07:31:45 #startmeeting 07:31:45 Meeting started Wed Oct 5 07:31:45 2011 UTC. The chair is dobedobedoh. Information about MeetBot at http://wiki.debian.org/MeetBot. 07:31:45 Useful Commands: #action #agreed #help #info #idea #link #topic. 07:31:47 #topic Meeting attendees 07:31:49 Please put #info in front of your name 07:31:51 #info dobedobedoh is Andrew Nicols - LUNS Ltd, UK 07:32:03 #info anzeljg is Gregor An�elj 07:32:09 #info anitsirk is Kristina Hoeppner, Catalyst IT, NZ 07:32:11 #info richardm is Richard Mansfield 07:32:26 #info elky is Melissa Draper - Catalyst IT, NZ 07:32:30 #info pxh is Piers Harding, Catalyst 07:32:55 I think that's everyone at the moment 07:32:56 Reminder: Please put ".." on a line by itself at the end of a long multi-line speech 07:32:59 to indicate that you've finished 07:33:02 .. 07:33:04 #topic Items from previous meeting 07:33:07 #info dan_p LUNS to investigate adding a new mahara integration project to run selenium tests 07:33:25 I'm fairly sure that dan_p hasnt' had a chance to work on this one any more yet, so I'll action it again for the next meeting 07:33:28 #action dan_p LUNS to investigate adding a new mahara integration project to run selenium tests 07:33:49 Is there anything else to discuss arising from the previous meeting that I've forgotten about? 07:33:59 #info fmarier is Francois Marier 07:34:05 heh, sorry fmarier ;) 07:34:09 :) 07:34:55 dobedobedoh: doesn't look like it from last time's minutes 07:35:14 Excellent :) 07:35:19 #topic Where should we put unit tests? [fmarier and dobedobedoh] 07:35:34 Do you want to go first fmarier? 07:35:56 so i put this down because i thought that the test directory (top-level directory) would be a place for this 07:36:13 and because we don't need to include unit tests in the tarballs we give to people 07:36:27 (the test directory is excluded by the build script) 07:36:44 but you seem to think it would be better to put them somewhere else 07:36:44 .. 07:37:11 I think that it could be beneficial to allow for unit tests to be placed within each plugin type 07:37:16 So for example, to write unit tests for an artefact, we should place tests in /artefact//tests/phpunit 07:37:23 This has the benefit that plugins can come fully self-contained copmlete with unit tests 07:37:47 And if anyone wants to have a plugin included in core, we can require that it comes complete with unit tests 07:38:01 Other projects work along similar lines - e.g. moodle 07:38:11 we could require that no matter where the tests live though 07:38:15 We can easily modify the build script to ignore any 'tests' directory 07:38:44 true, but it would make it easier for anyone using a VCS to have all of the tests easily executable within one location 07:38:50 and for those who use git submodule for example 07:38:53 .. 07:39:39 i'm not opposed to it, but it does make the tarball building script and the test running scripts more complicated 07:39:39 dobedobedoh: so in your scheme, where should the core tests go? 07:39:52 i personally doubt that many third-party plugings will come with unit tests 07:40:04 but it would certainly be nice 07:40:21 For libraries in htdocs/lib/tests; and for anything in the top level directory, /htdocs/tests 07:40:36 I'm not overly concerned as to where they go ultimately 07:40:50 But I do think we should make it easy for developers to write plugins with unit tests 07:41:39 .. 07:41:40 I think I'm with dobedobedoh on this. 07:42:01 I'm not too worried about where they go 07:42:08 dobedobedoh: do you want to look into writing the test running script and modifying the release script to look into the right places? 07:42:16 yeah I can do :) 07:42:23 in that case, it's all yours I think :) 07:42:33 #action dobedobedoh to look at modifying the release script, and test running script 07:42:37 heh okay 07:43:03 There are still a few minor issues with the unit tests 07:43:25 At present if you try and build test coverage it gets a trifle unhappy as it tries to test every UI file as well as libs 07:43:31 and complains about files being re-declared 07:43:53 But I think we should be able to get most of core tested now :) 07:44:00 .. 07:44:09 Is this what was crying with the mysql stuff? 