IRC log of #schooltool for Monday, 2011-11-14

*** menesis has quit IRC01:49
*** menesis has joined #schooltool02:00
*** ignas has quit IRC02:11
*** menesis has quit IRC08:19
*** menesis has joined #schooltool10:17
*** menesis has quit IRC14:20
*** ignas has joined #schooltool15:06
*** menesis has joined #schooltool15:52
*** replaceafill has joined #schooltool16:17
replaceafillhhmm i'm getting an evolution error after updating the gradebook16:28 (5, u'schooltool.gradebook', 5)16:28
replaceafillmake: *** [run] Error 116:28
replaceafillcan someone confirm that?16:28
aelknerreplaceafill, can tell me more about the error?16:31
aelkneris there a traceback?16:31
replaceafillaelkner, yes, let me paste it somewhere16:32
replaceafilli can't find the relevant line in evolve16:32
replaceafillah, it's above16:34
replaceafillhold one16:34
replaceafilli get that on the vps, i'll try to reproduce it in marcela's laptop16:36
* replaceafill 's laptop is officially dead :(16:36
aelkneri see what happened16:36
aelkneri need to fix the script to not expect there to be layouts for a given year, my bad16:36
replaceafillah ok16:37
th1aGood morning aelkner, replaceafill, yvl, menesis.16:37
replaceafillgood morning/afternoon16:38
th1aYes, afternoon too!16:38
th1areplaceafill:  What's the status of the book template?16:39
replaceafillalmost done, had some laptop issue last week, dead :(16:39
replaceafillwill push today16:39
replaceafillafter i transfer all my files to marcelas laptop16:40
th1aOK.  Get that on LP asap.16:40
yvlgood morning16:41
th1aThis is "no really, do the docs NOW" week for me.16:41
th1ahi yvl.16:41
* menesis hi16:41
th1aI've really only ever written under serious deadline pressure.16:42
th1aI'm pretty much kidding myself otherwise.16:42
th1amenesis, why don't you start us off.16:42
th1aAre we on track for 2.0 next Tuesday?16:42
menesisI have little to report16:42
th1aAnything we should be thinking about?16:42
menesiswe are on track16:43
menesisyvl has fixed all tests16:43
th1aOn track as long as I do my job now.  ;-)16:43
menesisand everyone fixed some bugs16:43
menesisit is halfway, and half bugs are fixed16:43
th1aWhat's our last day to fix bugs?16:43
menesisso I think we are doing ok16:43
menesisI scheduled next release for Thursday, Nov 1716:44
menesisthat would be a rc16:44
menesisI have tried to setup daily builds16:44
menesisfrom flourish branch to /dev ppa16:44
menesisexisting branches conflict, will have to make a copy16:45
menesisplanned how to transition branches and series  flourish -> trunk -> 2.016:46
menesiscreated a 1.7 directory on ftp for the old trunk16:46
menesisadded another versionator builder, so now versionator tracks precise and versionator-1.7 oneiric16:47
menesisadded flourish coverage builder16:47
menesisthis is all about
menesisthat is green for a few days now16:47
th1aLooks good.16:48
th1aSounds good.16:48
menesistoday synced zope packages from debian unstable to ubuntu precise16:48
menesisso finally completed the round-trip between ubuntu and debian16:48
menesiscompleted only half of packages16:49
menesisbut they need little work16:49
menesislooked at generated docs, tried to improve Makefile to not include buildout.cfg in bzr, etc. will see later when to merge that16:50
menesisdon't remember what else16:50
th1aOK, thanks menesis.16:51
aelknerreplaceafill, could you pull the gradebook trunk now and try again?16:51
yvlfixed some bugs16:52
yvlone left for me:
yvlbtw, a short personal note16:53
yvlI'd like to thank you guys for writing functional tests16:53
yvland menesis for pushing on fixing them16:53
yvlwe actually broke ST in some places while implementing new UI16:54
yvland tests helped to find those rare cases16:54
yvla big *thank you*!16:54
yvlit seems that keeping the old skin was a good bet.16:54
th1aaelkner:  Is this fixed?
