| *** th1a has quit IRC | 00:00 | |
| *** replaceafill has quit IRC | 01:17 | |
| *** menesis has quit IRC | 01:28 | |
| *** menesis has joined #schooltool | 01:36 | |
| *** menesis has quit IRC | 03:08 | |
| *** khildin has joined #schooltool | 09:01 | |
| *** th1a has joined #schooltool | 12:37 | |
| *** menesis has joined #schooltool | 12:41 | |
| *** replaceafill has joined #schooltool | 15:03 | |
| *** menesis has quit IRC | 15:17 | |
| th1a | hi replaceafill, yvl. | 15:30 |
|---|---|---|
| replaceafill | good afternoon | 15:30 |
| *** menesis has joined #schooltool | 15:31 | |
| yvl | good afternoon | 15:31 |
| yvl | and good morning replaceafill :) | 15:32 |
| replaceafill | :D | 15:32 |
| th1a | We're making replaceafill get up early. | 15:32 |
| th1a | OK so... Hopefully Glenda won't find a new bug today. | 15:33 |
| th1a | Perhaps we should debrief this process a bit. | 15:33 |
| th1a | Since I think we can all agree that it went pretty badly. | 15:33 |
| th1a | I would say one thing that made a difference compared to previous years, as much as I hate to say aelkner was right, was he always insisted on having the CanDo-ers working from a sandbox when they were testing. | 15:34 |
| th1a | And of course he was in the same time zone. | 15:35 |
| th1a | The trunk packages are great in general, but we ended up in a situation where Glenda was giving us one bug at a time which we were, hopefully fixing overnight, and then getting a little feedback, finding another bug, etc. | 15:36 |
| th1a | The turnaround was just too slow for all these small issues. | 15:36 |
| replaceafill | imho, the fundamental issue was us not paying attention to testing anymore | 15:37 |
| th1a | I'd say particularly if we have PAYING CUSTOMERS in the future, we'll need to be proactive about having times scheduled where we can quickly fix things. | 15:37 |
| th1a | replaceafill, yes, that too. | 15:37 |
| th1a | Definitely. | 15:37 |
| th1a | In particular the permissions change would have always been a nightmare testing-wise, right? | 15:38 |
| replaceafill | tests warned us about all this forbidden exceptions left and right | 15:38 |
| th1a | Well... I don't know. | 15:38 |
| th1a | Did we write off real bugs as just testing cruft? | 15:38 |
| replaceafill | th1a, what do you mean? | 15:39 |
| th1a | Well, when you change the permissions, by definition you're going to throw a bunch of bugs which mostly mean you have to change the tests. | 15:39 |
| th1a | They aren't really bugs, just obsolete tests. | 15:40 |
| replaceafill | well, you need to extend them, right? | 15:40 |
| replaceafill | changing permissions means modifying/adding functionality didn't exist | 15:40 |
| th1a | Yes. | 15:41 |
| th1a | To do it right would be a LOT of tests though, right? | 15:41 |
| replaceafill | indeed | 15:42 |
| th1a | Like four tests per object/action. | 15:42 |
| replaceafill | per role ;) | 15:42 |
| th1a | Right. | 15:42 |
| th1a | What would be the most efficient way to do that? | 15:42 |
| replaceafill | but we'd have a stronger app | 15:43 |
| replaceafill | no clear idea from me, on most efficient way | 15:43 |
| replaceafill | just do it! | 15:44 |
| replaceafill | :D | 15:44 |
| replaceafill | at least for new functionality | 15:44 |
| replaceafill | comboboxes, name change, etc | 15:44 |
| th1a | Well, it might be sufficient to use something selenium-based. | 15:44 |
| replaceafill | imho we already have a good infrastructure in place | 15:45 |
| replaceafill | we just have to use it | 15:45 |
| replaceafill | i really enjoy using it, and keep saying "well done Justas" | 15:45 |
| th1a | Should these be unit tests? | 15:45 |
| menesis | unit tests are much faster to run | 15:46 |
| menesis | and testing methods of the views can be done | 15:46 |
| menesis | and there are many tests doing that | 15:46 |