07:44:32 Nope... that was ADODB storing the password in $password as 'not stored' 07:44:45 ... lolwat 07:44:46 Which is fine, until you __sleep and __wakup 07:45:02 which mysql likes to do i hear. 07:45:10 As a 'security measure', when you Connect(), after connecting it stores the password as the string literal 'not stored' 07:45:17 Oh sorry... that's the other bug 07:45:38 That was caused by MySQL losing it's compatibility mode 07:45:44 It wasn't being re-applied on the __wakup 07:45:45 .. 07:45:58 it's a really good thing you caught that 07:46:11 It would have caused some very interesting issues 07:46:47 I've outlined my intentions WRT to the testing script at https://wiki.mahara.org/index.php/Talk:Developer_Area/Unit_Testing 07:47:06 I'd like to test any database upgrades that we do 07:47:14 as well as run the full unit tests on a fresh install 07:47:20 .. 07:47:50 you're going to *love* the timezone change then when we get to them 07:47:57 * fmarier should start using the cli installer 07:48:01 heh 07:48:17 Is there anything else to dicuss on unit testing before I move on? 07:48:47 I'm looking forward to the TZ changes being applied to mahara.org - that's for certain! 07:49:21 yeah me too! 07:49:41 "what do you mean it's 11pm? Oh right, that silly island thinks its time is important" :P 07:50:15 :p 07:50:22 elky: actually, i think it's texas time that's displayed on mahara.org because that's where the server is. ;-) 07:50:56 anitsirk, ah, my mistake. still :P 07:51:07 yeah, it was fine when it was NZ time :P 07:51:22 It's mightily confusing! 07:51:29 should just make it UTC and confuse everyone 07:51:54 anyways, we should probably move to the next agenda item, i think we're done talking about unit tests 07:52:01 yup 07:52:08 Since zzmonty isn't here yet, I'll jump to pxh's - you ready for it pxh? 07:52:36 #topic Web Services - an update, and request for API ideas [pxh] 07:52:40 yup 07:52:44 The floor is yours 07:53:07 ok - this was just an update on where things have got to with the artefact/webservice plugin 07:53:19 Web Services (https://wiki.mahara.org/index.php/Plugins/Artefact/WebServices) now has OAuth support including OOB. 07:53:19 I've also added personal user tokens - a variation on the wstoken as implmented by Moodle - where users can manage their own tokens for specific API calls that have been allocated for general use. 07:53:19 These general use API calls would typically be ones that only access the users own data eg: mahara_user_get_my_user. I'm hoping that having features like this will open up interesting scripting opportunities on systems like myPortfolio. 07:53:37 #info Web Services now has OAuth support including OOB 07:53:43 #link https://wiki.mahara.org/index.php/Plugins/Artefact/WebServices 07:54:10 #info Personal user tokens are supported and users can manage their own tokens for specific API calls 07:54:57 #info This should open things up more for scripting opportunities 07:55:07 #save 07:55:08 sounds great pxh 07:55:12 I've written quite a lot of unit tests 07:55:27 * dobedobedoh grins 07:55:50 the existing APIs are mostly centred around the ministry of Educations (NZ) requirements for myPortfolio 07:56:21 so it things like user CRUD, groups, and institution membership 07:56:30 ... not very user centric 07:56:48 #info Existing APIs mostly centred around myPortfolio 07:56:54 so - I was hoping to branch this out to APIs that are of a more generic usefulness 07:57:00 Sounds really handy - I've got a use for that at present 07:57:09 Alan McNatty is looking at using it for the next maharadroid release 07:57:29 yup 07:57:32 He's pretty keen to get the webservice plugin in core 07:58:22 there is work to do around that 07:58:39 yeah, i hope to have some free time soonish to help out 07:59:27 So at present, the API allows CRUD of groups, and institution membership (and presumably CRUD of users too) 07:59:41 As artefact/webservice is based on the Moodle webservice code base it needs going over - I've spent a lot of time going through it and cleaning out superfluous stuff, but there is still quite a lot to do - I think 07:59:42 But doesn't allow for users to do anything with the API 07:59:55 sorry to ask, pxh, but what does CRUD stand for? 07:59:59 dobedobedoh: yes 08:00:07 anitsirk: create, read, update and delete 08:00:13 thanks 08:00:27 anitsirk +1 - i didn't know that too 08:01:47 so - I'd like to field suggestions for APIs - or atleast have a place for peoples ideas on this front 08:01:49 .. 