* yvl done16:56
th1ayvl:  Would you like this one?
aelknerth1a, yes, i'll change the status16:56
yvlI'd rather postpone new features at this time16:57
th1aDo this one first yvl:
th1aYeah, you're probably right yvl.16:57
yvlI can try on that one16:57
yvlthere are reasons why it wasn't fixed for so long ;)16:57
th1aI'd like ti fixed...16:58
* yvl too16:58
yvlwhen an event comes from 2-3 calendars16:58
yvlwhich one should be picked as "main"?16:58
yvlsay, event in section calendar books a resource16:59
yvlyou're a leader of the resource16:59
yvland teacher of the section16:59
yvlso in personal calendar you get 3 events16:59
th1aActually, I'd say the main thing is taking these out of the personal calendars.17:00
th1aI wouldn't worry about the resource/section duplication.17:00
aelknerreplaceafill, sorry, didn't push, could you try pulling now?17:00
replaceafillaelkner, yeah i noticed17:00
th1aI don't know why they are in the personal calendar at all.17:00
replaceafillaelkner, it worked17:01
yvlpeople who are responsible for booking a lot of equipment want to look at all of calendars at the same time?17:01
th1aIf it is overlays, they have a choice.17:02
yvlit's not about a choic17:02
yvlit's about event duplication17:02
yvlnevermind, I'll do whatever17:03
th1aBasically, I think the answer to your question is "yes."17:03
th1aIf you want to track a bunch of equipment, you'll have a bunch of overlaid calendars.17:03
th1aNow, we could use a better interface for actually seeing a bunch of events, but that's a different problem.17:04
yvlno problem17:04
yvlI'll just clean up the most common case scenario17:04
yvlwill fix it properly at same later time17:04
yvlif anyone understands that there's a problem, that is ;)17:05
th1aOh, I think it is definitely a "wtf" issue.17:05
th1aFor me it is a "didn't we solve this four years ago?" issue.17:06
yvlno we did not17:06
th1aIf you run out of bugs poach some from replaceafill or aelkner.17:06
th1aasdf jkl17:06
yvlasdf it is :)17:07
th1aDamn mini scrollbars...17:07
* replaceafill doesn't like unity so far (two days using it)...17:07
th1aSo replaceafill and aelkner, please mark what you're working on as "in progress."17:07
replaceafillwill do17:08
th1aUnity is ok on my small laptop, doesn't work on multi-display desktop.17:08
th1aOK, thanks yvl.17:09
replaceafilllast week, i worked on the css of calendars17:10
replaceafilland buttons17:10
replaceafillplease look at
yvl(feel free to assign more bugs to me on occasion ;) )17:10
replaceafilland make sure they look ok for you17:10
yvlfirst thought - Windows XP17:10
yvlbut I actually like them17:11
replaceafillin the calendar, th1a and i worked on all the views, monthly, sidebar, weekly, yearly, etc17:11
replaceafillalso check the birth date in the person add from17:11
replaceafillwe styled the jquery calendar17:11
replaceafilland you can log in as a teacher and see the gradebook color changes17:12
replaceafillthe old green, yellow, red17:12
replaceafillfor cells17:12
replaceafilland scores17:12
replaceafillth1a, vinny didn't answer my question about the error "red"17:13
th1areplaceafill:  Those are automatically also used in the Journal too, right?17:13
th1areplaceafill:  Which question was that?17:13
replaceafillgo to the person add form, and hit submit17:13
replaceafillwith no data17:13
replaceafillyou'll get the errors in 3 fields17:13
replaceafilllook at the field's focus border17:14
replaceafillthe color vinny sent is too pink17:14
menesisI don't like how buttons stand out (the border: outset bit). I replaced it with 1px solid, and it looked better.17:14
th1aJust use the same red as the field at the top.17:15
replaceafillth1a, so i'll use the same red we use for the error message17:15
replaceafilli'd tell you the hex code but i don't have firebug yet :D17:15
replaceafilli also worked on the book style, as i said i'll push when i move my stuff here17:16
replaceafilli have many bugs for this week, so i'll work on that17:17
replaceafillth1a, i think that's it from me17:17
th1areplaceafill:  Feel free to just assign a couple to yvl.17:17