| menesis | selenium tests are also needed, for the javascript, ajax stuff and integration in general | 15:47 |
| replaceafill | +1 | 15:47 |
| menesis | but they are slow... we need some kind of fixtures, a basic school setup | 15:48 |
| replaceafill | we need redundancy :D | 15:48 |
| menesis | so that each selenium script does not need to create users/years/terms/timetables | 15:48 |
| replaceafill | i created a dumb script that uses the screen command | 15:48 |
| replaceafill | to run stests "in parallel" | 15:49 |
| replaceafill | pretty dumb, but it saves me around 20 minutes when running the cando test suite | 15:49 |
| replaceafill | it'll burn my cpu at some point but... :D | 15:50 |
| th1a | I'm happy with an external fan running under my laptop replaceafill. | 15:50 |
| menesis | "bin/test -j 3" is supposed to run tests in parallel | 15:50 |
| replaceafill | does it work? | 15:51 |
| menesis | but I haven't tried that | 15:51 |
| replaceafill | i remember trying it once and didn't | 15:51 |
| th1a | OK, so assuming we're focusing on permissions tests, how would we systematically throw these in for the additional roles? | 15:51 |
| th1a | Assuming we'd prefer to use unit tests. | 15:51 |
| menesis | no, unit tests won't help here | 15:52 |
| replaceafill | right | 15:52 |
| menesis | because problems arise like user can access section info page but not the name of the teacher | 15:52 |
| replaceafill | i can't imagine how to test a viewlet security for different roles using a unit test | 15:52 |
| menesis | so whole page is permission denied | 15:53 |
| th1a | OK, fine. | 15:53 |
| th1a | So is this something to jam into the existing stests or is this a little side project? | 15:55 |
| replaceafill | in my head, it's like what we did a couple of years ago for the gradebook | 15:55 |
| replaceafill | in the old skin | 15:55 |
| replaceafill | a little side project | 15:55 |
| replaceafill | for "extending coverage" | 15:56 |
| yvl | +1 | 15:56 |
| th1a | It seems like you set up a basic school, create users that are in one of the key groups, and just blast through the basic objects. | 15:56 |
| replaceafill | and in this case, for actually fixing the tests! | 15:56 |
| replaceafill | :D | 15:56 |
| th1a | OK. We just can't do it *right now*. | 15:56 |
| replaceafill | and i've always thought we need a check list | 15:56 |
| replaceafill | a test inventory | 15:56 |
| replaceafill | or something | 15:56 |
| replaceafill | item 31: test that clerk can access per skills report | 15:57 |
| menesis | a side project. a few new stests to check that a clerk can do what he is supposed to, including turning access options on/off | 15:57 |
| menesis | or that an administrator only sees read-only views | 15:57 |
| th1a | OK. | 15:58 |
| th1a | We can let replaceafill or menesis do that after the release. | 15:58 |
| replaceafill | +1 | 15:58 |
| yvl | +1 | 15:58 |
| menesis | also we need tests for student views because they have been broken occasionally and no one noticed | 15:59 |
| th1a | I would point out though that a lot of what tripped up Glenda wouldn't have been found by tests. | 15:59 |
| th1a | i.e., the requirement to distinguish between two identical skillsets from different years. | 15:59 |
| replaceafill | that was new functionality needed imho, not a bug | 16:00 |
| th1a | Also there was a lot of Glenda and Dave feeling like things were difficult and spending time trying to figure out things that were just bugged. | 16:00 |
| replaceafill | well... adjusting cando to their weird use case :D | 16:00 |
| th1a | i.e. that seems difficult because IT DOESN'T WORK. | 16:01 |
| menesis | yeah, sure, one can't think of every possible scenario. then you write/extend a test to check that particular case | 16:01 |
| th1a | If it worked it would be much easier. ;-) | 16:01 |
