08:02:49 #startmeeting 08:02:49 Meeting started Thu Nov 27 08:02:49 2014 UTC. The chair is rkabalin_. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:02:49 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:03:01 #info Welcome to the 38th Mahara developer meeting. Please state your name and where you are from for the minutes. If you are saying something very long, please put ".." on an empty line by itself to indicate your statement is finished. 08:03:07 #info ghada-laptop is Ghada El-Zoghbi at Catalyst IT, Sydney, Australia. 08:03:10 #info anitsirk is Kristina Hoeppner, Catalyst IT, Wellington, New Zealand 08:03:12 I got some mashed potatoes and a roast chicken from the grocery store, just finishing that up as we speak 08:03:12 #info rkabalin is Ruslan Kabalin, Lancaster University, UK 08:03:13 #info tobiasz is Tobias Zeuch, developer at KIT, Karlsruhe, Germany 08:03:17 #info robertl_ is Robert Lyon, Catalyst IT, Wellington, NZ 08:03:25 #info aarowlaptop is Aaron Wells at Catalyst IT in Wgtn NZ 08:03:28 #info son is Son Nguyen, Catalyst IT, Wellington, NZ 08:04:11 is that everyone I guess 08:04:19 ok 08:04:20 #topic Items from last meeting 08:04:24 #info robertl_ to add info about sslverify 08:04:37 I've updated the notes here with more help information 08:04:43 https://wiki.mahara.org/index.php/Developer_Area/Pushing_Git_Commits#Not_Signed-off-by 08:05:11 it's the item after the not signed off by 08:05:20 https://wiki.mahara.org/index.php/Developer_Area/Pushing_Git_Commits#Server_certificate_verification_failed 08:05:26 #info robertl_ updated the notes at 08:05:36 #link https://wiki.mahara.org/index.php/Developer_Area/Pushing_Git_Commits#Server_certificate_verification_failed 08:05:52 looks good. 08:06:22 ah I copied that url cause of item 2 on list from last meeting :) 08:06:42 #undo 08:06:42 Removing item from minutes: 08:06:50 which is the right one? 08:07:10 rkabalin_: i think you had the correct one. 08:07:15 ok 08:07:22 #link https://wiki.mahara.org/index.php/Developer_Area/Pushing_Git_Commits#Server_certificate_verification_failed 08:07:32 that's the correct one for point 1 08:07:32 looks good the me 08:08:07 thanks robertl_ 08:08:28 ok, next item 08:08:31 #info robertl_ to improve the documents about needing to sign-off a patch 08:08:40 that's my link 08:08:51 #link https://wiki.mahara.org/index.php/Developer_Area/Pushing_Git_Commits#Not_Signed-off-by 08:08:56 ta 08:09:10 also done as I can see 08:09:14 yep 08:09:18 thanks robertl_ 08:09:42 next one 08:09:44 #info kristina to update wiki with links to more of the Mahara groups 08:09:47 #info a list of Mahara groups that are known to me are now listed at https://wiki.mahara.org/index.php/Community_resources/Mahara_users_on_social_media 08:09:49 https://wiki.mahara.org/index.php/Community_resources/Mahara_users_on_social_media 08:10:00 feel free to add others if you know of other places 08:10:29 #info I am also in the process of renaming the "Marketing" section to "Community resources" as marketing does not really fit with what we have in that section 08:10:45 looks good 08:11:24 thanks @anitsirk 08:11:56 if there are no comments, I move to next item 08:12:08 sure. 08:12:11 ok 08:12:14 #info kristina to explain the change in the numbering scheme and then post it wherever I find 08:12:17 #info the announcement is made at 08:12:17 https://mahara.org/interaction/forum/topic.php?id=6530 08:12:25 hi tony. 08:12:33 hello tonyjbutler 08:12:38 Hi everyone 08:12:43 it also lins to aarowlaptop's tracker item. 08:12:51 *links 08:13:15 so far i've have very positive feedback from the users about the version numbering change. 