IRC log of #schooltool for Monday, 2010-10-11

*** alga has quit IRC04:19
*** replaceafill has joined #schooltool05:24
*** aks has joined #schooltool06:18
aksreplaceafill: hello06:19
*** replaceafill has quit IRC07:55
*** menesis has joined #schooltool09:21
*** alga has joined #schooltool09:34
*** yvl has joined #schooltool09:56
*** aks has quit IRC10:33
*** menesis has quit IRC11:27
*** aks has joined #schooltool12:02
*** krishnagiri has joined #schooltool12:20
*** menesis has joined #schooltool12:49
*** alga has quit IRC13:25
*** krishnagiri has quit IRC13:27
*** krishnagiri has joined #schooltool13:27
*** krishnagiri has left #schooltool13:52
*** aks has quit IRC14:15
*** Lumiere has quit IRC15:06
*** Lumiere has joined #schooltool15:10
*** Lumiere has joined #schooltool15:10
*** yvl has quit IRC15:19
*** alga has joined #schooltool15:47
*** yvl has joined #schooltool16:17
*** replaceafill has joined #schooltool16:25
*** th1a has joined #schooltool16:27
*** th1a has quit IRC16:30
*** th1a has joined #schooltool16:30
th1ahello aelkner, replaceafill, yvl, menesis.16:31
replaceafillgood morning/afternoon16:32
th1aLets all focus our attention on:
th1aKey dates: toolchain uploaded October 21.16:34
th1aPresumably that's when the repositories open?16:34
th1aSo in two weeks I'd like to be uploading packages.16:35
th1aFeature freeze: February 24th.16:35
th1aI'd like to be set comfortably before that point.16:36
yvlmenesis, maybe you remember when Lucid PPAs were opened?16:36
yvlon toolchain upload, right?16:36
yvlerm, I meant Maverick16:36
yvldo you remember when they did open?16:39
th1aWell, let's go on that assumption for the moment.16:39
* th1a now has his coffee at hand... ;-)16:40
th1aSo we have half of October, November, most of December and January.16:40
th1aFor new stuff.16:40
th1a(this mostly effects yvl at this point)16:41
yvland half of Feb for bugfixes?16:41
yvlthe smaller ones16:41
th1aThat's what I'm thinking.16:41
th1aSo... priorities.16:41
th1aHere's what I see:16:41
th1a* report framework;16:42
th1a* general user experience improvements;16:42
th1a* folding changes from pilots back into core;16:42
th1a* finishing timetables;16:43
th1a* on the packaging side, I'd like to have a clearer SSL story.16:43
th1aI mean actually,16:43
th1aI'd just like to have a clearer port 80 story.16:43
th1anot SSL.16:44
yvljust to throw some things on the table...16:44
yvl* speed optimizations (as in tinkering with schooltool.relationship, catalogs)16:45
yvlthere was something else... (sorry, I managed to get sick, hence a bit hazy)16:46
yvlah, right16:46
yvl* show possible access rights when redirected to login screen (maybe)16:46
menesisI need:16:47
replaceafillth1a, multi-school deployments are not a priority yet, correct?16:47
th1ayvl:  Those would be nice to have, but probably not higher priority than what I had.16:48
yvlth1a, I agree16:48
menesis* use free fonts in reports16:48
replaceafill+1 on free fonts!16:48
th1areplaceafill: Well... I would say not quite yet.16:48
replaceafillgot it16:48
aelknerdefinitely report framework16:49
menesis* convert init scripts to upstart, find a solution for multiple sites on one server16:49
th1aI mean, aelkner and replaceafill should be mostly freed up by next spring.16:49
yvl* (also probably lower priority)16:50
th1aWhat's the significance of init > upstart?16:51
menesis* Rename schooltool.stapp200* to something sensible16:51
th1aYes, please.16:51
replaceafillyvl, would it be possible to speed tests up?16:51
yvlprobably yes :)16:52
replaceafillyvl, i mean, maybe change some to unit tests instead of ftests16:52
yvlespecially those packages that have extensive functional tests and very little unit test coverage16:53
replaceafill* i18n'ed usernames :)16:54
yvlwhat is that?16:54
replaceafillusernames with non-ascii letters16:54
menesisupstart is a new way to define system services and when to start them. init.d script is just a bash script and our is too complex16:54
menesisreplaceafill: that's a bug in zope.16:55
