| [12:36:18] | * gusaus has quit (Ping timeout: 245 seconds) |
| [17:51:10] | * David_Hernandez has joined #aegir |
| [17:57:39] | * boshtian has joined #aegir |
| [19:04:50] | * ybabel has joined #aegir |
| [19:09:13] | * colan has quit (Ping timeout: 245 seconds) |
| [19:16:29] | * colan has joined #aegir |
| [19:43:01] | * ybabel has quit (Ping timeout: 260 seconds) |
| [19:43:05] | * ybabel1 has joined #aegir |
| [19:45:26] | * ybabel1 is now known as ybabel |
| [23:40:00] | * zombiebeard has joined #aegir |
| [23:43:49] | * noecc has joined #aegir |
| [00:27:57] | * ybabel has quit (Remote host closed the connection) |
| [00:28:07] | * ybabel has joined #aegir |
| [00:55:41] | * noecc has quit (Ping timeout: 260 seconds) |
| [01:18:56] | * noecc has joined #aegir |
| [01:38:17] | * noecc has quit (Ping timeout: 256 seconds) |
| [02:05:20] | * theMusician has joined #aegir |
| [02:55:10] | * David_Hernandez has quit (Quit: Saliendo) |
| [03:24:39] | * boshtian has quit (Quit: boshtian) |
| [03:42:28] | * noecc has joined #aegir |
| [05:03:19] | * colan waves |
| [05:03:25] | <helmo> | Scrum time.... ergonlog1c bgm jonpugh colan gboudrias memtkmcc |
| [05:03:38] | <jonpugh> | so is that the hangout or the appear in? |
| [05:03:38] | <jonpugh> | :D |
| [05:04:32] | <colan> | we've been texting as a record is autogenerated, even though i'd prefer voice myself. |
| [05:04:33] | <helmo> | i'n in appear in just in case ... |
| [05:04:53] | <colan> | can we record the session there? can't remember. |
| [05:05:27] | <helmo> | both have advantages ... nut no appear in does not offer a recoring option |
| [05:06:09] | <jonpugh> | helmo: can you post a link and/or put it in the topic |
| [05:06:24] | <colan> | jonpugh: https://appear.in/aegir |
| [05:06:42] | <jonpugh> | ty |
| [05:07:54] | <helmo> | I've been preparing for a release in https://www.drupal.org/node/2790229 |
| [05:07:55] | <hefring> | https://www.drupal.org/node/2790229 => [meta] 3.8 release (bugfix/patches) [#2790229] => 3 comments, 2 IRC mentions |
| [05:09:25] | <helmo> | I'll just start rabling here ... |
| [05:09:34] | <helmo> | I already merged the boa unfork branch |
| [05:10:36] | <helmo> | We need to fix or revert https://www.drupal.org/node/2774029#comment-11740645 ... |
| [05:10:37] | <hefring> | https://www.drupal.org/node/2774029 => Allow services to be disabled [#2774029] => 5 comments, 1 IRC mention |
| [05:10:39] | <colan> | yes, thanks. looks like we all have time to post patches, but not review them. :) |
| [05:11:05] | <colan> | reverting it makes sense until it's fixed i think. |
| [05:12:39] | <colan> | helmo: how extensive is the test converage that you & jonpugh have been working on? |
| [05:12:40] | <helmo> | The weird thing is that my attempt to reproduce was missing the provision part of the patch ... after applying that it came back green again ... which made me wonder if travis is using the right provision code |
| [05:13:01] | <helmo> | colan: far from complete :( |
| [05:13:22] | <colan> | well, if it actually caught something, thumbs up, right? :) |
| [05:14:32] | <colan> | looking into that provision issue would be a good idea then. |
| [05:15:55] | <helmo> | jonpugh: would it be possibe to access a screenshot, or the full html page where behat gives a failure? |
| [05:16:15] | * colan is reviewing the release notes. |
| [05:16:25] | <jonpugh> | The easiest way is to add a step "Then print last response" |
| [05:16:50] | <jonpugh> | then the HTML output will appear in travis logs |
| [05:17:04] | * gusaus has joined #aegir |