08:13:34 i've also announced it in various social media groups 08:14:07 I am not a big fan of that change, but it is OK ;) 08:14:30 i know, rkabalin_. you were the only odd one out ;-) 08:14:54 * rkabalin_ became a conservative after moving to the UK 08:15:01 )) 08:15:03 lol 08:15:19 ok, next item 08:15:23 maybe you should move to new zealand :-D 08:15:36 actually. next topic 08:15:38 ha 08:15:50 #topic Do we need the "Signed off by" line for commits anymore? (AaronW) 08:15:59 Ah, that's me 08:16:20 AaronW, the floor is yours 08:16:21 We have a coding guideline that says to use a "Signed off by" line on all our commits 08:16:40 yep 08:16:41 #info We have a coding guideline that says to use a "Signed off by" line on all our commits 08:16:46 This is what some heavy-weight projects like the Linux kernel use, in order to certify that the person in the "Signed-off by" line is the author of the commit 08:17:16 though we have author field in git as well 08:17:17 But I think it's over-kill for us. We should just make the "author" line in the commit hold the author, and use the "Signed-off by" as an optional add-on when there are multiple people who contributed to a commit 08:17:40 That's pretty much it. :) 08:17:40 #idea from aarowlaptop: We should just make the "author" line in the commit hold the author, and use the "Signed-off by" as an optional add-on when there are multiple people who contributed to a commit 08:17:53 Does anyone disagree? 08:17:54 aarowlaptop: I agree with that. 08:18:06 I agree 08:18:21 what happens when you cherry-picking old commits like when updating 3rd party plugin? 08:18:23 it's a good way to indicate when there are multiple authors 08:18:24 Fine by me (now I've got into the han 08:18:38 ...into the habit of doing it ;-) 08:19:16 well, cherry-picking doesn't include a new signed off 08:19:22 right now the code is rejected, when it's not signed off, but would it be difficult to reject the multiple-commit automatically when the second author doesn't use the signed-off field? 08:19:51 Well, a git commit only has exactly one "author" field 08:20:04 So there is no way to auto-detect when multiple people contributed to a commit 08:20:17 I'd just remove the check for the "Signed-off" line from the precommit checks entirely 08:20:33 I think this should be a responsibility of author/commiter to specify everyone involved. 08:20:44 Or alternately we could have it as a warning when the "committer" line is different from the "author" line 08:20:50 in the commit message you mean? 08:21:10 rkabalin_^ 08:21:46 aarowlaptop: we can have multiple signed-off lines. 08:21:48 but gerrit would still be able to tell who uploaded the patch anyways, right? Also for the follow-up versions of the patch I mean 08:21:49 Currently it is so to some extend, because as tobiasz2 mentioned, you would not check that there more than one person has been involved, it is responsibility of author or commiter 08:22:09 ghada, yes 08:22:32 ghada: yep, multiple sign-off is possible 08:22:34 Yeah, probably best to just rely on human judgement for this 08:22:47 aarowlaptop: I agree 08:23:03 i don't mind that you force the signed-off. it's just a habit that we get into. 08:23:17 but, that's me... 08:23:31 Yeah, it's not too bad. It just feels like an unnecessary extra step to me. 08:23:51 it is fine to have "Signed-off by" as an optional 08:23:52 What does the sign-off line bring to the table for us? 08:24:09 I'm fine with that as well, as long as gerrit tells me, who uploaded each patch set 08:24:31 @anitsirk: nothing really, it is just a policy 08:24:38 Yep, every commit always has an "author" field, and a "committer" field (which is the person who actually uploaded the commit) 08:25:21 so: is it an empty policy from by-gone days that is not needed anymore? 08:25:37 it seems like new people committing to gerrit do stumble upon it and it causes confusion. 08:26:01 anitsirk: yeah, it got me once. 08:26:07 then it definitely needs to be optional 08:26:10 if forcing the signed-off line does not add any value then why keep it? 08:26:26 It was introduced in the wake of the SCO lawsuit, in regards to linux 08:26:35 It's a legal thing to do with copyright. See http://stackoverflow.com/questions/1962094/what-is-the-sign-off-feature-in-git-for 08:26:50 heh, just reading that :) 08:27:05 :-) 08:27:11 yep, but it is up to the project whether respect that or now 08:27:20 s/now/not 08:27:30 Yep, it's meaningless otherwise. 