June 25th, 2015 Symfony Docs Meeting

log.txt

-06-25 10:59:33] <weaverryan> o/ everyone!
[2015-06-25 10:59:55] <javier_eguiluz> hi :)
[2015-06-25 11:00:01] <weaverryan> I meant to email a proposed topic list... but I'm suddenly babysitting a 2 year old today :)
[2015-06-25 11:00:38] <weaverryan> Here are 4 topics (some may not need much discussion) that I had in mind from previous conversations and emails:
[2015-06-25 11:00:42] <WouterJ> hi Ryan and Javier (and xabbuh!)
[2015-06-25 11:00:46] <weaverryan> A) docs workflow
[2015-06-25 11:00:46] <weaverryan> B) auto-deploy platform.sh
[2015-06-25 11:00:46] <weaverryan> C) linking to more 3rd party bundles
[2015-06-25 11:00:46] <weaverryan> D) fixing forms, doctrine, security
[2015-06-25 11:01:26] <xabbuh> hi :)
[2015-06-25 11:01:33] <WouterJ> (I probably have to leave in 45 minutes)
[2015-06-25 11:02:01] <weaverryan> Good, then we'll be efficient!
[2015-06-25 11:02:15] <weaverryan> So, let's get started :)
[2015-06-25 11:02:18] <weaverryan> A) docs workflow
[2015-06-25 11:02:28] <weaverryan> which includes:
[2015-06-25 11:02:30] <weaverryan> - more mergers/voters
[2015-06-25 11:02:30] <weaverryan> - needing to make many changes - better collaboration
[2015-06-25 11:02:30] <weaverryan> - abandoned PR's
[2015-06-25 11:02:39] <weaverryan> and anything else to improve how we work
[2015-06-25 11:02:59] <weaverryan> From the email, I think one easy win is to make xabbuh a merger
[2015-06-25 11:03:23] <WouterJ> +1 for making xabbuh a merger. I'm having less and less time to help maintaining the docs
[2015-06-25 11:03:24] <weaverryan> And a second win would be to use the 2 +1's for merging system from the core
[2015-06-25 11:03:25] <javier_eguiluz> +1 (if he agrees)
[2015-06-25 11:03:44] <xabbuh> +1 for both suggestions
[2015-06-25 11:03:57] <weaverryan> (unlike the core code, there are many really small things, and the fastest person to see them should just merge them and give a happy note)
[2015-06-25 11:04:12] <weaverryan> Beautiful! Thanks xabbuh - we should have done it a long time ago :)
[2015-06-25 11:04:25] <weaverryan> On the 2 +1 thing, I think that should not apply to small/obvious chances
[2015-06-25 11:04:27] <weaverryan> changes*
[2015-06-25 11:04:32] <WouterJ> I also agree with introducing voting. At the moment, many PRs are awaiting on the native-speaker-review, this can be solved if all Ryan has to do is commenting +1
[2015-06-25 11:04:33] <weaverryan> we should just merge those
[2015-06-25 11:04:59] <weaverryan> WouterJ Exactly - sometimes I'm on my phone :)
[2015-06-25 11:05:00] <xabbuh> I also agree with merging small things immediately
[2015-06-25 11:05:45] <weaverryan> Perfect
[2015-06-25 11:05:46] <WouterJ> I think we can apply the same rules as previously: The 3 of us can merge once: (a) it's a small/obviously correct change; (b) it has 2 +1's and 0 -1
[2015-06-25 11:06:06] <WouterJ> where (b) is then in fact replacing what was up to ryan behaviour
[2015-06-25 11:06:09] <WouterJ> s/behaviour/before
[2015-06-25 11:06:22] <xabbuh> indeed
[2015-06-25 11:06:30] <weaverryan> Well said :)
[2015-06-25 11:06:34] <javier_eguiluz> @WouterJ +1 it makes a lot of sense to do that
[2015-06-25 11:07:11] <WouterJ> ok, weaverryan do you sent a mail to fabien to add xabbuh to the gh contributors, then I'll create a PR updating the docs :)
[2015-06-25 11:07:34] <xabbuh> I already have access to it
[2015-06-25 11:07:47] <WouterJ> ah, cool :)
[2015-06-25 11:07:48] <weaverryan> I thought you might - even easier :)
[2015-06-25 11:07:49] <xabbuh> I just didn't have to use it before
[2015-06-25 11:08:08] <weaverryan> Well, now you're officially dangerous
[2015-06-25 11:08:39] <WouterJ> as for the voters, what about adding Javier as a voter?