| [05:17:15] | <jonpugh> | beyond that you need to write some code to catch the error and save the file in the behat FeatureContext class. |
| [05:17:30] | <helmo> | can we do that only when an error occurs? |
| [05:17:31] | <jonpugh> | but you'd have to save it somewhere remotely because travis builds disappear after run |
| [05:19:10] | <jonpugh> | I've done it with Behat in the past but never saved it anywhere :| |
| [05:19:13] | <jonpugh> | you can only get a screenshot if you fire up a selenium container |
| [05:20:17] | <jonpugh> | but again, you need to have a place to save it. |
| [05:20:38] | <jonpugh> | I'd love to try and get it to save to the github PR... |
| [05:21:37] | <helmo> | normally reproducing it locally in docer would be good enought ... but I just installed it and it seems to run fine |
| [05:22:42] | <helmo> | I'v also tried h_wordpress again this afternoon, after fiddling a bit to get wp-cli in the right place it worked nice. |
| [05:23:11] | <helmo> | I'm tempted to just add it to the golden contrib list of the makefile ... it's listed as experimental anyway.... thoughts? |
| [05:24:49] | <bgm> | helmo: to be honest I'm not sure how much time I'm ready to commit to hosting_wordpress |
| [05:24:49] | <gusaus> | wordpress working on aegir? |
| [05:24:55] | <bgm> | it also has very limited functionality |
| [05:25:02] | <bgm> | i.e. install and disable, mostly |
| [05:25:41] | <bgm> | although if there are patches I'm more than willing to review and merge. I do use it in prod. |
| [05:25:44] | <colan> | in that case, maybe we should invest more in a generic solution via hosting_kubernetes. |
| [05:26:02] | <colan> | ergonlogic says he got wordpress working with it. :) |
| [05:26:19] | <helmo> | bgm: thanks for mentioning that ... |
| [05:27:37] | <colan> | https://github.com/kubernetes/kubernetes/tree/master/examples/mysql-word... |
| [05:28:17] | <helmo> | we can also just mention it in the release notes and evaluate again with the nect release. |
| [05:28:21] | <bgm> | depends which resources we're investing :) |
| [05:28:30] | <colan> | bgm: yours? :P |
| [05:28:46] | <bgm> | hah, well, I'm not migrating to kubernetes tomorrow :) |
| [05:29:08] | <bgm> | (maybe in 6-12 months, not closing the door either) |
| [05:29:44] | <colan> | ergonlogic demoed it last week. i think it was only me & memtkmcc that were around for it. that seems to be the aegirNG we've been looking for. |
| [05:30:57] | <colan> | it can host anything that's in kubernetes yaml files / docker containers. that's the concept. |
| [05:31:11] | <bgm> | yep, I understand, and I think it's a good long term solution |
| [05:32:09] | <colan> | i worked on it with him to get it off the ground, but i'm not going to have too much time in the next little while myself. so feel free to help him with it, folks. :) |
| [05:32:32] | <bgm> | I'm just a bit spread thin, and most of my Aegir time goes into hosting_civicrm, tiny bit in hosting_wordpress. If people start adopting hosting_wordpress, they need clear expectations for now (either fund the work, or submit patches) |
| [05:33:59] | <colan> | bgm: how about it we put a note up there saying that support is minimal, and it's being deprecated long-term by the kubernetes/aegirNG more generic app container solution? |
| [05:34:20] | <bgm> | works for me :) |
| [05:34:35] | <colan> | it'll lower expectations, and possible drive folks to help with the new thing. |
| [05:34:45] | <colan> | or at least start asking questions. |
| [05:35:00] | <jonpugh> | to be clear we're just talking release notes, right? |