08:27:40 so from a legal point of view if some one signs of on a patch they are saying it's their work 08:27:57 kind of 08:28:09 #info the Signed-off line as introduced in the wake of the SCO lawsuit, see http://stackoverflow.com/questions/1962094/what-is-the-sign-off-feature-in-git-for for example 08:28:16 Moodle for example does not requre that 08:28:28 Only if they're made aware of what it means in the dev docs. 08:28:57 So, it's probably good to have it optional. And, when there are multiple authors, use it to indicate multiple authors. 08:29:21 #idea probably good to have it optional. And, when there are multiple authors, use it to indicate multiple authors 08:30:06 I'm ok with that. we'll need to update the wiki 08:30:26 I agree with that 08:30:27 do the other present approved code reviewers agree? 08:30:31 I'll volunteer to update the wiki 08:30:33 I agree 08:30:50 robertl_ and rkabalin_ ? 08:30:59 sounds ok to me 08:31:09 #action aarowlaptop update the wiki with "Signed-off" line necessity changes 08:31:14 I agree 08:31:20 rkabalin_: can you please make an #agreed item to note down the decision? 08:31:30 #agreed use the "Signed-off by" as an optional add-on when there are multiple people who contributed to a commit 08:31:45 * robertl_ just reading about the -S flag for signing commit with your GPG key 08:31:57 and aarowlaptop should also make the change then for gerrit 08:32:06 and wondering if that could be useful 08:32:28 #action aarowlaptop change gerrit to accept commits without "Signed-off" line 08:33:14 we used GPG signing on git tags in the past in some internal projects 08:33:39 I guess with having gerrit as a gateway it's not necessary 08:33:49 to sigh the release 08:33:50 @robert: making the GPG signing mandatory might be a big obstacle for community programmers 08:33:56 yeah 08:34:04 I agree with tobiasz2 08:34:17 yeah, now that I've read about it - i'm not keen for it 08:34:29 ok, shall we move the next topic? 08:34:48 sure 08:34:53 sure, I m ok with that. 08:34:58 #topic Do we stay with Dwoo or move to Smarty (or something else)? (robertl_) 08:35:25 Why do we need to move? 08:35:47 we are looking at updating a number of things in mahara, templates/css/javascript 08:36:06 to make things more streamlined/easier for development 08:36:16 I think the main issue was that Dwoo was not supported much anymore, right, robertl_? 08:36:27 Dwoo is old and unsupported 08:36:28 yep 08:36:40 #info robertl_ is looking at updating a number of things in mahara, templates/css/javascript 08:36:43 there does appear to be one guy who has been pushing updates lately 08:36:43 our front-end devs throw in Twig as good alternative. it seems to be popular 08:37:02 so was thinking either smarty3 or twig 08:37:08 so it's not completely dead, just moribund ;) 08:37:16 as both seem to be well supported/modern/fast 08:37:22 heh, I remember the time when we agreed to move to Dwoo )) 08:37:23 i haven't used twig so can't really contribute to this 08:37:25 https://bugs.launchpad.net/mahara/+bug/966001 for some background reading 08:37:49 Smarty is closer to Dwoo so it's probably easier to convert to (we could just run our Smarty -> Dwoo conversion scripts in reverse ;) 08:38:00 changing to smarty 3 would be the least painful as the templating syntax is very similar 08:38:00 but, if that's where things are moving, maybe we should move with it too 08:38:11 and we still have smarty() function )) 08:38:29 twig is being used for projects like Drupal (as part of the Symfony2 framework.) 08:38:32 rkabalin_: Do you remember the rationale for the switch from Smarty to Dwoo? 08:38:33 i actually like smarty... 08:38:56 If we were to do it as part of a framework it would make more sense but twig vs smarty3 stand alone they both seem to be as full featured and speedy as each other. 08:38:59 It was regarded as outdated or something 08:39:02 #info Smarty is closer to Dwoo so it's probably easier to convert to (we could just run our Smarty -> Dwoo conversion scripts in reverse) 08:39:22 My impression from the docs, is that at the time, Smarty was still sticking to compatibility with PHP 4 08:39:40 i know that zend framework is moving away from smarty 08:40:17 but that was years ago, and since then Smarty has gone to Smarty 3, which is PHP 5 based 08:40:27 Smarty 3 requires a web server running PHP 5.2 or greater. 