| th1a | That's where we suffered from not having real time communication and quick fixes. | 16:01 |
| th1a | Dave was also extra-heated over the duplicate skills thing because he felt it was only a problem because we'd blown off his recommendation to put skills in years. | 16:02 |
| th1a | Glenda has actually been pretty understanding about the whole thing. | 16:03 |
| th1a | She did a great QA job too... it just unfolded painfully incrementally. | 16:03 |
| replaceafill | hopefully, we'll hear less from them during the next year, right? | 16:04 |
| replaceafill | it's not that i want to get rid of them | 16:04 |
| th1a | Hopefully everything will work perfectly after today. ;-) | 16:04 |
| replaceafill | it's just that she did a lot of testing! | 16:05 |
| th1a | She also started a bit late. | 16:05 |
| th1a | But I do think it is a good warning for how to avoid problems with *paying* customers. | 16:05 |
| replaceafill | true | 16:05 |
| th1a | OK, any other last words on this? | 16:06 |
| th1a | It is good that all the bugs were pretty small, and we did have good turnaround fixing them. | 16:07 |
| th1a | OK, yvl? | 16:08 |
| yvl | ok | 16:09 |
| yvl | I was not able to reproduce timetabling bugs | 16:09 |
| yvl | the second one looks like a duplicate | 16:09 |
| yvl | also there was a remake in that area of journal | 16:10 |
| yvl | what else... | 16:10 |
| yvl | added displaying of contact info if any is entered | 16:10 |
| yvl | and looked at various bits and pieces | 16:11 |
| *** khildin has quit IRC | 16:11 | |
| yvl | also, I'll commit extended traceback views today | 16:11 |
| yvl | with plugins and their versions enumerated | 16:11 |
| yvl | and some info like python version, OS, etc. | 16:12 |
| th1a | Ah, good. | 16:12 |
| yvl | bug reports miss that info most often | 16:13 |
| yvl | "so what version of schooltool are you running" | 16:13 |
| yvl | also, I'm adding a "Submit to Launchpad button" | 16:13 |
| th1a | Nice. | 16:14 |
| yvl | that will open proper project to report the bug | 16:14 |
| yvl | and pre-fill tags with "auto traceback" | 16:14 |
| yvl | and summary with the traceback | 16:14 |
| yvl | sorry, additional comments with traceback | 16:14 |
| yvl | it also makes user enter summary and search for similar bugs before submitting | 16:15 |
| yvl | but traceback is not lost | 16:15 |
| yvl | and... that's it. | 16:15 |
| * yvl done | 16:15 | |
| th1a | OK. Yes, I don't see any other bugs that are low-enough hanging fruit for yvl to fix now. | 16:16 |
| th1a | So you can get back to the relationship status thread when you're ready. | 16:16 |
| yvl | kk | 16:17 |
| th1a | Thanks yvl. | 16:18 |
| th1a | replaceafill? | 16:18 |
| replaceafill | ok | 16:19 |
| replaceafill | http://dev.schooltool.org:7081 | 16:19 |
| replaceafill | manager:schooltool | 16:20 |
| replaceafill | Server -> Settings -> Name Sorting | 16:20 |
| replaceafill | need some wording there th1a | 16:20 |
| replaceafill | if you change the "Field", you'll see the different tables react accordingly | 16:21 |
| replaceafill | well, most tables ;) | 16:21 |
| th1a | I'd say "Sort by:" | 16:21 |
| replaceafill | ah ok | 16:22 |
| th1a | Maybe just that. | 16:22 |
| th1a | I don't know that the hint helps. | 16:22 |
| replaceafill | kk | 16:22 |
| replaceafill | i just wanted to "warn" the user | 16:22 |
| replaceafill | columns may swap | 16:23 |
| replaceafill | i'll get rid of it | 16:23 |
| replaceafill | then you can go to any section | 16:23 |
| replaceafill | http://dev.schooltool.org:7081/schoolyears/2012-2013/SY/sections/6320-15-4-5 | 16:23 |
| replaceafill | for example | 16:23 |
| replaceafill | click the Gradebook link there | 16:24 |
| replaceafill | or the CanDo Gradebook | 16:24 |
| replaceafill | and the columns should also react | 16:24 |
| replaceafill | i have a question here th1a | 16:24 |