| [05:35:21] | <jonpugh> | hosting_wordpress shouldn't be included in aegir unless it's planned to be there long term I think |
| [05:35:43] | <colan> | agreed. i was thinking of putting the note on the hosting_wordpress project page. |
| [05:35:44] | <helmo> | jonpugh: true |
| [05:35:49] | <bgm> | it could be in the hosting_wordpress description, which is showed in the settings page |
| [05:36:10] | <bgm> | most people won't read the release notes, they'll just want to enable the option in aegir |
| [05:36:13] | <colan> | wherever it's seen, i guess. |
| [05:36:27] | <colan> | ok, then in there makes sense. |
| [05:36:35] | <bgm> | jonpugh: I think it should be planned as a long term objective, but that's just me :) |
| [05:37:13] | <colan> | it is, but it'll be even longer if we don't start moving in that direction. ;) |
| [05:37:46] | <colan> | though, ergonlogic may be done given how much time he's been putting into it lately. :) |
| [05:38:05] | <jonpugh> | as soon as you put that in there, people will try it and expect it to work and be supported as is into the future, and complain when it's not, then you're on the hook for an upgrade path... |
| [05:38:37] | <jonpugh> | As a policy, let's start putting in mentions of notable contribs in the release notes so people can check things out on their own |
| [05:39:18] | <colan> | good idea. |
| [05:39:19] | <jonpugh> | we're very close to being able to deploy anything on docker, so I'd say let's point to hosting_wordpress for those who want to experiment |
| [05:39:43] | <jonpugh> | And as a segue, let's have more visible release notes, like in our blog which could live in the Atrium site I created this week |
| [05:39:55] | <jonpugh> | http://aegir.atrium.devshop.build/ |
| [05:40:21] | <colan> | we could add a paragraph explaining the relationship between the two of them. short vs. longer term. stress that experimental stuff is experimental. |
| [05:41:08] | <bgm> | I feel like there's too much difference between aegir & aegirNG, that at this point it's probably better to focus on AegirNG |
| [05:41:17] | <bgm> | it sends contradictory signals to users |
| [05:42:20] | <colan> | so let's just mark it deprecated then, and link to the experimental project. users can then make their own decisions. |
| [05:42:49] | <colan> | nobody's stopping them from using hosting_wordpress if they want. |
| [05:43:02] | <bgm> | you can mark hosting_wordpress as deprecated when aegir 3.x will be deprecated :) |
| [05:43:39] | <colan> | i wish. ok, sorry. |
| [05:44:04] | <colan> | bgm: it's your project, what do you think? |
| [05:44:27] | <colan> | or rather, how do you want to proceed? |
| [05:44:45] | <bgm> | I don't have an objection to including as golden-"experimental"-contrib if it has a clear description that it's really minimal. |
| [05:44:48] | * memtkmcc has joined #aegir |
| [05:45:00] | <bgm> | at the same time, we need a clear message to users, and if we start to promote aegirNG, probably better to focus on that. |
| [05:45:02] | <helmo> | It's now fairly neutral on http://docs.aegirproject.org/en/3.x/release-notes/3.8/ ... the project page can explain it in more detain I guess |
| [05:45:51] | <bgm> | cool release notes |
| [05:45:51] | <helmo> | We'll hold off on adding it to the makefile |
| [05:46:57] | <colan> | helmo: maybe just add this to the end of what you have for kubernetes: "... to allow Aegir to host any containerized application, not just Drupal." |
| [05:47:20] | <colan> | ? |
| [05:47:25] | <colan> | looks good overall. |
| [05:47:52] | <memtkmcc> | oh, thanks for merging boa unfork stuff! |
| [05:47:59] | * theMusician has quit (Quit: theMusician) |