08:40:31 and Dwoo has nearly ground to a halt 08:40:45 ok. 08:40:55 We do have a commit in gerrit, from a year or two back, where one of our Catalyst devs worked on a Twig conversion 08:41:13 Actually I think it's 2 or 3 years back 08:41:16 smarty sounds like it's very much community driven - kinda like us :) 08:41:34 http://meetbot.mahara.org/mahara-dev/2012/mahara-dev.2012-03-27-07.41.html 08:41:35 as stated above, i think Smarty would be the easiest way forward for the project. 08:41:48 #info the last time we discussed templates was in March 2012 (see above link) 08:42:31 which resulted in chris to start converting the templates (we abandonded the patches for the time being) 08:43:09 I guess I'd like to try out Twig firsthand so I know more about it 08:43:20 I see that Twig is released under the new BSD license. 08:43:22 Smarty, as far as I can tell, is nearly identical to Dwoo in functionalit 08:43:23 y 08:43:32 yes 08:43:46 Dwoo is a fork of smarty 08:43:51 is BSD licence ok for us? 08:43:57 license that is 08:44:01 Yep, you can put BSD libraries in a GPL project 08:44:12 #info Smarty, as far as aarowlaptop can tell, is nearly identical to Dwoo in functionality. Thus, moving to it shouldn't be too difficult. 08:45:21 so, I do not mind if we move back to smarty 08:45:36 * rkabalin_ is not conservative in the morning 08:45:55 One advantage of Twig over Smarty, is that it's newer & trendier, thus making Mahara more desirable to developers 08:46:48 do we know what extra features it offers over Smarty? 08:46:51 nope 08:47:10 At least I don't. :) 08:47:17 but it is as already popular as Smarty or would it be another thing to learn for most community developers? 08:47:20 aarowlaptop will you be willing to make a sort of investigation of differences and post a wiki with comparison of benefits and issues we might have 08:47:20 robertl_ a good list for GPL compatibility is https://www.gnu.org/licenses/license-list.html 08:47:29 It is true that we don't want to jump to another project where we'll wind up being the only users of it 08:47:45 on the other hand, Twig has been around at least since 2012 now? So maybe it has staying power? 08:47:51 Sure 08:47:54 from what I've read smarty has more build in functions 08:47:54 I can look into it 08:48:27 Twig has been around since 2009, and had its most recent stable release (1.16) this past July 08:48:47 just compare all aspects of potential concern, e.g. license, easiness of moving to, reliability, etc. 08:48:48 oh, and Symfony uses Twig 08:49:15 twig and symfony are written by same folks I believe 08:49:24 http://stackoverflow.com/questions/15011094/twig-vs-smarty-benchmark 08:49:26 Smarty is on version 3.1.21, and had its last release this past October. So it's still going strong too. 08:49:56 #action aarowlaptop will compare all aspects of Smarty and Twig and write wiki on this change proposal 08:50:03 using Twig would make it easier for us at Catalyst to maybe persuade some of our Drupal devs to care about Mahara ;) 08:50:07 ^ is that OK aarowlaptop? 08:50:17 rkabalin: Yes 08:50:20 thanks 08:50:32 http://umumble.com/blogs/php/smarty-vs-twig%3A-performance/ 08:52:27 from the article, "Smarty has worked in 60 times faster." 08:52:35 hmm, those results say smarty is faster 08:52:36 good to know 08:52:44 ok, we need to move, but we will get back to this discussion once we have more information. aarowlaptop feel free to add agenda item rather than 'item from last meeting' when you are ready, so we could discuss in details and vote 08:52:48 though the test is 3 years old 08:52:52 #info According to http://umumble.com/blogs/php/smarty-vs-twig%3A-performance/ Smarty is 60 times faster than Twig 08:52:55 ah 08:53:03 will do 08:53:15 yeah, it's old. 08:53:32 thanks aarowlaptop 08:53:34 but we should be able to do our own tests based on their ones 08:54:20 ok, next item 08:54:32 #topic Move to a CSS framework instead of keeping our custom responsiveness (@anitsirk) 08:54:47 ok. put your reading glasses on if you haven't already. ;-) 08:54:48 #info One standing item on our loose roadmap is to improve usability. Another one is to make it easier for designers to create a theme for Mahara. Currently, Mahara uses as custom implementation of responsive design that is becoming difficult to maintain and to expand. 