[2015-06-25 11:09:01] <WouterJ> (and maybe Iltar if he wants?)
[2015-06-25 11:09:20] <weaverryan> Oh yes, that's up to Javier - I think clearly he has plenty of experience and activity to merit being a voter
[2015-06-25 11:09:20] <xabbuh> good idea imho, Javier seems to read everything anyway :)
[2015-06-25 11:09:34] <weaverryan> (and be rescuing a lot of old PR's)
[2015-06-25 11:10:05] <Iltar> Heya, sorry I was at the story
[2015-06-25 11:10:52] <javier_eguiluz> I don't know if more voters is better for the workflow ...
[2015-06-25 11:11:02] <javier_eguiluz> ... but if you agree, I'd be honored to be able to vote
[2015-06-25 11:12:42] <weaverryan> We'll see :). Let's start with 4 voters. I would be for adding Iltar in the future if it'll help things (and if he wants to be a voter)
[2015-06-25 11:12:48] <weaverryan> Next problem on this topic: minimizing the back-and-forth for contributors and reducing "abandoned" or "stalled" PR's
[2015-06-25 11:12:49] <xabbuh> I think it helps, at least when they valuable feedback like you do
[2015-06-25 11:13:28] <Iltar> weaverryan, on the docs? If desired I can always help out, but I'm not really busy with the docs atm
[2015-06-25 11:13:38] — Iltar is actually have a week off this week
[2015-06-25 11:13:56] <weaverryan> Iltar: That's my impression too - you're somewhat active, but not super active
[2015-06-25 11:14:01] <weaverryan> Well, glad you're joining us anyways :)
[2015-06-25 11:14:09] <javier_eguiluz> regarding the abandoned PRs ... I'm taking over them and finishing them (most of the times the needed work is minimal)  I don't know if this is a good strategy and I don't know if the people whom I "steal" PRs agree with this
[2015-06-25 11:14:35] <WouterJ> okay, we'll test with 4 voters first then
[2015-06-25 11:14:48] <weaverryan> I actually think we should formalize the "stealing" of PR policy - e.g. after 2 pings and 2 weeks of no reply, it's available for "stealing" 
[2015-06-25 11:14:54] <weaverryan> of course, we will find a better term for this ;)
[2015-06-25 11:14:56] <Iltar> weaverryan, I got into GameForge (game publisher) as Board Moderator and Applied to be part of QA, so I was rather busy :)
[2015-06-25 11:15:02] <WouterJ> javier_eguiluz, from my experience, as long as you keep their commits, they are very happy to have someone take over
[2015-06-25 11:15:24] <WouterJ> having long open PR isn't a nice experience for contributors
[2015-06-25 11:15:42] <weaverryan> I think so too - we just say "thanks, your commits are included on the new PR" - I think we're already doing that nicely :)
[2015-06-25 11:16:00] <xabbuh> I agree, we could write this down somewhere so that others could feel comfortable too to take over stalled PRs
[2015-06-25 11:16:07] <weaverryan> exactly
[2015-06-25 11:16:31] <WouterJ> isn't "taking over PRs" actually a way to work around a problem (long open PRs), rather than a fix of that problem?
[2015-06-25 11:16:43] <weaverryan> strictly in theory, we could even have something that shows up PR's that need a ping or should be re-labeled as "ready to be taken over"
[2015-06-25 11:16:54] <xabbuh> two pings after the last change each with a break of two weeks would mean that a PR has not been worked on for at least four weeks, that sounds okay to me
[2015-06-25 11:17:08] <weaverryan> WouterJ Can we fix it if the person who contributed doesn't respond with the changes?
[2015-06-25 11:17:17] <xabbuh> WouterJ: depends on the PR
[2015-06-25 11:17:39] <xabbuh> we sometimes have PRs where we do the review with a long delay, in this case it's more our fault
[2015-06-25 11:18:10] <weaverryan> yep, it's sometimes our fault for sure
[2015-06-25 11:18:19] <WouterJ> well, we can fix some of the trouble causing people to stop working on it. For instance, some PRs get a review long after the PR was opened, others have 4 rounds of reviews (which makes me as a reviewer even desperate), etc.