| [05:48:00] | <helmo> | colan: ok i'll add that |
| [05:49:15] | <jonpugh> | Ok, I finally have visual proof of concept of the new tasks ajax: https://drive.google.com/file/d/0B3MXiVV32vTNeTRuTTVHS0pOeGM/view |
| [05:50:25] | <memtkmcc> | helmo: can you review my latest two patches, or should I just commit them? I don't want to mess anything around release time, though… |
| [05:50:42] | * ybabel has quit (Quit: ybabel) |
| [05:50:45] | <memtkmcc> | I mean |
| [05:50:47] | <memtkmcc> | https://www.drupal.org/node/2820747 |
| [05:50:48] | <hefring> | https://www.drupal.org/node/2820747 => Nginx - Support for Accelerated Mobile Pages (AMP) [#2820747] => 2 comments, 1 IRC mention |
| [05:51:00] | <colan> | jonpugh: looks great! so it refreshes every second? |
| [05:51:00] | <memtkmcc> | https://www.drupal.org/node/2820486 |
| [05:51:01] | <hefring> | https://www.drupal.org/node/2820486 => Nginx - do not log media files requests [#2820486] => 2 comments, 1 IRC mention |
| [05:51:30] | <jonpugh> | colan: yup. |
| [05:51:45] | <jonpugh> | vue.js is awesome |
| [05:51:55] | <colan> | i like how the counter keeps going after the task is done :) |
| [05:51:55] | <jonpugh> | notice all the available tasks appear dynamically |
| [05:52:06] | <jonpugh> | timeago plugin turns any datestamp |
| [05:52:11] | <jonpugh> | into dynamic counter |
| [05:52:19] | * colan nods |
| [05:52:34] | <jonpugh> | well <time datetime="2016T01234"></time> |
| [05:52:36] | <colan> | if the're all doing that every second, could get annoying though. |
| [05:53:00] | <colan> | should stop on completion, and then update every minute? |
| [05:53:07] | <jonpugh> | I have it configured to only show one unit, so only tasks that have happened in the last minute will visibly change |
| [05:53:23] | <colan> | oic. |
| [05:53:24] | <jonpugh> | not needed, this little plugin handles the updates |
| [05:53:57] | <jonpugh> | https://github.com/rmm5t/jquery-timeago/wiki/Sites-that-use-Timeago |
| [05:54:06] | <jonpugh> | craigslist, arstechnica, github status |
| [05:54:29] | <colan> | you can add aegir. |
| [05:55:15] | <helmo> | memtkmcc: I can't really speak on nginx patches ... colon? |
| [05:55:24] | <helmo> | colan: ^^ |
| [05:56:08] | <colan> | i've been looking at those as they come in, but the recent ones i have not as it's stuff i haven't been working with. |
| [05:56:12] | <colan> | e.g. AMP |
| [05:56:19] | <jonpugh> | that last one Looks like a small enough change to me |
| [05:57:20] | <colan> | maybe just commit that stuff. when i update, it'll break, then we'll know. :) |
| [05:57:42] | <colan> | but i'm fairly confident given that memtkmcc has been runnning this stuff for a while now. |
| [05:57:49] | <memtkmcc> | colan: it is not really about AMP, more about not breaking indexing sites using paging |
| [05:58:05] | <memtkmcc> | simple stuff, in fact |
| [05:58:12] | <colan> | memtkmcc: then just commit it :) |
| [05:58:23] | <memtkmcc> | okay, thanks! |
| [05:58:58] | <colan> | not that it's entirely up to me... but if nobody's reviewing it, and it works for you, go for it. |
| [06:00:22] | <colan> | personally, i commit simple stuff directly, and bigger things wait for review. like the conversation we're having about https://www.drupal.org/node/2819021 |
| [06:00:22] | <hefring> | https://www.drupal.org/node/2819021 => Set new temporary file directory variable for Drupal 8 [#2819021] => 8 comments, 1 IRC mention |
| [06:01:09] | <colan> | maybe that's a good model. |
| [06:01:15] | <colan> | thoughts? |
| [06:02:00] | <colan> | memtkmcc: ^^^ |