08:54:56 #idea Therefore, we are looking into moving to use a CSS framework. 08:54:57 Bootstrap is one of the more famous ones, but there are plenty of others. We've talked with front-end developers at Catalyst to get an understanding of what they'd recommend. 08:55:08 #info We also considered the argument that Moodle uses Bootstrap and so should we to make it easier, see https://mahara.org/interaction/forum/topic.php?id=6553 08:55:08 What we need to keep in mind is that the framework should suit Mahara and its purpose and not solely choose it because a different project uses it. 08:55:08 So here are a few arguments for and against to two top contenders Bootstrap and Gumby. 08:55:19 http://getbootstrap.com/ 08:55:19 #info Pro Bootstrap: Large community, lots of people use it, lots of extensions 08:55:19 #info Con Boostrap: Moodle is only on 2.3 and thus a very different version from Bootstrap 3 (different enough to get confusing) 08:55:19 http://gumbyframework.com/ 08:55:36 #info Pro Gumby: Very flexible as there are several ways of doing things (CSS, HTML, semantic grid), seems to be easier to work with (but that might be subjective), easier to style for CSS straight away than Bootstrap -> can do more in the CSS layer, e.g. you can add more to the templates yourself 08:55:36 #info Con Gumby: smaller community than Bootstrap (but it's not a one-man project) 08:55:48 #info As for the argument that many people already know Bootstrap: Bootstrap components an be used in Gumby and vice versa as they are compatible to a degree. 08:55:49 #info Another alternative would be to use Base which is a leightweight CSS framework that doesn't come with its own JavaScript like Bootstrap and Gumby do. 08:55:49 http://matthewhartman.github.io/base/ 08:56:00 #info We should also figure out what we value in a framework that we choose: is it Flexibility (Gumby), community (Bootstrap) or lightweight (Base)? 08:56:00 I'd be interested to know whether there are any other arguments for or against a certain framework that could help in the decision making. We might have the opportunity to get started on this piece of work very soon and thus are looking for a decision sooner rather than later. We do know that there is not just the CSS framework that would need to be looked at but also the templating system as well as JavaScript. 08:56:00 #info We are also planning on giving our themes a refresh, but would only tackle this once we've implemented a CSS framework so that the work of creating the themes wouldn't have to be done twice. 08:56:07 .. 08:57:54 wow... 08:57:56 also if the css framework also includes js - will it be compatible with the current js in mahara 08:57:57 Why do we look at what Moodle is using? 08:58:33 robertl_ not with mochikit, but should be fine for most of jQuery as Bootstrap uses its own version of jQuery. 08:58:35 did you want to have the new framework for 15.04? 08:58:49 I'd like to pick a project with an establish community to decrease the likelihood of our library getting orphaned 08:59:08 lightweight is also good 08:59:12 or, ease of use at least 08:59:16 rkabalin_ the argument was that it might be easier for designers to theme mahara if the framework is the same as moodle. but from the front-end devs' opinion, it's a weak argument 08:59:28 ok 08:59:43 I personally prefer Bootstrap 08:59:55 and I have some experience using it 08:59:56 ghada: that depends entirely on how much time we can get from the front-end devs as it is most likely at least 1 month of work. 08:59:58 would those actually make it easier for theme developers or would it be another framework-language to learn besides css? 09:00:24 it is not a language tobiasz2 09:00:31 just a set of css rules 09:00:39 so we are thinking of starting now, but if it shapes up not to be finished for 15.04, we'll wait until 15.10. it's not just the implementation of the CSS framework, but also making changes to the templates and also creating new themes. 