[2015-06-25 11:18:19] <weaverryan> that's one of those problems that's shared with the core code
[2015-06-25 11:19:11] <weaverryan> Yea, we need to limit the rounds of changes
[2015-06-25 11:19:21] <weaverryan> sometimes, I'll come to a PR after a few rounds of changes, and it's still not righ
[2015-06-25 11:19:24] <weaverryan> right*
[2015-06-25 11:19:29] <weaverryan> So I'll merge and make the changes myself
[2015-06-25 11:19:37] <weaverryan> or merge and open a new PR if those changes are big enough
[2015-06-25 11:19:43] <xabbuh> yeah, probably it's better to just merge something after two rounds of review and tweak things ourselves
[2015-06-25 11:19:44] <weaverryan> I think that's a good idea
[2015-06-25 11:20:07] <weaverryan> Agreed - as long as the merger can make those tweaks immediately (so as to not forget)
[2015-06-25 11:20:38] <WouterJ> yeah, I still think letting people finish the PR itself is great. But for some small changes, merging and fixing probably is better
[2015-06-25 11:20:46] <WouterJ> or when we already had a round of fixing
[2015-06-25 11:20:55] <weaverryan> So, limit 2 rounds of changes? Meaning: if the user has made updates after comments, we can only make comments once more
[2015-06-25 11:21:25] <weaverryan> And this doesn't have to be a hard rule - if it's someone I know like Iltar, I might make him get it perfect - we can judge by how active and excited the user is
[2015-06-25 11:21:26] <WouterJ> To be honest, I remember Ryan fixing all my PRs some years ago and I don't know if I would be happy if he had commented instead.
[2015-06-25 11:21:35] <weaverryan> If there was a long delay after the first round of changes, then maybe we limit to just 1 change
[2015-06-25 11:22:12] <weaverryan> WouterJ Do you mean fixing them was better than commenting, or the opposite?
[2015-06-25 11:22:24] <WouterJ> fixing better than commenting
[2015-06-25 11:22:37] <weaverryan> Cool, I would hope so
[2015-06-25 11:22:58] <weaverryan> I want to be able to say "this is so awesome! I fixed some meaningless small details, but 99% of it was perfect!"
[2015-06-25 11:23:02] <xabbuh> yeah, I think it's too frustrating if you don't see any progress
[2015-06-25 11:23:17] <xabbuh> at a certain point you get the feeling that the PR will never get accepted
[2015-06-25 11:23:34] <WouterJ> I would actually propose one round of changes
[2015-06-25 11:23:51] <WouterJ> it's enough to introduce people to our standards and to feel like their contributing is important to us
[2015-06-25 11:24:08] <javier_eguiluz> when you talk about "round of changes" ... you mean 1 round per reviewer or 1 round in total?
[2015-06-25 11:24:31] <WouterJ> but it still allows to merge quickly. (and saying "Great job fixing the comments! I made some other small changes [...]." is nicer than "Great job fixing the comments! I've added some more...."
[2015-06-25 11:24:52] <WouterJ> most of the time, we review at more or less the same moment
[2015-06-25 11:25:15] <Iltar> I love making my PR perfect :D
[2015-06-25 11:25:23] <weaverryan> One round of changes as a soft rule - the exception being if it's not mergeable after 1 commit (the problems are fundamental - e.g. the PR is perfect)
[2015-06-25 11:25:37] <javier_eguiluz> @Iltar "perfect" is the enemy of "done" ;)
[2015-06-25 11:25:37] <weaverryan> Iltar: That's good feedback - we've gotta try to see how excited the person is in making those first changes
[2015-06-25 11:25:47] <weaverryan> This clearly won't be a perfect rule :)
[2015-06-25 11:25:55] <weaverryan> Javier - I like your point ;)
[2015-06-25 11:26:13] <Iltar> I know, I'm a perfectionist, it's dangerous if I don't stop myself
[2015-06-25 11:26:18] <xabbuh> I think we shouldn't write this rule down, just keep it in mind when thinking about whether to merge or to comment
[2015-06-25 11:26:27] <weaverryan> xabbuh: I was just about to say that :)
[2015-06-25 11:26:34] <weaverryan> And we can revisit
[2015-06-25 11:26:49] <weaverryan> So the goal is to merge as soon as something has the +1's and is "mergeable" - we make small tweak changes afterwards
[2015-06-25 11:26:58] <xabbuh> +1
[2015-06-25 11:27:44] <weaverryan> It's getting back a little bit of the "speed" of running a smaller project - that's nice :)
[2015-06-25 11:27:46] <WouterJ> let's test with that indeed (I see some problem with voting, as the voter !== merger, meaning that additional changes of the voter are lost, but let's see how it works in the current state)
[2015-06-25 11:28:10] <weaverryan> agreed
[2015-06-25 11:28:23] <weaverryan> I'd like to move onto a different topic - anything else on this one for now?