th1aHow hard might it be to fix in Zope?16:55
th1aSo... politics!16:58
menesisone line fix :)16:58
th1aIt is funny since Unicode was such a priority from the start with Zope 3.16:58
menesisit is more effort than that, test, release, etc.16:58
th1aBut... not a difficult initial coding job.17:00
th1aMoving on.17:00
replaceafillok, my priority one question for this week:17:00
replaceafillis RML translatable?!?!?17:01
replaceafilli set i18n:* stuff in the rml template17:01
replaceafillthe strings are extracted17:01
replaceafillbut they are not translated when the pdf is generated :(17:01
th1aOh, dear.17:01
replaceafilli mean, i translated stuff using the view (context, request) approach17:02
replaceafilland just using replace in the request17:02
th1aOK, well, figuring that out is necessary, to be sure.17:03
replaceafillbecause there were a few strings, Class, Teacher, etc17:03
replaceafillbut i was thinking about cando for instance, where one template has a loooot of text17:03
yvlit should be translatable17:04
yvlif not, it's a serious bug17:04
replaceafillyvl, update-translations picks the string in the pt17:04
replaceafilli translate it, compile the strings17:05
replaceafilland nothing :(17:05
replaceafilli even used the "template" attribute trick17:05
replaceafillyou taught me :)17:05
replaceafilland the content is still english17:05
replaceafillsame happens with the current pdf footer in default page templates17:05
replaceafill"Created with SchoolTool"17:06
th1aIt makes sense that it wouldn't be a priority for the rml implementation.17:06
replaceafillPage X17:06
replaceafillah and the date too17:06
th1aWho else cares about making out of the box translatable reports?17:06
yvlI'm pretty sure it's my fault then17:06
replaceafillwhich could be localized17:06
yvlplease file a bug and assign it to me17:06
replaceafillyvl, ah ok17:06
yvlwell, it's not RMLs fault anyway17:06
replaceafillok, moving on...17:07
th1aWe just need to wire it in differently?17:07
yvlI probably wired it wrong17:07
replaceafilli finished the "add multiple students, teachers, administrators" forms17:07
th1ayvl: Whew.17:08
yvlso yes, th1a, test and fix.  Shouldn't be hard17:08
replaceafilli also started with the "global" calculations in cando17:08
replaceafillin cambodia17:08
replaceafillcalculations that take into account all the sections the student is enrolled in17:09
replaceafillMonthly Average Score of the Semester17:09
replaceafillSemester Exam Average Score17:09
replaceafillAverage of the Semester17:09
replaceafilland Annual Average Score17:10
replaceafillalso, dwelsh reported a security issue in cando which i looked at17:10
replaceafillbut didnt find anything unusual17:11
replaceafilli asked david about the specifics, but apparently the issue was reported by someone else17:11
th1aWas I cc:ed on dwelsh's mail?17:11
*** alga has quit IRC17:11
th1aI don't remember that.17:11
replaceafillth1a, not on dwelsh's, i cc you on my response17:11
th1aAh, I see it.17:12
replaceafillso, david said he would get more information and get back to me17:12
aelknerreplaceafill, security issue?17:12
replaceafillaccording to matt gallagher a student was able to see other student's user ids and demographics data17:13
replaceafillalthough he said he didnt confirm the report17:13
th1aYou couldn't repeat it with their Data.fs?17:14
replaceafillno, i couldnt17:14
replaceafillthe user sees what she's supposed to see17:14
th1aWhere are we in the communication loop with Cambodia?17:14
replaceafillah, i had a doubt about asking for scanned forms17:15
th1aYou showed them changes and they said "fine," like, two weeks ago?17:15
replaceafilli should asked them in english of course17:15
replaceafillbut i wanted them to see them in khmer too17:15
th1aOh, right.17:15
replaceafillthen i went to the translation issue in the pdf17:15
replaceafillso, fixing the translation issue in the pdfs would help me for this week17:16
replaceafilli can give it a try if yvl's busy17:17