| th1a | I think trying to explain it would probably lead to more confusion than just doing it. | 16:24 |
| replaceafill | oh wait | 16:24 |
| replaceafill | the section roster | 16:24 |
| replaceafill | it currently uses a full name column + username | 16:25 |
| replaceafill | should we split that accordingly? | 16:25 |
| replaceafill | id cards use the sorting now | 16:25 |
| th1a | The report? | 16:26 |
| replaceafill | ys | 16:26 |
| replaceafill | yes | 16:26 |
| replaceafill | http://dev.schooltool.org:7081/schooltool.task.messages/ReportProgressMessage/download/section_roster_6320_13-09-18-03-15.pdf | 16:27 |
| replaceafill | yvl, menesis here's my diff so far: | 16:28 |
| replaceafill | http://bazaar.launchpad.net/~replaceafill/schooltool/name_change/revision/3669?remember=3664&compare_revid=3664 | 16:28 |
| th1a | I don't feel like it is necessary or desirable to change the layout of the report but you might want it to respect the sort preference. | 16:28 |
| th1a | Or, you WOULD want it to respect the sort preference. | 16:29 |
| replaceafill | th1a, it does respect the reference | 16:29 |
| th1a | OK. | 16:29 |
| th1a | It is fine then. | 16:29 |
| replaceafill | but again, it's unclear when you use the title | 16:29 |
| replaceafill | title of the person, i mean | 16:29 |
| replaceafill | ok | 16:29 |
| replaceafill | on the gradebook | 16:29 |
| th1a | Will this be in the trunk PPA soon replaceafill? | 16:29 |
| th1a | If you are looking at a printed roster you probably want it to be readable. | 16:30 |
| replaceafill | eehh i still need to work on the journal, virginia and intervention | 16:30 |
| th1a | OK. | 16:30 |
| replaceafill | so, the sorting question of the gradebook | 16:31 |
| replaceafill | do we need the "Sort by" options in the two columns | 16:31 |
| replaceafill | ? | 16:31 |
| replaceafill | the original Name column had that option | 16:31 |
| replaceafill | that allowed you to sort ascending/descending | 16:31 |
| replaceafill | on the column | 16:31 |
| replaceafill | if we want to keep it, i need to update that functionality | 16:32 |
| th1a | Hm... | 16:32 |
| replaceafill | imho, just respecting the setting would be enough.... | 16:32 |
| th1a | Let's say no and see if anyone complains. | 16:32 |
| replaceafill | ok | 16:33 |
| replaceafill | and i also think we need to make the columns a little "wider"? | 16:33 |
| replaceafill | this css is using the original widths | 16:34 |
| replaceafill | around 180px for both columns | 16:34 |
| th1a | OK. | 16:34 |
| replaceafill | but the activity/points part complicates it a little bit | 16:34 |
| replaceafill | since you need more room for the second column | 16:34 |
| replaceafill | in this case i used a long first name and last name so you could see how the overflow is hidden | 16:35 |
| replaceafill | ok, cando gradebook is mostly done | 16:35 |
| replaceafill | still need to check the zillion reports :P | 16:35 |
| replaceafill | but since they're based on the persontable, it shouldn't take long | 16:36 |
| th1a | OK. | 16:36 |
| replaceafill | so i'll work on that | 16:36 |
| replaceafill | ah | 16:36 |
| replaceafill | jelkner requested an emergency fix for quiz | 16:37 |
| replaceafill | which i did last night | 16:37 |
| replaceafill | he made a mistake | 16:37 |
| replaceafill | because he's now a site manager and a teacher in his instance... | 16:37 |
| replaceafill | he created and deployed a course worksheet by accident | 16:37 |
| replaceafill | because he hadn't seen the option before :/ | 16:38 |
| replaceafill | users... | 16:38 |
| replaceafill | anyway | 16:38 |
| * replaceafill done | 16:38 | |
| th1a | Ah. | 16:38 |
| th1a | OK. | 16:38 |
| th1a | Thanks replaceafill. | 16:38 |
| th1a | menesis? | 16:38 |
| menesis | I got a cold over the weeked, so was working from the bed | 16:39 |
| menesis | still, I got something done | 16:39 |