[2015-06-25 11:28:31] <javier_eguiluz> let's move on
[2015-06-25 11:28:33] <WouterJ> let's move on
[2015-06-25 11:28:37] <weaverryan> B) auto-deploy platform.sh
[2015-06-25 11:28:38] <xabbuh> we can always refine the workflow if we see where it doesn't fit
[2015-06-25 11:28:45] <weaverryan> very quickly - we are ready to do this, but...
[2015-06-25 11:28:59] <weaverryan> I need to get permission from Fabien that he's ok with us using and "promoting" that we're using a commercial service
[2015-06-25 11:29:07] <weaverryan> I don't expect and issue, but I've already emailed him
[2015-06-25 11:29:30] <weaverryan> And that's all - I'll move on now to the next topic, before we lose Wouter :)
[2015-06-25 11:29:30] <WouterJ> fabien originally sent the mail to me, so he at least should be ok with us using it
[2015-06-25 11:29:44] <xabbuh> I could imagine that he wouldn't be happy with offering space for advertising banners, the rest didn't sound too critical
[2015-06-25 11:29:55] <weaverryan> Oh yea, good point
[2015-06-25 11:30:08] <weaverryan> Yea, we wouldn't do that - just README, documentation, etc
[2015-06-25 11:30:15] <weaverryan> the stuff we would have done naturally in order to explain how it works
[2015-06-25 11:30:17] <xabbuh> alright
[2015-06-25 11:30:33] <weaverryan> If we don't hear in a few days, I think we'll just do it and say that javier_eguiluz said we could :p
[2015-06-25 11:30:36] <xabbuh> so I agree to move on, nothing else we could do here right now
[2015-06-25 11:30:40] <weaverryan> Next topic
[2015-06-25 11:30:42] <weaverryan> C) linking to more 3rd party bundles
[2015-06-25 11:31:01] <weaverryan> javier_eguiluz: this was something Javier messaged me about, and he's already done it in a few places
[2015-06-25 11:31:05] <weaverryan> I'm not sure there's much to discuss
[2015-06-25 11:31:11] <weaverryan> but I wanted to bring it up briefly
[2015-06-25 11:31:31] <weaverryan> Javier - anything you want to say on this? Other than "we should do more of this"?
[2015-06-25 11:31:43] <WouterJ> Somewhat related, I would propose to reopen https://github.com/symfony/symfony-docs/issues/4871 again
[2015-06-25 11:31:49] <xabbuh> so, this is about whether or not we link to other bundles if it fits, or something more than that?