th1aThanks replaceafill.17:17
replaceafillah sure17:17
aelknerok, so i finished up the report sheets hide feature17:18
aelknerluckily, david used the niepa demo and found a bug before we released it17:18
aelkneri had forgotten to include the report card layouts in the evolve script, so i added that17:19
aelkneralso, fortunately, replaceafill just found out that the new code crashed if a year has no layouts17:19
aelknerso i just fixed that, too17:19
aelkneralso, i added error messages to the deploy form17:19
aelknerbefore, it would just do nothing if no template selected17:20
aelknerthen we added the title field which is also required, so both fields have error messages now17:20
th1aHm... maybe you should write some functional tests trying to break this.17:20
th1aOr just generally keep trying to break it.17:21
aelknerwhich part?17:21
aelknerthe UI for deploying?17:21
aelknerwith the error messages?17:21
th1aThe "it crashes if you haven't also done X" kind of bugs.17:22
aelknerdon't know of any that match that description at the moment17:22
th1aaelkner has no sense of QA.17:23
th1aI'm just saying, it seems like you missed a lot of possible scenarios.17:24
th1aAre you sure there are no other ways to break this?17:24
aelkneri don't think we can be of that for any part of schooltool17:25
aelknersure of that, i mean17:25
aelknerbut writing funcitonal tests for deploying and hiding would make sense17:25
aelkneri haven't done any selenium tests yet because of the pressure to get the hide/unhide done17:26
aelknerbut if we have time, i could do that next17:26
aelknerbut let me finish my report17:26
th1aWell, it would probably be worth a day at this point.17:26
aelkneri fixed the following bug:17:27
aelknerit was crashing on reimporting contacts17:27
aelkneri also added a reimport tests to sample_data.txt, checking the the same data was there17:27
aelknerafter reimport, that is17:28
aelknerand on last thing, the niepa report card request problem17:29
aelkneri still can't reproduce it17:29
aelknereven though david's link does cause it, that link should never happen17:29
th1aIn what sense?17:30
aelknerwell, perhaps the session timed out before he hit Download17:30
aelknerand the modal dialog can't handle that17:31
aelknerbut we have other problems with our report request dialogs17:31
aelknerthey still don't handle the situation when the user is logged out17:31
aelkneryvl, can you check this link out?17:31
th1aWhat does "that link should never happen" mean?17:32
aelknerif you request a report card, no problem17:32
aelknertry opening a second browser instance and log out there17:33
aelknerthen go back to the one that still thinks it's logged in17:33
aelknertry requesting the report card then17:33
yvlto sum up17:33
aelknerit brings up a dialog with login17:33
yvlbtw, what's the password there, aelkner ?17:34
th1adon't post that in public.17:34
aelknerstandard manager password shoudl work17:34
yvlno it does not17:34
aelkneroh, my browser has it memorized, one sec17:35
aelkneryvl, i emailed it to you17:36
yvlah, that one17:37
yvlyes, a known problem17:37
yvldialogs don't handle such cases17:37
yvlI've tried to bring that up several times, but...17:38
* yvl has nothing more to say :)17:39
aelkneryvl, that whole area of the code could use your love17:39
aelknernot just the reportLink stuff, butalso the modal dialog stuff17:39
th1aSo what happens?  You get a login form jammed into the dialog?17:40
aelknerit is difficult to figure out each time a developer needs to add a modal17:40
aelknerth1a, yes17:40
th1aIs that the problem David had?17:40
aelkneryvl, i was hoping you would not only clean up th code , but also document it some for us17:40
th1aI mean, it is a bug, but it shouldn't be too hard to recover from for the user.17:40
aelknerth1a, i can't say specifically what davd did17:40
yvlnever got the time for that, sorry man :|17:40
aelknerhe could have gone all the way to calling up the request dialog17:41
aelknerthen going away until the session timed out17:41
yvlcall up the dialog, go to lunch - and here you go :)17:41