| menesis | most of the time was spent tweaking the schooltool init script | 16:39 |
| menesis | so that it starts on a fresh install | 16:39 |
| menesis | and restart works | 16:40 |
| th1a | Thanks for pushing the little fix last night. | 16:40 |
| menesis | I ended up adding a "sleep 5" before starting | 16:40 |
| th1a | I figured it needed something like that. ;-) | 16:40 |
| menesis | to let the supervisor die | 16:40 |
| menesis | and I think it finally works well | 16:41 |
| menesis | part of this was testing and debugging on a virtual precise box | 16:41 |
| menesis | but turns out the failure to start celery was local | 16:42 |
| menesis | and solved by a reboot | 16:42 |
| menesis | so I pushed 2.5 packages to precise dev ppa | 16:42 |
| menesis | ... | 16:42 |
| menesis | without a warning notice and no announcement that this has been done | 16:43 |
| menesis | should I do that? | 16:43 |
| th1a | Uh... | 16:43 |
| th1a | I suppose we should! | 16:43 |
| th1a | It is the *dev* PPA, so I'm not sure what the procedure should be. | 16:44 |
| menesis | the release notes are updated and describe the new features and architecture | 16:44 |
| th1a | OK, I'll post a notice. | 16:44 |
| th1a | I'll handle it. | 16:45 |
| menesis | but the conffiles have changed so the root might need to take manual action to resolve conflicts | 16:45 |
| menesis | the advise is to anser Yes (install version provided by package) | 16:46 |
| menesis | what else | 16:46 |
| th1a | Oh... Glenda needs to know that? | 16:46 |
| th1a | I guess that came up before. | 16:46 |
| menesis | those questions arised when upgrading to trunk | 16:47 |
| menesis | ppa | 16:47 |
| th1a | Yes. | 16:47 |
| menesis | and several times | 16:47 |
| th1a | I remember now. ;-) | 16:47 |
| menesis | but its not written what to do for others that use the ppa | 16:47 |
| th1a | OK. | 16:47 |
| th1a | Yes, could you add that to the notes? | 16:48 |
| menesis | yes | 16:48 |
| menesis | next, I went over the projects in launchpad and set trunk as the development focus | 16:48 |
| menesis | should have done that earlier | 16:48 |
| menesis | some of your branches may point to the wrong location on launchpad now | 16:49 |
| menesis | flourish/trunk | 16:49 |
| menesis | so if you get "diverged" error check where they are being pushed and push --remember ...trunk | 16:50 |
| menesis | also I have upgraded a few libraries, made some zope releases that had accumulated fixes | 16:51 |
| menesis | nothing important, haven't seen anything break yet | 16:52 |
| menesis | yeah | 16:52 |
| * menesis done | 16:53 | |
| th1a | OK. Thanks menesis. | 16:53 |
| th1a | Next week same time, gentlemen. | 16:54 |
| th1a | Have a good week/end. | 16:54 |
| * th1a drops the bag of gravel. | 16:54 | |
| replaceafill | thanks everybody | 16:54 |
| replaceafill | have good week/end | 16:54 |
| th1a | replaceafill can crawl back into bed for a nap. ;-) | 16:54 |
| replaceafill | :D | 16:54 |
| replaceafill | breakfast time ;) | 16:54 |
| yvl | see you all next week :) | 16:56 |
| *** khildin has joined #schooltool | 17:40 | |
| *** replaceafill has quit IRC | 17:43 | |
| *** pablo___ has joined #schooltool | 18:38 | |
| pablo___ | hi | 18:38 |
| pablo___ | parent access? | 18:39 |
| th1a | Hi pablo___. Not really at this point. | 18:54 |
| th1a | They can log in using the student account. | 18:54 |
| *** menesis has quit IRC | 20:39 | |
| *** menesis has joined #schooltool | 20:45 | |
| *** menesis has quit IRC | 22:12 | |
| *** th1a has quit IRC | 22:54 | |
| *** mibofra has joined #schooltool | 23:03 | |
| mibofra | uhm.. hi :) | 23:03 |
| mibofra | I'm a student of an Italy scientific high school, my question is simple... is schooltool prepared to manage a school of 2000 students? | 23:06 |
| *** menesis has joined #schooltool | 23:44 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!