| [06:02:33] | * theMusician has joined #aegir |
| [06:03:02] | * anarcat has quit (Quit: offlining) |
| [06:03:16] | <colan> | in other news, i released a beta1 for https://www.drupal.org/project/hosting_services |
| [06:03:45] | <colan> | so feel free to play with it. |
| [06:04:02] | <colan> | that's been in alpha for as long as i can remember. |
| [06:04:58] | <colan> | there was some talk about including it in golden contrib, but i'm not sure how that would work if it has some dependencies. |
| [06:05:11] | <helmo> | nice |
| [06:05:30] | <helmo> | we'd also have to add the dependancies |
| [06:05:51] | <colan> | services services_api_key_auth hosting_variables |
| [06:06:04] | <memtkmcc> | colan: I think that helmo suggestion is a good one, since we already have separate templates for core versions, why not to hardcode their stuff there (for tmp path) instead of trying to fix the current logic? |
| [06:06:47] | <colan> | memtkmcc: you mean just do everything in the template? |
| [06:08:09] | <memtkmcc> | colan: as much as possible, I think, if we don't need to make it dynamic anyway? we only need to make the site name in the path dynamic, in fact |
| [06:08:19] | <helmo> | the left part of the '=' |
| [06:08:42] | <colan> | sure. |
| [06:08:56] | <memtkmcc> | yeah |
| [06:09:17] | <colan> | right, okay. that's what he wrote. :) |
| [06:09:26] | <colan> | forgot... |
| [06:11:37] | <colan> | anyone have anything else? |
| [06:13:28] | <helmo> | I'll try to get the release out tomorrow... |
| [06:13:58] | <colan> | excellent! i'll be in the co-op channel if anyone wants to discuss co-op stuff. cheers! |
| [06:14:43] | <helmo> | jonpugh: if you have a moment please checkout why https://travis-ci.org/aegir-project/hosting/builds/169324314 is failing ... the whole testsuite runs fine in my local docker dev env |
| [06:15:31] | <ergonlogic> | sorry all, I thought the scrum was at 3pm |
| [06:16:32] | <colan> | helmo: and thanks for getting https://www.drupal.org/node/2595809 in there. |
| [06:16:33] | <hefring> | https://www.drupal.org/node/2595809 => Find (d8) profiles without a .profile [#2595809] => 15 comments, 3 IRC mentions |
| [06:17:58] | <jonpugh> | hmmmm |
| [06:18:13] | <colan> | ergonlogic: ok for the co-op meeting though? other channel. |
| [06:18:16] | <jonpugh> | colan: helmo: I might be seeing a new bug caused by that |
| [06:18:31] | <colan> | :( |
| [06:18:35] | <jonpugh> | https://usercontent.irccloud-cdn.com/file/t7BP6wND/Install%20profile%3A%... |
| [06:18:59] | <helmo> | hmm :( |
| [06:19:15] | <colan> | nice. what happends if you install it? :) |
| [06:19:29] | <jonpugh> | good question! |
| [06:20:04] | <jonpugh> | readme gets added as a package and saved as the install profile |
| [06:20:22] | * colan puts on a helmet in case bad things happen. |
| [06:20:59] | <jonpugh> | nah just fails to install "readme" profile |
| [06:21:04] | <jonpugh> | same thing if a profile is missing |
| [06:21:20] | <jonpugh> | however this is a start reminder of how difficult it is to see the errors for a task |
| [06:21:34] | <jonpugh> | took like 10 clicks to find out why it went red :D |
| [06:22:28] | <colan> | can anyone else reproduce? |
| [06:22:54] | <colan> | given that you said "I might be seeing..." |
| [06:23:23] | <jonpugh> | let me reinstall hostmaster from 7.x-3.x real quick |
| [06:23:39] | <colan> | actually, let me know if i can test that. think i'm running that patch on dev. |
| [06:23:44] | <colan> | er, let me see |
| [06:26:22] | <jonpugh> | yeah, same thing |
| [06:26:33] | <jonpugh> | https://usercontent.irccloud-cdn.com/file/ImJ2828e/brand%20new%20reinstall |