th1aIsn't he getting some kind of error?17:42
th1aWhat does "that link should never happen" mean?17:42
aelkneri'm saying, as a result of some ind of bug, that link ends up in his url area17:42
aelknerand since it is a bad link, it renders the bad html view rather than the pdf17:43
th1aWhat's the correct URL?17:43
aelkneri'm not sure, the whole thing gets subverted in the modal dialog js process17:44
aelknerit's difficult to follow, so i can't say17:44
aelknernote when you request the report, you don't see the url with post data in it17:45
th1aSo basically, we have no idea what the problem is.17:45
aelkneryou just get the pdf17:45
aelknerthat's why i'm asking yvl to look at it, but you have to agree that he has time for that17:45
th1aI don't see any reason to think this has anything to do with the dialog issue.17:46
aelkneroh, i thought i gave a pretty good one17:46
th1aDoes it allow you to reproduce it?17:47
aelkneri haven't reproduced it, but it's not easy to try all the scenarios17:47
aelknerand if there are scenarios that require going away and timing out, they are difficult to set up17:48
aelkneryvl, can we simulate this kind of thing with selenium tests?17:48
th1aI just forwarded David's screenshot to replaceafill and yvl.17:49
th1aHave a look at that.17:49
th1aYou're rendering a dictionary?17:49
th1aI guess I was wrong with my first guess because I was thinking it was trying to render PDF syntax as html, but I didn't look very closely.17:50
yvlits JSON17:51
yvlI wonder why/how17:52
aelkneryvl, again, it's the whole modal js stuff17:52
aelknerit needs your love :)17:52
yvlyes it does need some love :)17:52
aelknercan you simulate timeout in selenium tests?17:53
th1aSo that's the package that should be going to the dialog.17:53
yvlah, I see what happened17:53
yvlthe dialog link is rendered with href to the view that returns JSON'ed modal dialog contents17:54
yvlbut the link is replaced with javascript that calls modal dialog as soon as html is loaded17:54
yvlusers just needed to hit it before any js is executed17:55
yvla slow laptop running tons of stuff might help ;)17:56
aelknerthis is part of what i mean by 'we could use developer docs for this stuff'17:56
yvlI think there are more places that may have the same problem17:56
th1aOr perhaps a slow connection?17:56
th1aVery slow.17:56
yvl+ chromium17:56
yvlit tends to render eagerly17:56
aelkneryes, david definitely has a slow connection17:57
th1aSo anyhow, can we fix this specific issue?17:57
yvl0.5-1 days17:58
th1aOK.  Can you file the bug yvl?17:58
th1aOK, thanks yvl.17:58
aelknercan you add some docs, maybe refactor for simplicity a little?17:58
th1aI don't think docs would have helped with this.17:58
replaceafillyvl, if you touch flourish.js, can i make a request? :)17:58
* yvl is sorry to say - maybe17:58
yvlsure replaceafill  :)17:59
aelknerok, never mind, perhaps we could discuss developer docs after the release17:59
replaceafilladd the ability to open dialogs without the a.href :)17:59
yvlok ;)17:59
th1aI guess you should just move on in your bug queue then, aelkner.18:00
aelkneryes, we had a lot of other discussion, so i didn't get to that yet18:01
aelkneri have two that i've identified18:01
aelknernote the comment menesis made at the end18:01
aelknerif the user enters 1234 for a user id18:02
aelknerand the Calc/Excel app saves 1234.018:02
aelknerand the user doesn't notice that and moves on18:02
aelknerwe had agreed that the code that is looking for a string and finds a number18:02
aelknerthat it would say i couldn't find a string18:03
aelknerth1a, it's your call, i could auto convert to string18:03
aelknerand let the user find out about the problem18:04
aelknerwhen 1234 elsewhere is saved somehow as a string18:04
aelknerand therefore, wouldn't match the 1234.018:04
aelknerkeep in mind 1234.0 is a valid user id18:05
th1aI'm not sure what we can do other than save as strings and read as strings.18:05
th1aIs there anywhere we actually want integers?18:05
th1aOr floats.18:05
aelknerdemos perhaps18:05
aelkneri'm sure there is more than one place18:06