[2015-06-25 11:31:52] <javier_eguiluz> to me the real issue with bundles it's not about the Symfony Docs ... but about bundles themselves (they need more exposure, more contributors, better documentation, etc. .... but that's a discussion for another moment)
[2015-06-25 11:32:35] <WouterJ> xabbuh: I think it's about what javier did in a PR some days ago, adding some small tip boxes linking to interesting/popular 3rd party bundles
[2015-06-25 11:32:46] <weaverryan> Javier - yea, that is a bigger issue indeed, one that I would like to talk about (but not now)
[2015-06-25 11:33:51] <xabbuh> giving tips about bundles that could be useful is imho a good thing
[2015-06-25 11:34:06] <weaverryan> Ok, so that's an easy win - let's continue to look where we can do this
[2015-06-25 11:34:08] <xabbuh> linking to other blog articles seems a bit difficult to me
[2015-06-25 11:34:24] <weaverryan> About 4871, it's difficult to me, but if we see a spot where it makes sense, let's consider it
[2015-06-25 11:34:26] <javier_eguiluz> one quick question: would you agree to add short cookbook articles about bundles
[2015-06-25 11:34:52] <javier_eguiluz> I'm thinking about something very popular that Symfony developers need and only HWIOAuthBundle provides: how can I longin in my app with Twitter/Facebook/Google
[2015-06-25 11:35:04] <weaverryan> For example, if we had a cookbook on Vagrant, we might link to http://www.whitewashing.de/2013/08/19/speedup_symfony2_on_vagrant_boxes.html, which is basically an industry standard blog post at this point and a trusted author
[2015-06-25 11:35:26] <weaverryan> javier_eguiluz: Interesting
[2015-06-25 11:35:54] <xabbuh> hm, I think that absolutely depends on the topic
[2015-06-25 11:35:55] <weaverryan> It's an easy +1, but the question is: is that article basically a very short entry with a link to docs on HWIOauth or do we actually put the tutorial in the Symfony docs
[2015-06-25 11:35:57] <WouterJ> adding a cookbook just to link to a bundle: -1, adding a cookbook explaining how one would implement something like and then say "There is a bundle for this!" +1
[2015-06-25 11:36:23] <xabbuh> usually, I would say it's better documented where there are people who really know anything about the bundle
[2015-06-25 11:36:40] <xabbuh> on the other hand, the Symfony docs can make those topics more popular
[2015-06-25 11:36:47] <weaverryan> "adding a cookbook just to link to a bundle" I'm +1 for this. Better if we have an article showing you how to do it without the bundle AND a link, but sometimes using the bundles is the only thing that makes sense
[2015-06-25 11:36:53] <javier_eguiluz> I was thinking about putting the article in Symfony docs: something like: install the bundle as explained in their documentation . Now, open securit.yml and add this, create your user entity, etc.
[2015-06-25 11:37:30] <WouterJ> imo, we shouldn't document bundles in our documentation (like we do for Doctrine, which kind of is a big mistake imo, but can't be changed anymore)
[2015-06-25 11:37:31] <javier_eguiluz> the same with another very popular and common need: uploading files. Short article explaining VichUploadBundle
[2015-06-25 11:37:33] <weaverryan> javier_eguiluz: I think maintenance and versioning is a problem. And to put it another way, if we use our efforts to write articles on other docs, then we can help shape and improve those
[2015-06-25 11:37:44] <xabbuh> I would feel better if we could group several use cases in one article and then link each use case to a dedicated article in the bundle's docs
[2015-06-25 11:38:33] <javier_eguiluz> ok ... anyway my idea was only to dedicate efforts to the few popular things in web development: Oauth integration, file uploading, payments, etc.)
[2015-06-25 11:38:35] <weaverryan> Yep: I like: Cookbook article that explains the issue, introduces the bundle, and links to a "cookbook" on their documentation for this
[2015-06-25 11:38:38] <Iltar> I think it would be very useful to link to new bundles which are an addition to the community
[2015-06-25 11:38:51] <Iltar> Bundles that solve common problems and don't bring too much issues along
[2015-06-25 11:39:07] <weaverryan> javier_eguiluz: I couldn't agree more that we should absolutely have entries about these topics, even if they are lightweight links
[2015-06-25 11:39:29] <Iltar> I would ask consent from the writer as the maintainer might get a lot more work load which he or she might not have the time to deal with
[2015-06-25 11:39:57] <Iltar> It would also be nice (in case of multiple bundles) to give multiple options to the user
[2015-06-25 11:40:36] <weaverryan> Haha, Javier (and I) hate options ;) - unless they really cover different cases
[2015-06-25 11:40:58] <weaverryan> Ok, so how can we wrap this up to some productive steps?
[2015-06-25 11:41:01] <Iltar> I know, but why would bundle A be listed and not bundle B?
[2015-06-25 11:41:02] <javier_eguiluz> @weaverryan exactly my words! A "clear path" always win over the "illusion of availability"
[2015-06-25 11:41:36] <weaverryan> Iltar: because we +1 A over B based on our best experience/opinion
[2015-06-25 11:41:36] <Iltar> If they both do the same, how will you determine which is better suited?
[2015-06-25 11:42:04] <weaverryan> just subjective opinion. Someone needs to make that decision, should it be us? Or a beginner trying to solve their problem?