replaceafill(i know he is :D)17:17
replaceafilland last thing, i kept looking a the fedora 13 packages17:17
replaceafillthey work now, with some little issues about starting the server17:18
th1aThanks for keeping an eye on the Fedora packages.17:18
replaceafilli sent a few comments back to aks17:18
th1aI'd rather have you do the rml fix if yvl can point you in the right direction.17:18
yvlsure, replaceafill - in app/browser/ PDFView.__call__()   xml = self.template()  should produce translated rml17:19
* replaceafill looks...17:19
replaceafillth1a, that's it from me17:19
th1aThanks replaceafill.  So you're going to request those forms?17:19
replaceafilli just wanted to discuss it first17:20
aelknerthings were pretty slow last week, so i decided to take a couple days off17:21
aelknerbut first i added cando and schooltool.intervention to schooltool.niepa17:21
aelknerluckily, integrating cando skin was successful17:22
aelknerthen i installed schooltool.niepa as a sandbox on douglas' server17:22
th1acomponent architecture ftw!17:22
aelkneri started it as a daemon, and it's serving on port 1000017:22
aelkneri recommend using that server to deliver view changes that david will request17:23
th1aWell, it is pretty much up to him.17:23
aelknerin other words, i saw the emails gong back and forth17:23
th1aI don't think he can really rely on an internet connection.17:23
th1aI'd just like him to have a box which he doesn't constantly fiddle with.17:23
aelknerso that's why he wants to serve it up on a laptop17:23
th1aI believe so.17:24
th1aHe can't be completely dependent on it.17:24
aelknerwell, we could give him sandbox instructions for schooltool.niepa17:24
th1aAt this point, he just needs to get his CanDo packages running.17:25
th1aI don't want to start over again.17:25
aelkneryeah, but schooltool.niepa is cando-ready now17:25
aelknerand we can add to it as he needs it17:25
th1aI don't really feel like starting him on a new sandbox, even though in the short term it might be a quicker fix.17:25
th1aWe may have to do it if we can't figure out how he broke his current .deb setup.17:26
aelknerwe don't have a schooltool.niepa deb yet17:27
th1aIt could get pretty arcane if he doesn't know what he's overwritten with easy_install.17:27
th1aI'm feeling a pretty long way from worrying about pushing him customizatinos.17:27
aelkneroh, i see17:27
th1aHe just needs to learn how to *use* SchoolTool.17:28
aelknerwith the rest of the time, i turned my attention to the calendar bug that results from our FET related data17:28
aelknerit turns out that the calendar view code uses lots of zope magic17:29
aelkneri think traversal adapters are great in theory, but overused in practice17:29
aelkneranyway, i've gotten most of the way through tracking down where the view code goes17:29
aelknerand i'm trace debugging now to get down to where the data goes wrong17:29
aelknerit's a case where the app/browser view code calls into (via multi-adapter query)17:30
aelknersome timetable/browser code17:30
aelkneri suspect that it will come down to a timetable data model issue17:30
aelknerbut i don't have the specics just yet17:30
aelkneris it ok if i keep pushing on with this to get it off my plate?17:31
th1aNot forever.17:31
aelknerit won't take forever17:31
th1aDo we know if the time is being modeled wrong in data or just presented wrong?17:31
aelkneri think the presentation might be messed up by the data, but again, i'm not sure yet17:32
aelkneri need just a little more time17:33
th1aI mean, if it is being modeled wrong, it should probably just be left for yvl's overall refactoring.17:33
th1aIf it is hard.17:33
th1aBasically, check in with me as you get to the source.17:33
aelknerremember, the timetable was entered via the XLS import17:33
th1aOh, yeah, so *that* might be the source of the problem.17:34
aelkneragain, maybe17:34
th1aOK, just let me know as you go.17:34
yvlby the way - can anyone give a link to the calendar bug?17:34
aelknerwill do17:34
yvlthank you17:34
