a fine bike shedding thread that gets everyone involved. ;) > Right now, we need to manually vet each change before sending it to > buildbots, because they are shared mutable environments that we can't afford > to have running untrusted code automatically. If we could switch to Travi

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Hynek Schlawack
>>> A lot of projects do follow this workflow, and maybe it will be fine for >>> us. The real question is; is FreeBSD support really worth it for the cost >>> to contributors, since that's the only platform we currently support but >>> can't test? >> >> I'm guessing that we have more FreeBSD

Re: [Twisted-Python] overview: new review queue venue

2016-05-22 Thread Glyph
> On May 21, 2016, at 11:36 PM, Amber Hawkie Brown > wrote: > > >> On 22 May 2016, at 14:32, Glyph wrote: >> >> >>> On May 21, 2016, at 11:21 PM, Amber Hawkie Brown >>> wrote: >>> >>> On 22 May 2016, at 14:15, Glyph wrote: Sorry, I guess I wasn't clear. I know that PRs

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:42 PM, Craig Rodrigues wrote: > > On Sat, May 21, 2016 at 11:37 PM, Glyph > wrote: > > I realize it can feel like that sometimes, but Google Analytics suggests the > large majority of our visitors (45%) are on Windows. By contrast, 0.0

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:42 PM, Daniel Sank wrote: > > I can speak for ~20 scientific research groups who use Twisted via LabRAD's > python API . A lot of us use or deploy to > Windows at least some times. So that's around 200 people you've never heard > of

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:24 PM, Daniel Sank wrote: > > That does leave the question of whether you want people to use merge commits > or to rebase their branch on the latest master before merging so that you > always get a fast-forward. Github will never fast-forward, and I never want to see

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Craig Rodrigues
On Sat, May 21, 2016 at 11:37 PM, Glyph wrote: > > I realize it can feel like that sometimes, but Google Analytics suggests > the large majority of our visitors (45%) are on Windows. By contrast, > 0.05% are on FreeBSD. Granted, that's a *very *high percentage of > FreeBSD clients for the Inter

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Daniel Sank
I can speak for ~20 scientific research groups who use Twisted via LabRAD's python API . A lot of us use or deploy to Windows at least some times. So that's around 200 people you've never heard of who use Twisted on Windows :)

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:24 PM, Daniel Sank wrote: > > Squashing commits is essential to making useful commit histories. Nope. It's just a handy hack to work around the commonly-used, broken history viewers (like Github's own) that can't correctly present multi-parent commits. If you use So

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:37 PM, Glyph wrote: > > the large majority of our visitors Sorry. "plurality" :). -glyph___ Twisted-Python mailing list Twisted-Python@twistedmatrix.com http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:30 PM, Amber Hawkie Brown > wrote: > > >> On 22 May 2016, at 14:23, Glyph wrote: >> >> A lot of projects do follow this workflow, and maybe it will be fine for us. >> The real question is; is FreeBSD support really worth it for the cost to >> contributors, since t

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 14:32, Glyph wrote: > > >> On May 21, 2016, at 11:21 PM, Amber Hawkie Brown >> wrote: >> >> >>> On 22 May 2016, at 14:15, Glyph wrote: >>> >>> Sorry, I guess I wasn't clear. I know that PRs are presently a potential >>> alternative to a diff, and that we are still u

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:21 PM, Amber Hawkie Brown > wrote: > > >> On 22 May 2016, at 14:15, Glyph wrote: >> >> Sorry, I guess I wasn't clear. I know that PRs are presently a potential >> alternative to a diff, and that we are still using Trac for ticketing. I >> want to make it possible

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 14:23, Glyph wrote: > > A lot of projects do follow this workflow, and maybe it will be fine for us. > The real question is; is FreeBSD support really worth it for the cost to > contributors, since that's the only platform we currently support but can't > test? I'm gues

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 14:24, Daniel Sank wrote: > > All, > > > Please do not use squash commits. See > > http://mjg59.dreamwidth.org/42759.html. > > Squashing commits is essential to making useful commit histories. Are you > just saying not to use Github's built-in feature which squashes eve