[2015-06-25 11:42:17] <weaverryan> I think there will be few cases where there are 2 high quality libraries solving the same problem
[2015-06-25 11:42:29] <weaverryan> But if there really is, then we could list both - but I don't see it happening much
[2015-06-25 11:42:38] <Iltar> I know, it was just an edge case :)
[2015-06-25 11:42:39] <weaverryan> (unless again, 2 libs really cover 2 different use-cases)
[2015-06-25 11:42:48] <WouterJ> if the bundles are exactly the same, I would open an issue in one of them if collaboration isn't better ;) Otherwise, we can always add some options like: If you need to do A, use X, if you need B use Y
[2015-06-25 11:42:49] <weaverryan> Iltar: you and your edge cases ;)
[2015-06-25 11:42:56] — Iltar has experience with them
[2015-06-25 11:43:07] <javier_eguiluz> out of memory I can only think of a use case where two differente bundles are very popular: KnpPaginator and PAgerfanta
[2015-06-25 11:43:14] — Iltar mumbles something about murphies law
[2015-06-25 11:43:42] <Iltar> javier_eguiluz, that would be one case
[2015-06-25 11:43:45] <weaverryan> So, let's try to wrap this up. I think we're agreed on these articles in theory, but not what they contain exactly
[2015-06-25 11:44:07] <javier_eguiluz> that's right
[2015-06-25 11:44:10] <Iltar> Knp would be an easy choice because of the "Knp", but I would recommend PagerFanta over the knp variant
[2015-06-25 11:44:30] <weaverryan> I use Pagerfanta o:)
[2015-06-25 11:44:42] <Iltar> Same, works pretty well (also under the hood)
[2015-06-25 11:44:52] <Iltar> Anyway, going OT
[2015-06-25 11:46:14] <weaverryan> I would prefer to have the actual tutorial inside the bundles themselves. But, in the interest of getting things done faster, I would be for putting the tutorial in the cookbook entry itself, and then trying to "move" it to the bundle later
[2015-06-25 11:46:26] <weaverryan> In a perfect world, it'd be in the Symfony docs, because we have more eyes on it and it renders a lot better
[2015-06-25 11:46:43] <weaverryan> the 2 problems being: that's more for us to maintain and handling the bundle's versioning is an issue
[2015-06-25 11:47:26] <xabbuh> which could be solved by rendering more bundles' docs on symfony.com
[2015-06-25 11:47:43] <xabbuh> which in turn means more work for Javier/SensioLabs
[2015-06-25 11:47:52] <javier_eguiluz> @xabbuh we're on it ... there is an updated bundles doc section pending to be merged/deployed
[2015-06-25 11:48:00] <weaverryan> Yea... and really, if it's such an important topic, than I think we *do* want the bundle's docs on symfony.com
[2015-06-25 11:48:16] <WouterJ> why is writing an article for bundle docs harder than writing one for the symfony docs?
[2015-06-25 11:48:23] <weaverryan> Javier - would that make the process to add a new bundle or a new version for a bundle more automatic?
[2015-06-25 11:49:02] <weaverryan> WouterJ: Mostly because some of them are poorly maintained, or their docs might be such a mess that we really need to organize things in  their docs first. But, that's probably a good problem
[2015-06-25 11:49:04] <javier_eguiluz> @weaverryan nope ... but that's not a problem ... we prefer to add bundles manually. Our long-term intention is to host the docs of the Top100 bundles
[2015-06-25 11:49:21] <weaverryan> Javier: What about new versions at least?
[2015-06-25 11:49:29] <javier_eguiluz> I know that there are 2,600 bundles out ther ... but most of them are irrelevant, duplicates, abandoned, etc.
[2015-06-25 11:49:30] <weaverryan> (I agree that adding new bundles is no big deal to be manual)
[2015-06-25 11:49:35] <WouterJ> (versioning is a problem)
[2015-06-25 11:50:11] <javier_eguiluz> what if we only shoe the most recent version?
[2015-06-25 11:50:25] <weaverryan> I think that's a problem, in practice :/
[2015-06-25 11:50:34] <weaverryan> e.g. there are 2 very different versions of FOSUserBundle
[2015-06-25 11:50:56] <weaverryan> What about the YML configuration file approach you were talking about on Twitter with lsmith?