08:02:34 do you want suggestions here/now; or to the wiki page? 08:03:36 I don't want to hold people up - people can catch me later - and I can organise them onto a wiki page 08:03:53 but if the creative juices are flowing then don't let me stop you 08:04:54 I guess one thing which might be helpful is the ability to copy pages 08:05:12 both from a user to become a template, and from user to user (e.g. administrator to all members of a group) 08:05:25 But that's probably more a web interface request than API now I think about it... 08:06:27 ok - noted 08:07:06 Thanks pxh :) The WS API looks good 08:07:57 Unless there are any objections, I'll move on to the next topic then... 08:08:58 .. 08:09:02 #topic Wiki - Adding extensions to make adding/editing pages easier [zzmonty] 08:09:24 I don't think that zzmonty is here, does anyone want to discuss this in her absence? 08:09:33 (I've seperated these into two topics btw) 08:10:07 we should probably do it next time if she's around 08:10:14 okay 08:10:26 I'll park it straight onto the next meeting wiki page 08:10:29 #topic Wiki - Request to allow additional extensions (.zip) to make it easier for user contributed content [zzmonty] 08:10:38 I guess the same is true for this one fmarier? 08:10:49 i've replied to that one on the forum already 08:11:03 #info dobedobedoh to add zzmonty's items to the agenda for the next meeting 08:11:11 basically zip files should be on mahara.org where we can control quotas 08:11:23 I'm not sure whether this item was added pre or post your comment on the forum 08:11:28 on the wiki, if we allow zip files, then people have unlimited storage for any kind of content at all 08:11:48 we can't afford to be a public dropbox :) 08:12:11 (on the old wiki, i found lots of spam pages and also personal files from random people) 08:12:34 I don't understand people. 08:12:34 .. 08:12:45 Okay, well I'll park these for the next meeting then 08:13:25 Which brings us on to the next topic 08:13:30 #topic Next meeting time/date and Chair 08:13:48 and i've got something else for the "any other business" part too 08:13:52 Next month's meeting will be an evening UTC meeting 08:14:07 okay... AOB will be next ;) I accidentally reversed the order 08:14:28 Anyone have any suggestions for next month? 08:14:30 Tuesday 1st November @ 19:30 UTC (19:30 GMT/08:30 NZST) 08:14:35 #idea Tuesday 1st November @ 19:30 UTC (19:30 GMT/08:30 NZST) 08:15:05 what would it look like the week after? 08:15:24 because the week after would be after daylight saving kicks in 08:15:27 8th November? 08:15:29 Ah okay 08:15:34 the week after or even tuesday evening GMT would be better 08:16:21 isn't daylight savings the last sunday in october? 08:16:24 * anitsirk checks 08:16:29 So that'd be Tuesday 8th November @ 18:30 UTC (19:30 GMT/08:30 NZST) ? 08:17:08 yep. daylight savings ends on 30 october 2011 for europe 08:17:12 dobedobedoh: no 08:17:16 http://www.timeanddate.com/worldclock/fixedtime.html?iso=20111108T1930 08:17:23 UTC doesn't do daylight savings :) 08:17:38 Which bits wrong? 08:18:38 * dobedobedoh spots it 08:19:02 19:30 UTC is: 19:30 London time (Tue) and 08:30 NZ time (Wed) 08:19:06 #idea Tuesday 8th November @ 19:30 UTC (19:30 GMT/08:30 9th NZST) ? 08:19:36 it's better to never print out more than one date 08:19:46 that way you can never have conflicting ones :) 08:20:08 * elky mutters something about timezones. 08:20:17 * fmarier once missed a Mahara meeting because of wrong conversions in the minutes of the previous meeting 08:20:26 time is fine for me 08:20:33 works for me 08:20:35 #agreed 08:20:43 #agreed 08:20:51 Any volunteers for chair? 08:21:48 #info Next developer meeting to be held on Tuesday 8th November @ 19:30 UTC - http://www.timeanddate.com/worldclock/fixedtime.html?iso=20111108T1930 08:22:38 i'm guessing we've all done it once? 08:22:47 If no-one comes forward soon, I'll pick someone unsuspecting... 