09:00:45 it is pretty easy and intuitive 09:01:14 tobiasz2: it would help with consistency and thus make it easier for themers 09:01:32 but how do they integrate JavaScript? Are that just be JavaScript sources that have to be included and that work well with the css-rules? 09:02:29 There is JS library that you can use to make some certail elements behave the way you like - e.g. menu dropdowns 09:02:34 i'm just thinking of the clients who paid to have a theme implemented. They'll have to redo it. Also, there was a big jump in themes from 1.7 to 1.8. So, many clients had to redo their theme then. We'll be asking them to redo it again - in a relatively short period . 09:02:35 it is well-documented 09:02:38 tobiasz2: bootstrap for example comes with prepackaged sets of components of jQuery 09:02:58 JS in bootstrap is not JS framework, it is pretty limited in fact 09:03:04 ok, thanks 09:03:11 it is just facilitating functions 09:04:05 can we maintain bootstrap and legacy themes at the same time? 09:04:06 ghada: good point. on the other hand, we do want to refresh mahara and entice people to start / continue using it. with our current responsive theme that does become limiting 09:04:19 I 09:04:28 but, if it's getting too difficult to maintain it, maybe we need to move forward with it... 09:04:35 I'd assume the existing themes would still work 09:04:48 as they would be css/js based anyway 09:04:51 good question robertl_. I was going to ask it. 09:04:54 rkabalin_ apparently you could, but that could make it quite confusing 09:05:21 they just wouldn't be as swish as the new bootstrap ones 09:05:21 robertl_ but the classes would change 09:05:22 @anitsirk well, for backward compaibility purposes 09:05:44 e.g. if someone willing to use the theme they developed with future versions 09:06:05 if we don't change the namings on html-level and introduce a new base-theme, the old themes should not be affected, right? 09:06:15 the idea for implementing a framework is to remove all existing CSS and start from scratch and only implement as much as needed so that there are no unnecessary left-over things. 09:06:51 but will that mean changing the template files or just the css files? 09:07:08 heh, what if someone paid for the theme and now we say - guys it is not compatible at all, it needs to be re-written completely 09:07:17 we should also change the template files 09:07:46 isn't that what moodle did from 1.9 to 2.x for example? 09:08:01 Well, it's already kind of a headache to upgrade a theme from one release of Mahara to another, right? 09:08:13 yes, there was a huge jump from 1.7 to 1.8, but it did not mean that the theme had to be redone entirely. it could be upgraded at less cost than creating a new theme. 09:08:14 yep, template files change is not avoidabe 09:08:22 One of the benefits of moving to a framework is that (in theory) it should make subsequent theme upgrades easier 09:08:43 ... and maybe new theme developments 09:08:59 there are always some adjustments to make from one version to another, but often they are reasonably small. only 1.7 to 1.8+ was bigger 09:09:01 for that matter, if we switch to a framework, it should make it easier to produce custom themes than currently 09:09:04 and going from 1.10 to 15.04 is the perfect time for upgrade pain 09:09:14 hah, good point 09:09:20 the large jump in numbers calls for it 09:09:32 new theme development should be easier for sure as there'll be more consistency, and we could also implement a living style library more easily 09:10:25 I think it is good idea to move to the framework, just need to make it not to disappoint people, may be announce it now that there will be a move to the new theme framework 09:10:45 a backwards-compatibility theme would in theory be possible... 