[2015-06-25 11:50:57] <javier_eguiluz> OK ... we'll figure out the details
[2015-06-25 11:51:05] <xabbuh> there will be more bundles with more than one maintained version
[2015-06-25 11:51:15] <weaverryan> Ok cool - that's good enough for now
[2015-06-25 11:51:39] <weaverryan> So, we *will* add these very important "How do I login with FAcebook, etc" type of articles
[2015-06-25 11:51:52] <weaverryan> But the actual tutorials will be inside the bundle, which will be rendered in symfony.com anyways
[2015-06-25 11:52:27] <weaverryan> javier_eguiluz: Should we make an issue on the docs with a check-list of some of the most important ones?
[2015-06-25 11:52:34] <weaverryan> Then we can check in monthly at this meeting with progress
[2015-06-25 11:52:49] <javier_eguiluz> I think it's a very good idea!
[2015-06-25 11:53:30] <weaverryan> Ok, we have 7 minutes to look at the last topic briefly :)
[2015-06-25 11:53:37] <weaverryan> D) fixing forms, doctrine, security
[2015-06-25 11:53:48] <weaverryan> Last hack-day, we identified these as the biggest "problems" for people
[2015-06-25 11:53:58] <weaverryan> Obviously, we can't fix them or find a fix in 6 minutes
[2015-06-25 11:54:01] <WouterJ> isn't it a better idea to move this topic to slot (a) next month?
[2015-06-25 11:54:05] <weaverryan> but perhaps we can find one small thing to do
[2015-06-25 11:54:14] <weaverryan> I'd be fine with that honestly
[2015-06-25 11:54:23] <WouterJ> as it's quite a bit topic (during the hackday, we spent an hour just identifing them)
[2015-06-25 11:54:39] <xabbuh> the only quick fix I can think about is documenting the outstand options
[2015-06-25 11:54:49] <xabbuh> but that probably won't help much
[2015-06-25 11:54:53] <javier_eguiluz> a quick comment: I think forms is "unsolvable", security could be simplified a lot and doctrine is "solvable" but with a lot of work
[2015-06-25 11:55:43] <weaverryan> javier_eguiluz: "unsolvable" isn't very actionable ;)
[2015-06-25 11:56:12] <xabbuh> I would have said that Doctrine would be harder to tackle
[2015-06-25 11:56:35] <weaverryan> Anyways, I'm happy that it sounds like people have some ideas
[2015-06-25 11:56:56] <WouterJ> imo, from a doc perspective, forms needs a lot more "What if I want ..." articles, Security more global overview article (there is none at the moment) and Doctrine needs a better documentation it self (at the moment, the good docs are incomplete, on symfony.com, and the "less good" docs are complete, on doctrine-project.org)
[2015-06-25 11:57:12] <weaverryan> You have to remember: a lot of the documentation was how Ryan figured out how to do/explain things 4 years ago. That means that how we explain things and what we explain might be far from ideal
[2015-06-25 11:57:42] <weaverryan> WouterJ I'll not all of this for next time
[2015-06-25 11:58:16] <WouterJ> at the moment, I am doing a lot of work on the Security component (learning the internals, closing issues, etc). So I expect to review most of the Security articles somewhere in this period as well
[2015-06-25 11:58:19] <WouterJ> already sent some PRs
[2015-06-25 11:59:00] <weaverryan> WouterJ: That's fantastic :)
[2015-06-25 11:59:45] <weaverryan> And of course, I think my Guard system will save everything for security in 2.8 :p (I'm going to release a lib/bundle soon too, so people can try/find edge cases before 2.8)
[2015-06-25 11:59:57] <weaverryan> Ok guys, we're out of time and topics
[2015-06-25 12:00:10] <weaverryan> thanks for coming! Nice to have time to focus on docs strategy
[2015-06-25 12:00:26] <weaverryan> And I feel like there are topics/innovations for at least a few more months worth of meetings :)
[2015-06-25 12:00:27] <xabbuh> just a quick question, did any of you heard of this conference http://www.writethedocs.org/conf/eu/2015/ before?
[2015-06-25 12:00:31] <javier_eguiluz> thank you all! It was a productive meeting.