08:22:52 yeah I think so 08:23:36 i'll do it then 08:23:37 don't think so 08:24:03 #info Next meeting to be chaired by fmarier 08:24:06 erm, no i haven't done it yet 08:24:13 i'll do it, sorry 08:24:18 ooh okay :) 08:24:19 #undo 08:24:19 Removing item from minutes: 08:24:26 #info Next meeting to be chaired by richardm 08:24:27 ... 08:24:43 have you done it elky? 08:24:58 #topic Any other business 08:25:01 richardm: https://wiki.mahara.org/index.php/Developer_Area/Developer_Meetings/Chair_Duties 08:25:04 dobedobedoh, not yet. 5 minutes after i get out of bed isn't the best idea either ;) 08:25:07 (in case you haven't seen it) 08:25:14 nex time perhaps 08:25:21 as in, the one after richardm 08:25:22 fmarier: thanks 08:25:25 heh indeed. I did suggest a while back that we chair in our own evenings 08:25:27 then the december meeting may be yours. that'll be in the evening ;-) 08:25:40 sounds good 08:25:47 I've got a couple of questions 08:25:47 fmarier - you had an AOB? 08:25:58 #info anzeljg wants to let everybody know that his development Mahara install at http://mahara.ledina.org/DEV/ no longer exists 08:26:00 so the point i wanted to bring up is bug numbers in commits 08:26:40 anzeljg: You oky to come after fmarier? 08:26:48 basically, i think it's a good thing and we've been pretty good at including them, but at the same time, i think that it shouldn't be a strict requirement 08:26:50 yes, sorry to interrupt 08:27:33 fmarier, should we be committing things that don't exist in the bugtracker? 08:27:37 because it might discourage people from code cleanups and small trivial fixes if they have to take the pain of filing a bug report for them 08:27:52 trivial fixes such as small typos. 08:28:16 there are clear-cut cases like: you're adding a new feature, or you're fixing a bug that has been nagging a lot of users 08:28:32 these ones should obviously have bugs so that they're more visible to users and admins 08:28:58 so i think that it should be a discretionary thing on the part of the reviewers 08:29:02 yeah i think we already decided they shouldn't be a strict requirement a couple of meetings ago 08:29:16 yeah - I'm inclined to agree 08:29:18 if a reviewer thinks that it would be useful, then it probably should have one 08:29:19 fmarier, should we ask them to use a certain string so we know not to try marry it to the bugtracker? 08:29:43 but i've seen some reviews where it's like "no bug link, -1" 08:29:58 always creating a bug for fixing one typo in a lang string doesn't merit a bug report. i did a couple of times, but it seems like too much hoopla for a couple of letter changes. so fmarier suggested i can clean up lang strings without creating a bug for each which would be awesome and less disruptive on the tracker. 08:29:58 .. 08:30:12 ok, which just means we need to decide our expectations and document it. 08:30:34 ok 08:31:13 it's already documented in the "how to review" wiki page: https://wiki.mahara.org/index.php/Developer_Area/How_to_Review_Code#Commit_message 08:31:36 Is that everything you wanted to raise fmarier? 08:32:06 if we want to use the bug numbers to keep track of changes for bugs, then having typo fix commits have, say (typo fix) instead of (bug #xxxxxx) would help avoid the "no link, -1" 08:32:09 i've added two examples to that wiki page 08:32:30 elky: i don't think it needs to be computer-readable 08:32:38 * iarenaza says hi 08:32:43 hi iarenaza 08:32:45 if the commit is obvious, then the reviewer should pick up on that 08:32:54 hi iarenaza 08:33:05 reviewing requires good judgment 08:33:34 +1 08:33:52 oops - sorry 08:33:54 fmarier, of course, but spending half an hour trying to find a bug it relates to when launchpad goes the speed of a snail is also not idea. 08:33:58 ideal* 08:34:32 it's just a thought 08:34:34 well if the commit is trivial, then you don't have to search for a bug 08:34:40 or create a new one 08:34:58 so i think we're saying the same thing 08:35:33 I think we can probably manage without bug numbers and use our judgement if it's a trivial change 08:36:32 I think it can't hurt to encourage submitters to identify if it's related to a bug or not 08:38:07 I think people generally will 08:38:24 I'm going got suggest that we move on to anzeljg's AOB now as this topic seems to be circling 08:38:40 ok 08:39:16 anzeljg: The floor is yours 08:39:46 what is the answer about pChart? 