09:10:57 just update the 1.10 "raw" theme to work in 15.04 09:11:17 rkabalin_ i would announce it at feature freeze / around that time once we know whether we can get things in for 15.04 or work towards 15.10 09:11:18 or possibly an automated upgrade the theme script 09:11:19 and then make your "old-school" themes inherit from bc-raw instead of raw 09:11:38 and keep consistent with html-classes like "js-hidden", which are somtimes set in the code if I remember correctly 09:12:36 yeah, it'll depend on how much of the theme change turns out to be strictly in themes themselves, and how much is in PHP 09:12:36 aarowlaptop, that would be good for intermedate version (that supports both bootstrap and legacy), and then legacy support can be removed in the future 09:13:21 any strong feelings towards bootstrap, gumby or something else besides rkabalin_? 09:13:37 the legacy theme sounds good 09:13:54 i agree with the legacy theme. 09:14:16 #idea see if it were possible to keep a legacy base theme of which existing themes could inherit so they don't have institutions don't have to create a new theme immediately 09:14:28 if you remove it right away without giving users breathing space, you may find that it's not taken up 09:15:16 #info ghada: if you remove it right away without giving users breathing space, you may find that it's not taken up 09:15:24 One thing that came up on the Mahara Barcamp: to have an external theme creator 09:16:11 like a clickable generator that produces a theme you can put into your mahar installation without having to have a greator knowledge of CSS 09:17:06 )) 09:17:07 tobiasz2: great idea! can you whip one up for us... ;-) 09:17:22 the theme generator sounds good - but who to build it :) 09:17:30 I'd wait until you decide on a framework ;) 09:17:35 heh 09:17:41 So it'd be a bit like the skin generator? ;) 09:18:00 Doesn't Mahara have a UI for customizing colors and stuff in the theme? 09:18:02 tobiasz2 good news with bootstrap this can be possible, because css rules are strict there 09:18:13 yeah, I guess. Maybe with a preview of one or two pages 09:18:17 aarowlaptop: yes for institutions to have a browsing theme 09:18:28 it is only for people from that institution though. it's not a full-fledged theme 09:18:35 and doesn't work under "No institution" 09:19:21 Do we need to decide now which framework we choose? 09:20:19 well, like we said before, we need to choose one that will be around for a while. 09:20:19 i'd like to see if there are any big objections to a particular framework or if we can leave it up to the front-end experts to hash it out with aarowlaptop 09:20:37 no objections from me 09:20:43 me too. 09:20:57 one thing to consider is how much double-up/overlap there will be with jquery/jquery-ui 09:21:36 will we end up with multiple things all doing the same job - like we have now with mochikit/jquery 09:21:46 robertl_ as far as I remember jquery is standard there, just extended 09:22:19 so, it does not break anything you may find in the standard jquery lib downloaded from jquery site 09:23:23 it's getting a bit late. should we leave it up to the front-end'ers? 09:23:23 so, it should be fine to use jquery that comes with bootstrap for anything 09:23:31 yep 09:23:59 #action leave CSS framework porposal up to the front-end experts to hash it out with aarowlaptop 09:25:29 but I assume dev people agree it will be a good move forward considering we keep legacy support for a while 09:25:53 yes, we really do need legacy support - that's my opinion. 09:26:10 yes, I agree 09:26:56 #info we need to try hard to keep legacy support while implementing CSS framework 09:27:05 ok, next item 09:27:14 #topic Message composition when replying to an internal_notification (tobiasz2) 09:27:19 ok, I quickly wanted to discuss the reply-mechanism for internal notifications after installing the multirecipient notification. The patch is here: https://reviews.mahara.org/#/c/3891/ 09:27:34 just a small recap: 09:27:38 The multirecipient notification uses different tables with their own ids. When replying to a multirecipient notification, the new notification is still linked to the parent notification. But the link doesn't work across the different notification systems. 