th1aWe can't truncate .018:06
aelknerwhat if bob.0 is valid?18:06
aelknershouldn't 1234.0 be valid?18:06
th1aAs a string.18:06
th1aI'm not sure what the question is.18:06
aelkneri'm saying trunkcating the .0 would be hostile to the user who wants it18:07
th1aIf excel turns the year into a number and adds .0 we can't really do anything about that.18:07
th1aWe just have to read it as text, right?18:07
th1aI think we're all in agreement.18:07
aelknerand accept it, that's what i was suggesting18:07
aelknerit's just that users can enter lots of numbers without looking18:08
aelknerlike accountants do18:08
aelknerso the problem will be caused by them and we won't protect them18:08
th1aAs far as I know, there was never even an alternative proposal, so yes, go ahead.18:08
aelkneractually you and i had already discussed this and came to the 'dn't allow integer ids' decision18:09
aelknerperhaps that was hasty of us at the time :)18:09
th1aIs that the same decision we've made now?18:09
aelknerno, we decided to change it18:09
aelknerif the use enters 1234, doesn't see that it got converted18:10
aelknerthey will jsut have funny schooltool error messages18:10
aelknerthings like, '1234' is not a valid user id18:10
th1aMoving on...18:10
th1aAnything else?18:10
aelknerso, cool, i can fix that one18:11
aelknerone more:18:11
aelknerteacher should be able to see what the student would see18:11
aelknerwhat should the menu item title be?18:11
aelknerright now all the menu has is Score18:12
aelknerhow about 'Student view'?18:12
aelknerit's not a report, it's the html table view that the student gets18:13
aelknerstudents can't change of course, but they can view their grades18:13
aelknerit's the /mygrades view18:14
aelknerone more step back18:14
aelkneroh, you are suggesting now, i see18:14
aelknerit's not a summary18:14
aelknerit's the student's gradebook view18:14
th1aIt is a detailed student view.18:15
th1aCOmpared to the regular view.18:15
th1aIt is a HTML report.18:15
aelknerno, it's just a view18:15
th1aMake it "Report."18:15
aelknera student clicks the Gradebook tab18:15
th1aAnd we'll add a dialog and PDF version later.18:15
aelknerthat's how they see it18:15
th1aFrom the teacher's point of view it is a report.18:16
th1aThe fact that it was written for the student originally should be completely irrelevant.18:16
aelkneri think that conflates the word, report a bit18:16
th1aIt IS A REPORT.18:16
aelknerpdf is the most typical report18:16
th1aWe have differnent types.18:16
aelknerthese result in downloads18:16
th1aYou can have a HTML report.18:16
th1aWell, this one.18:17
aelknerin schooltool, i mean18:17
aelknerno, other than this one18:17
th1aIt is REPORTING.18:17
aelkneryou don't have to shout since i'm obviously going to do whatever you want me to do18:17
th1aJust make it 'Report' and we'll add to it later.18:18
aelknerso those two bugs will be my next tasks18:18
aelkneri'll make sure to in progress for now on18:18
aelknerif i remember :)18:18
* yvl coughs18:19
aelknerit's like when Herman Cain says he'll take a lie detector test18:21
th1aHad to go to the door.18:21
aelknerthen says, 'if i think it is appropriate'18:21
th1aSee you guys Wednesday.18:21
yvlthanks guys!18:21
th1aWe should have the front page template Wednesday.18:21
aelknercya guys then18:21
* th1a drops the bag of gravel.18:22
replaceafillthanks everybody18:22
*** replaceafill has quit IRC19:07
*** replaceafill has joined #schooltool19:10
th1areplaceafill:  Also, I'll need to be able to have the latest colors for doing screenshots, so make sure they're in trunk or somewhere else I can get to them.19:11
replaceafillth1a, they are19:11
replaceafillsomehow, i missed this comment from menesis
replaceafillsorry menesis19:12
th1aYeah, I'd like to see that.19:13
replaceafillth1a, submit with no data
replaceafilli don't like errors very much...19:18
replaceafillth1a, just applied the button change too19:20
* replaceafill likes the outset better :)19:21
replaceafillalthough i like how the 1px solid looks in chrome :)19:22
replaceafillwhatfont, nice :)19:29