| [06:27:11] | <jonpugh> | ohhh wait |
| [06:27:15] | <jonpugh> | it's in provision? |
| [06:27:46] | <helmo> | yes the patch was for provision ... but I'm also seeing it now |
| [06:27:59] | <jonpugh> | attempting a revert and reverify |
| [06:31:25] | <jonpugh> | hm, still picks it up |
| [06:31:26] | <helmo> | I guess we'd have to parse the yml file and check for type: profile ?? |
| [06:32:09] | <jonpugh> | how does it parse for modules/themes now? |
| [06:35:13] | <helmo> | the readme only seems to be shoing for d7 platforms ... |
| [07:12:55] | <colan> | helmo: jonpugh: i reverted & am still getting it. could it be from something else? |
| [07:14:38] | <colan> | after i reverified platforms again, but didnt' try a new install. |
| [07:16:14] | <jonpugh> | No idea, I don't know if that's the cause, it just clicked they are both profile related? |
| [07:34:57] | * theMusician has quit (Quit: theMusician) |
| [07:39:12] | * mengi has quit (Quit: Leaving.) |
| [07:46:22] | * theMusician has joined #aegir |
| [07:57:17] | <ergonlogic> | so, as far as aegirNG goes, I've been trying to write up a bit of a spec, based on the resource/application/deployment model |
| [07:58:19] | <ergonlogic> | I'm thinking of starting a new D8 distro (probably actually called "aegir") with a focus initially on the server/task stuff |
| [07:58:32] | <ergonlogic> | since that's pretty much a hard blocker for progress in D7 |
| [07:58:45] | <ergonlogic> | our current tasks can;t attach to anything but nodes |
| [07:59:15] | <ergonlogic> | whereas we should probably shift almost everything to entities in D8 |
| [08:04:09] | <jonpugh> | I messed around with composer and d8 in a new branch in hostmaster: 8.x-4.x. There's nothing there yet but conversion of .info to .info.yml and adding a composer.json. I was playing around with using composer to require hostmaster, so I added it to packagist: https://packagist.org/search/?q=aegir |
| [08:04:26] | <jonpugh> | so you might want to look at that ergonlogic |
| [08:04:44] | * Guest53930 has quit () |
| [08:05:11] | <jonpugh> | I would consider the name of the install profile carefully. I've debated this a lot myself, and if we name anything aegir we'll always be saying "aegir the install profile" instead of just "hostmaster" |
| [08:05:21] | * elijah has joined #aegir |
| [08:14:38] | <ergonlogic> | perhaps, but I meant more broadly, as in everything using that namespace (i.e., 'aegir_tasks', 'aegir_deployments', 'aegir_role_apache', etc.)k |
| [08:15:35] | <ergonlogic> | whether it's 'aegir' or 'aegir_profile'... I prefer the former, but I suppose the latter is clearer |
| [08:36:30] | * theMusician has quit (Quit: theMusician) |
| [08:42:13] | * memtkmcc has quit (Quit: Leaving.) |
| [08:50:56] | * theMusician has joined #aegir |
| [08:57:43] | * theMusician has quit (Quit: theMusician) |
| [09:00:38] | * theMusician has joined #aegir |
| [09:05:20] | * zombiebeard has quit (Quit: zombiebeard) |
| [09:32:53] | <colan> | helmo: jonpugh: can you guys test https://www.drupal.org/node/2595809 ? it would be great if we could get it into the release. |
| [09:32:54] | <hefring> | https://www.drupal.org/node/2595809 => Find (d8) profiles without a .profile [#2595809] => 19 comments, 4 IRC mentions |
| [09:33:41] | <colan> | i think i nailed it, but my environment's a bit messed up right now & i won't be able to rebuild until next week. |
| [09:33:51] | <colan> | so i don't completely trust it. :) |
| [09:57:48] | * theMusician has quit (Quit: theMusician) |
| [10:38:51] | * theMusician has joined #aegir |