09:27:58 Because of that, right now you don't see the internal notification in the message log above the notification when replying to an internal user notification. I didn't offer a "reply" link for internal notifications but a "send message to" link in the first patches. 09:28:15 When Robert and Kristina suggested to move the reply buttons to a separate column, I started displaying a reply-link to the internal notifications as not to confuse users with too many different icons. But Aaron noticed that people could hit "reply" for an internal notification without actually see the content of the notification they are replying to. 09:28:41 One thing I'd like to remind of is that this is a temporary issue, by the time, most messages people reply to will be multirecipient notifications and you'll have the complete dialog on the compose-screen. For the moment I have 4 ideas to tackle that (all are already in the comments on the patch): 09:28:51 - making the message content, recipient list and the list of reply-buttons open and close simultaneously, triggered by clicking on either the message subject, the extend-button behind the first recipient or into the reply-column 09:29:07 - introduce a third button "send message to sender" and display that for internal notifications, because the message thread is lost and it's not a full reply any more with link to the old notification 09:29:18 - pass the internal notification to sendmessage.php and display the old thread, hinting, that the message will start a new thread 09:29:29 - or migrate the old message thread when replying, so the thread is actually preserved. Though that might raise race conditions, when both participants reply to messages from the same thread simultaneously 09:29:41 ok, any thoughts? 09:29:42 .. 09:29:57 What do you consider an internal notification? User message or others? or do you mean the "old" notifications from before multi-recipient times? 09:30:00 I like option 1 09:30:27 yep, internal notification would be the old notifications from before the multirecipient stuff 09:32:30 i wouldn't introduce a third button. that screen is already very busy and trying to find an icon that is sufficiently different may not be easy. people would most likely have to read the tooltip anyway 09:32:34 what do you mean open and close simultaneously? 09:33:23 @ghada: right now you click on the header and only the message opens. And you enlarge the recipient list and only that opens. I'd make it so that you open both simultaneously 09:33:40 i see. 09:34:11 thinking of how thunderbird works - I click a message line and am given the message and a toolbar above that message 09:34:23 which to me seems a good idea in general, though it still lets you reply to the message without opening it at all. Though for the old internal notifications it's not really a reply anyway 09:34:27 i agree with anitsirk. we shouldn't introduce another button. 09:35:05 so yeah opening the message to see the reply to buttons make sense 09:35:30 it seems the simplest thing to do. 09:35:38 yeah, I also thought of the thunderbird menu, but that'd be a bigger change, opening a menu bar. Also, you'd have to move the mouse a lot to open a notification and then search the button to reply in the menu bar 09:36:55 ok, I'll make a patch for option 1. Thanks for the feedback :) 09:37:00 before the reply link was text an end of message - but if we have reply buttons at start of message it would be more useful 09:37:22 #idea to implement: making the message content, recipient list and the list of reply-buttons open and close simultaneously, triggered by clicking on either the message subject, the extend-button behind the first recipient or into the reply-column 09:37:54 I am still thinking ))) 09:38:09 but yes, option 1 os fine 09:38:13 robertl_ so move the buttons away from their new column that we introduced with the patch (still pending)? 09:38:14 robertl_ wouldn't you then though also see all the recipient names? 09:38:51 man this page is complicated 09:38:54