replaceafillit's a chrome extension for finding fonts in the page19:53
replaceafillvery nice19:53
th1aAre you restarting that server?19:53
replaceafillah yes, hold on19:53
replaceafillth1a, restarted19:56
replaceafilli applied menesis patch for buttons19:56
th1atbh I like the new one better.19:56
replaceafillme 2 :)19:57
th1aWhat's the problem with the errors.19:57
replaceafilli like the shadow effect19:57
replaceafillah hold on, i reverted that change19:57
th1aYou like the menesis style buttons.19:57
th1aOK.  Cool.19:57
th1aSettled then.19:58
replaceafillwe can leave the outset ;)19:58
replaceafilldamn, i hate this keyboard!!!19:58
replaceafilleverything is different :(19:58
th1aleave the outset?19:59
th1aAre you messing with me?19:59
replaceafillno :)19:59
replaceafillok, refresh person add19:59
replaceafillsubmit with no data20:00
replaceafillthe fields with errors look too red to me20:00
th1aIs it the same red as the text?20:01
th1aAnd the border above?20:01
th1aI think you're nit-picking then.20:01
th1aIt looks fine to me.20:01
replaceafillwill push these changes then20:01
replaceafillis it me or the hovers color in menesis patch is different20:09
* replaceafill sets the old style to see20:09
*** menesis has quit IRC20:10
replaceafillah! right, we used light teal for the secondary buttons20:10
th1aWhat was I going to say...20:11
replaceafillok, after 5 hours comparing colors, i'll leave our work from last week on buttons :D20:11
th1adid you update the favicon to the transparent one Vinny game us.20:12
replaceafilllet me check20:12
th1aIt still has a little noise around the edges.20:12
th1a(that could be in his too)20:12
replaceafillhhmm i don't see it in gimp
th1aIt is the gray green pixels in the corners.20:16
th1aAnd for that matter the gray around the hexagon.20:16
th1aThey show up against a dark background.20:16
th1aLike an inactive firefox tab on Ubuntu.20:16
th1aHowever, it is probably better than the alternatives.20:16
th1aWhich would make the inner hexagon invisible against a dark background and probably screw up the shape of the circle.20:17
th1aDoes your firefox have a dark background for inactive tabs?20:18
replaceafilllet me set the old ubuntu theme20:18
replaceafilli changed it to a ligth gray20:18
replaceafillthat's the dark one20:18
th1aThe current one would look better against a light gray background, which is the more important case.20:18
replaceafillah! i see it th1a20:18
th1aIt doesn't look BAD.20:19
replaceafillyou mean this:
replaceafillmy firefox doesnt show dark inactive tabs by default20:20
th1aHow do you feel about it?20:22
replaceafilli'll try to clean it :)20:22
replaceafilldont like it either20:22
th1aI guess you can fiddle with it bit by bit pretty easily.20:22
th1aThe problem is you'll probably turn it into a hexagon.20:22
th1aWhich may look worse in just the regular contexts.20:22
th1aWe could also set a background, but I think that looks worse in general too.20:23
replaceafillmaybe messing with the scaling options we can get the circle20:23
th1aI'm dubious.20:24
th1aThis is probably not a good use of your time.20:29
replaceafillyou're right, it ends up looking like an hexagon :(20:32
replaceafillok, i'll move on then20:32
th1aI think it is good enough.20:32
replaceafillth1a mozilla uses a bigger image as their icon
replaceafill<meta content="" name="og:image">20:48
replaceafilland starbucks sets 3 icons in one!
replaceafilllol, th1a
replaceafillsee the favicon :D20:56
th1aHm?  You've lost me replaceafill.21:01
replaceafillhold on21:01
th1aHow does it look on a light background?21:03
replaceafilllet me show you firefox21:04
replaceafillfirefox with active and inactive21:06
th1aHm... that's probably better.21:06
th1aI don't know.21:06
th1aI mean, it is weirdly shaped.21:06
th1aI think the dark tabs is a relatively uncommon case.21:07
replaceafillyeah, i'll leave the one from vinny21:08
* replaceafill goes to buy a sata-usb adapter21:11
*** replaceafill has quit IRC21:12
*** menesis has joined #schooltool21:14

Generated by 2.15.1 by Marius Gedminas - find it at!