08:39:57 In the meantime, I'm implementing eZ Components (Graph Component) insted of pChart 08:40:49 anzeljg: so far, i haven't received an answer from the author to my questions 08:41:08 so we should probably assume for now that it's not open source 08:41:13 and that we can't use it 08:41:14 #info fmarier hasn't heard back from the author of pChart about licensing conflicts 08:41:35 #info anzeljg implementing eZ Components instead 08:41:44 ok 08:42:06 and my other question - is there something new on skin support integration? 08:43:17 richardm: Do you have any info on this? 08:43:23 i haven't had time to look at it yet 08:43:34 * iarenaza Looking at http://sourceforge.net/projects/pchart/ it says it's GPL-ed 08:44:30 no, i've been too busy as well 08:44:31 iarenaza: http://www.pchart.net/license states that it's not for commercial application without buying a license which != GPL 08:44:49 but if the author doesn't get back isn't that also a sign that the project may not be very active? 08:45:15 anitsirk: the author got back to me on my initial query, but then i haven't heard back when i sent him a few questions 08:45:28 so i think the project is active 08:45:40 #info No update on skins support... ongoing 08:45:48 i would just want to clarify whether or not it's under the GPL 08:45:55 from what's on the website, it's not 08:46:40 anyways, i'll update the bug if/when i hear back 08:46:56 Is there anything else to discuss? 08:47:05 not from me 08:47:05 fmarier: the text from dobedobedoh's link is rather ambiguous, so a definitive answer from the author is really a good idea :) 08:47:25 not from me either (anything else to discuss) 08:47:25 just an idea: maybe we can also use eZ Components (Documnet 08:48:01 just an idea: maybe we can also use eZ Components (Document component) for PDF export 08:48:05 #link https://bugs.launchpad.net/mahara/+bug/721496 08:48:08 Launchpad bug 721496 in mahara "Using pChart as framework for rendering graphs in Mahara" [Wishlist,Won't fix] 08:48:14 that's the bug to subscribe to if you're interested 08:48:17 IIRC lamiette was working on PDF export and it's nearly complete 08:48:27 ok 08:48:43 Perhaps she can comment on that idea when she's available 08:48:51 fine with me 08:49:10 anzeljg: do you know if it requires having xwindows installed? 08:49:28 because that's one of the downsides of the package that lamiette used in her pdf export code 08:49:31 what? eZ Components? 08:49:44 many servers don't have X installed 08:50:01 anzeljg: yeah the pdf generation in eZ Components 08:50:03 http://ezcomponents.org/docs/tutorials/Document 08:50:13 will look... 08:51:05 it seems to be a wrapper for TCPDF or Haru 08:51:30 haven't find it, but see http://ezcomponents.org/docs/api/trunk/introduction_Document.html#writing-pdf 08:52:41 it requires haru - http://ezcomponents.org/introduction/requirements 08:52:59 Doesnt look like it, fmarier. at least, i can't find mention with a search of the site 08:53:56 Is there anything else to dicuss re ezcomponents/pChart ? 08:54:02 Document component: iconv and haru, Graph component: iconv, xml and dom 08:54:08 no 08:55:10 hello (apologies for lateness) 08:55:17 hi lamiette 08:55:21 hi 08:55:25 hi 08:55:28 I guess we need to confirm whether X is required for ezcomponents? Would someone like to take that as an action for the nex meeting? 08:55:29 hi lamiette 08:55:32 #help 08:55:38 Hi lamiette 08:56:09 fmarier: Btw, pchart 1.x (from pchart.sourceforge.net) is clearly GPL (the headers in the source code say so). It's 2.x (from pchart.net) that is "dual licensed". 08:56:15 PDF export was/is working last time I looked at it, but I wanted to fix some issues with general exporting of collections vs single pages 08:56:24 I opened a tracker item for it 08:56:41 so the whole thing still needs some work, and may be a bit dated now already 08:57:28 I've not had much time to revisit it with other work but am hoping to get some time in a few weeks 08:58:53 Would someone be able to look at ezcomponets re: xwin then? 08:59:54 I'm going to try and wrap the meeting up now 09:00:27 Are there any new AOB to discuss? 09:00:47 not from me 09:01:05 not from me 09:01:07 no 09:01:12 Well thank you all :) 09:01:16 thanks dobedobedoh 09:01:20 #endmeeting