* replaceafill stands corrected, rml *is* translatable...17:34
aelknerreplaceafill :)17:34
* replaceafill is only kind of stupid...17:35
yvlah yes, that one17:35
yvllet me remember where the code is17:35
aelkneranyway, th1a, i will check in with you very soon with more complete findings17:35
yvlI think that's not timetable's fault17:35
yvlIIRC it's just a rounding error17:35
yvl(events get into "next" 15 min slot)17:35
yvlnotice the times are 8:5017:36
yvltry changing them to 8:45 or 9:0017:36
yvlshould work well then17:36
aelkneryvl, very interesting17:36
th1aYeah, I tend to think it is not the model.17:36
aelknerbut what if the school starts a period at 8:5017:36
aelknerare you saying that they aren't allowed to do that?17:37
aelkneri mean, without the calendar looking strange?17:37
aelkneryvl says, "notice the times are 8:50...  try changing them to 8:45 or 9:00'17:38
aelkneris he saying, change the data?17:38
yvlif you changed the data, the problem should dissappear17:38
yvlit's a simple matter of rounding actually17:38
yvlsee app/browser/   snapToGrid()17:39
aelknerare you saying the school needs to change their data17:39
aelknerso that the period starts on 8:45?17:39
th1aHe's talking about just testing.17:39
yvlyes, th1a17:39
th1aSo we have overzealous snap to grid?17:40
th1aOr is it snapping to the hour instead of the period boundary?17:40
aelkneron days where it is out of line, the day fails to have a 7:00 period at the top17:41
yvlsorry, I don't remember at hand17:41
th1awell, that gives aelkner a good lead.17:41
th1aaelkner, I'm also anxious to get back to the Mpelembe report card.17:41
aelkneryeah, i'll definitely look at the snap to grid17:41
th1aWe can discuss that once you get this timetable bug done.17:41
yvlone more thing17:41
aelkneryeah, we can talk about that, too17:41
yvlalso check schooltool/timetable/browser/cal.py17:42
yvlDailyTimetableCalendarRowsView is actually the one used17:42
aelkneryeah, that's the code i refered to at the beginning of my report17:42
aelkneri know that is where i will see the problem17:43
yvlbut the problem was something.... that period times are "snapped" or something separately from the events themselves17:43
aelkneri'll post to the developers list so that everyone sees what i find17:43
th1aThanks aelkner.17:43
yvlsure, thanks17:43
th1amenesis:  So is 1.5.2 literally the Maverick release, or will it get a new number?17:43
aelknerthanks for the tips yvl17:43
menesisth1a: I don't know any more fixes made last week17:44
menesisso that's it17:44
th1aOK, fine with me.17:45
th1ayvl:  How are timetables coming?17:46
yvlsome good progress17:46
yvlcan't give the deadline though :|17:47
yvlbits and pieces are falling into place, it's just a pile of work :)17:48
th1aCan we shoot for the end of the month?17:48
th1aI'd like to put it to bed by then.17:49
yvlthey're kind of haunting me already17:49
th1aI believe that will do for today.17:49
th1aWe had a nice smooth release here, and we'll get an early start on package uploading to make the next one even smoother.17:50
th1aHave a great week gentlemen!17:50
* th1a drops the bag of gravel.17:50
replaceafillthanks everybody!17:51
aelknergreat week everyone17:51
yvlproductive week to you all :)17:51
menesisth1a, aelkner: I have made a new zambia package last week. Do I have to package any more pilots? Cambodia, Niepa, something else?17:55
th1aWell, I guess you could make a niepa package, essentially as a stub atm.17:56
th1aOr you could wait until it does more.17:56
menesisjust to make sure I haven't forgotten something17:56
replaceafillso far, i've been handling cambodia demo's using my branch17:57
replaceafillno prob about that though17:57
*** povbot has joined #schooltool19:01
*** menesis has quit IRC19:02
*** ignas has joined #schooltool19:05
*** ignas has quit IRC19:19
*** menesis has joined #schooltool20:06
*** ignas has joined #schooltool20:50
*** replaceafill has quit IRC22:25
*** ignas has quit IRC23:57

Generated by 2.15.1 by Marius Gedminas - find it at!