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:08 PM, Amber Hawkie Brown > wrote: > > The issue in 2016 is not actually the kernel; but OpenSSL and OpenSSH, among > other system libraries. Every new version of Fedora has been red on the > buildbots for this reason; OpenSSL changed, and we needed to fix our use of

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Daniel Sank
All, > Please do not use squash commits. See http://mjg59.dreamwidth.org/42759.html. Squashing commits is essential to making useful commit histories. Are you just saying not to use Github's built-in feature which squashes everything into a _single_ commit? If so, note that you can turn that GUI

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 11:04 PM, Adi Roiban wrote: > > I am not sure that Travis supports Python on OSX It does. We test on OS X quite extensively on , including a py2app app bundle. > and it might take some time until there is support for Python on Wi

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 14:15, Glyph wrote: > > Sorry, I guess I wasn't clear. I know that PRs are presently a potential > alternative to a diff, and that we are still using Trac for ticketing. I > want to make it possible to avoid using Trac for ticketing; perhaps switching > to github issues

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 10:52 PM, Craig Rodrigues wrote: > > On Sat, May 21, 2016 at 3:12 PM, Glyph > wrote: > Hooray! We're on github now. Next: there's the question of how to deal with > pull requests? > > A few people, including myself modified the text with

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Daniel Sank
Dear all, While I'm just a lurker, having used Twisted and Github for some time in a moderately sized team I would like to offer a couple comments: > This is exactly why issues and PRs should be separated. If you only have one > artifact - the PR - to represent both the issue and the potential s

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Amber "Hawkie" Brown
> On 22 May 2016, at 06:12, Glyph wrote: > > It is therefore tempting to map it into GitHub via labels and webhooks and > bot workflows. However, I think a better mapping would be this: > > • Proposing: Just open a pull request. Any open pull request should be > treated as part of the

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Adi Roiban
On 22 May 2016 at 02:04, Glyph wrote: > > > On May 21, 2016, at 4:00 PM, Werner Thie wrote: > > > > Thank you all for that great and massive move! > > > > Me running twisted almost exclusively on FreeBSD, what is the reason for > > > > > We could also potentially just replace our buildbot build

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Craig Rodrigues
On Sat, May 21, 2016 at 3:12 PM, Glyph wrote: > Hooray! We're on github now. Next: there's the question of how to deal > with pull requests? > A few people, including myself modified the text with Git instructions: https://twistedmatrix.com/trac/wiki/TwistedDevelopment#SubmittingaPatch The b

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 7:25 PM, Clayton Daley wrote: > > To qualify my comments, I've yet to contribute to Twisted because I don't > have a good enough grasp of its internals, but I have contributed to a > variety of Git-based OSS projects. I definitely get uneasy with the general > idea that

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Clayton Daley
To qualify my comments, I've yet to contribute to Twisted because I don't have a good enough grasp of its internals, but I have contributed to a variety of Git-based OSS projects. I definitely get uneasy with the general idea that we're trying to "replicate workflow A from Trac in tangentially rel

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 3:57 PM, Terry Jones wrote: > > Hi Glyph > > On Sun, May 22, 2016 at 12:12 AM, Glyph > wrote: > Reviewing: This is the potentially slightly odd part. I believe a review > that doesn't result in acceptance should close the PR. > > This fe

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Glyph
> On May 21, 2016, at 4:00 PM, Werner Thie wrote: > > Thank you all for that great and massive move! > > Me running twisted almost exclusively on FreeBSD, what is the reason for > > > We could also potentially just replace our buildbot build farm with a > > combination of appveyor and travis-c

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Werner Thie
Thank you all for that great and massive move! Me running twisted almost exclusively on FreeBSD, what is the reason for > We could also potentially just replace our buildbot build farm with a > combination of appveyor and travis-ci; this would remove FreeBSD from > our list of supported platform

Re: [Twisted-Python] overview: new review queue venue

2016-05-21 Thread Terry Jones
Hi Glyph On Sun, May 22, 2016 at 12:12 AM, Glyph wrote: > - Reviewing: This is the potentially slightly odd part. I believe a > review that doesn't result in acceptance should *close* the PR. This feels wrong to me. I find github pull requests very useful, in ways that it sounds like your sug