[2015-06-25 12:00:35] <weaverryan> Yes I have, but have never been
[2015-06-25 12:00:42] <weaverryan> there's one on the US too
[2015-06-25 12:01:38] <weaverryan> javier_eguiluz: btw, about HWIOAuthBundle, it's basically abandoned currently. I have been emailing with stylod and another guy who runs the HWI organization, but have literally gotten zero replies - been emailing them for almost 1 month
[2015-06-25 12:01:39] <xabbuh> I see, thanks
[2015-06-25 12:02:08] <javier_eguiluz> @weaverryan what about "taking over" the bundle under the FOS org?
[2015-06-25 12:02:25] <WouterJ> who of FOS will take it over?
[2015-06-25 12:02:37] <WouterJ> if you ask me, FOS is short on team members as well
[2015-06-25 12:02:51] <WouterJ> so is Knp, Sensio and almost all other "big" vendors
[2015-06-25 12:03:59] <weaverryan> So true - it's a problem that we should all think about and experiment with for improvements
[2015-06-25 12:04:08] <weaverryan> we are under-manned in every spot
[2015-06-25 12:04:25] <weaverryan> Javier - we are talking about potentially forking it, I'd just love to have a reply from the official guys
[2015-06-25 12:04:26] <WouterJ> anyway, leaving (finally :P). Thanks for the nice meeting! (Ryan, do you create a gist, or should I export the logs from my IRC client?)
[2015-06-25 12:04:30] <weaverryan> otherwise it's a hostile fork
[2015-06-25 12:05:09] <xabbuh> weaverryan: who will maintain the fork?
[2015-06-25 12:05:17] <weaverryan> Anyways, the guys from Elcodi are busy, but have interest in the project continuing and Soullivaneuh is interested in being on the team
[2015-06-25 12:05:22] <weaverryan> I have been emailing with Soullivaneuh as well
[2015-06-25 12:05:39] <weaverryan> Right now, it would need to be Soullivaneuh
[2015-06-25 12:05:50] <weaverryan> Better would be no fork, then it's stloyd and Soullivaneuh
[2015-06-25 12:06:02] <weaverryan> (but Soullivaneuh doesn't have rights yet)
[2015-06-25 12:06:21] <weaverryan> WouterJ: If you create an export, awesome. Otherwise I will :)
[2015-06-25 12:06:21] <javier_eguiluz> @weaverryan sorry for being "naive", but does a bundle like that require that much maintenance?  I mena, it's "only" about integrating the OAuth protocol. Its features are really focused and limited.
[2015-06-25 12:06:50] <weaverryan> Javier: check out the number of PR's and issues: https://github.com/hwi/HWIOAuthBundle
[2015-06-25 12:07:15] <weaverryan> It also has a bunch of work done on a new version, but it's not done yet. It needs a push to that new version. Then, I think it can be more stable for awhile
[2015-06-25 12:07:43] <javier_eguiluz> I don't know ... but Laravel's Socialité has a few commits and it works right for most of people: https://github.com/laravel/socialite
[2015-06-25 12:08:22] <xabbuh> at least, you need someone to reply to user questions, PRs, issues etc.
[2015-06-25 12:09:27] <weaverryan> Javier: I think it does a lot less (I mean that in a good way) - something that the Symfony community should strive for
[2015-06-25 12:09:40] <javier_eguiluz> @weaverryan +1,000,000
[2015-06-25 12:09:41] <xabbuh> otherwise you end with a whole bunch of open PRs, bug reports, feature requests and so on with nobody having knowing about the future steps the maintainers want to take
[2015-06-25 12:09:46] <Iltar> I'll try my best to help around the Security part (and others where I can), usually symfony-docs isn't too high on my priority list when having 5 minutes of waiting
[2015-06-25 12:09:50] <weaverryan> HWIOAuthBundle comes with controllers for finishing registration, etc 
[2015-06-25 12:13:43] <xabbuh> anyway, I have to leave too
[2015-06-25 12:13:47] <xabbuh> enjoy your evening
[2015-06-25 12:13:52] <weaverryan> Yep, later xabbuh!
[2015-06-25 12:13:59] ← javier_eguiluz left (uid41340@gateway/web/irccloud.com/x-tlginzrwlmsasesi)
[2015-06-25 12:14:03] <xabbuh> enjoy your day, Ryan :)