IRC log of #schooltool for Friday, 2009-11-20

*** dlobo has joined #schooltool00:32
*** replaceafill has joined #schooltool00:46
replaceafillping aelkner00:47
replaceafillaelkner, never mind i just got it :)00:49
*** dlobo has quit IRC00:49
*** menesis has joined #schooltool01:08
*** replaceafill has quit IRC01:13
*** dlobo has joined #schooltool01:34
*** dlobo has quit IRC02:20
*** krushik has quit IRC03:02
*** krushik has joined #schooltool03:09
*** alga_ has quit IRC04:08
*** pcardune_ has joined #schooltool04:14
*** pcardune has quit IRC04:31
*** replaceafill has joined #schooltool04:34
*** pcardune has joined #schooltool05:48
*** th1a has quit IRC05:48
*** dlobo has joined #schooltool05:49
*** dlobo has quit IRC07:05
*** pcardune has quit IRC08:44
*** alga has joined #SchoolTool10:15
*** replaceafill has quit IRC10:28
*** replaceafill has joined #schooltool10:38
*** alga has quit IRC11:18
*** alga has joined #SchoolTool11:58
*** yvl has joined #schooltool12:18
*** alga has quit IRC12:26
replaceafillyvl, ping14:02
yvlpong, replaceafill14:03
replaceafillyvl, basic python question14:03
replaceafillproperties are always calculated when they are called?14:03
replaceafillor they are just created once and cached or something?14:03
yvlproperty is just a way to call the function14:04
yvlso yes, every time14:04
replaceafillah, ok, just checking :)14:04
replaceafilli have a property that is accessing a changing list14:04
yvlZope has.....14:04
yvl"Lazy" decorator to deal with this thing14:04
yvlthat computes once14:04
yvlbut I strongly recommend not to use it14:05
yvlbecause it behaves weirdly in the debugger14:05
replaceafillno, but i want the property to always compute :)14:05
yvloh, ok then :)14:05
replaceafillthanks man14:05
yvlwelcome :)14:05
*** ignas has quit IRC15:16
*** jelkner has joined #schooltool15:23
*** th1a has joined #schooltool15:39
*** ignas has joined #schooltool15:45
* th1a goes to pick up aelkner.15:46
*** mgedmin has joined #schooltool16:00
*** dlobo has joined #schooltool16:00
th1ayvl:  ayt?16:32
yvlhi th1a16:32
th1aOK we're here.16:32
th1ahi aelkner.16:33
yvlI'm just finishing the last bug report16:33
yvl3-4 mins16:33
th1aWe were just discussing that fixing the ugly email hack is probably first on aelkner's list.16:33
ignasyvl, unccl oveguqnl!16:35
yvlignas: you ok there?16:35
ignasyvl, why, yes, I am fine, thank you ;)16:36
yvlthen thank you!!!16:37
yvlok, I'm ready16:43
yvlsorry for the delay16:43
th1aNow aelkner is filing a bug...16:43
yvlI'd like the failing test to be top priority16:45
yvlas long as it fails, other changes are not tested16:46
yvlunit tested16:46
yvlthe thing is, it should not be fixed by allowing empty FROM16:48
yvlit should be fixed by applying desired dublin core information16:48
yvlthen I'd suggest to move on the email hack16:49
th1aaelkner is looking at his bugs.16:49
aelkneryvl: do you need a couch to talk about your feelings on dublincore16:50
yvlnaah, AK will do16:51
yvlbut seriously, that should be replaced16:52
th1aDo we not want to use DC for creation time?16:54
th1aDo we not want to use it at all?16:54
yvlI'd like to see it gone entirely16:55
th1aAside from this specific point, we're also trying to establish that this is the correct way to do it in the future for the sake of data integrity?16:55
th1aHandling disappearing users, etc?16:55
yvlit's defintely a bit safer than strings stored in dublincore16:56
yvlas for the correct way of handling data integrity...16:56
yvlI don't have a final solution yet16:56
yvlbut that will have to be addressed16:57
th1aUltimately there is no solution for data you need being deleted from the database.16:57
th1aAside from generally trying to handle it as gracefully as possible.16:57
yvlthat is true16:57
yvlalso having relationships will give me a little bit more freedom16:58
yvlas the code sits in ST core, and in one place only16:58
yvlwe will have to address turning the plugins off at some point16:58
yvland I'd rather have a consistent mechanism that needs to be replaced16:59
th1aSo creation date just as a regular attribute instead of a DC attribute?16:59
yvlideally, yes16:59
yvlbut just fixing creator is also fine17:00
yvlas both creator and goal objects are ST objects17:00
yvlthen again, it will be easier to write one evolution script that adds relationship and extracts the date to an attribute17:01
th1aOK, I talked aelkner into it.17:02
aelkneryes, i will do what you just described17:03
yvlthanks, aelkner!17:03
aelknerwe decided that leaving dublincore in place (and just not using it) will make evolution much easier17:03
th1aSorry, we need to call Chris A. at SLA for a sec...17:05
th1aTrying to get their report cards out.17:05
* th1a is losing interest in supporting rich text entry...17:17
yvlit's a quite good feature17:19
*** alga has joined #SchoolTool17:19
yvlbut one should expect a lot of quirks with it17:19
th1aApparently so.17:19
th1aBut yes, I know it is something people want.17:20
yvlwell, it's one of those areas17:20
ignaswe use ckeditor over here, seems to be less f than fck editor17:20
yvlinternationalization is extremely difficult, nobody gets it right the first time, few get it right in first 200 times17:20
th1aI think they changed the name.17:20
yvlsame goes for time zones17:20
yvlsame for document formatting17:21
yvlthat's just my opinion17:21
ignasyou forgot security and UI ;)17:21
yvloh yes!17:21
yvland extentions17:21
ignasyou just have to look at it the Edison way, you just found one more way how *not* to do it ;)17:22
yvl"Hmm, it seems that ST is not a light bulb.  Oh well."17:22
yvlth1a, ping me when you're done with Chris17:25
th1aWhat a massive pain in the ass.17:33
th1aIt is all the project manager's fault... "hey, let's add a rich text widget."17:34
ignasit was me17:35
ignasLyceum asked me to add it17:35
ignasso I did17:35
th1aYes, it is not like people don't *want* it.17:35
th1ayvl:  We're going to have to take a couple more minutes to look at this problem now that we're all worked up.17:36
yvlgo ahead17:36
yvlif it's not too hard, send my some info on the problem, when you can17:36
yvlwhat data breaks what and where17:37
yvlhaving the data and pdfs would be nice17:37
th1aBasically, in Safari FCKEditor doesn't work.17:37
th1aSo they just get the regular text widget.17:37
th1aAnd then we lose their line breaks because we're expecting HTML paragraph breaks, (I guess).17:38
th1aSo we have to sniff that out or something.17:38
yvlI see17:38
yvlI kind of thought we don't support Safari that much17:38
th1aRIght now aelkner is just writing a little script to dump the data into a text file so we can see exactly what the data looks like.17:38
yvland IE17:38
th1aWell, we should.17:38
th1aBut we're pretty much dependent on their users reporting bugs.  ;-)17:39
yvlit is possible to do automated testing on that17:39
yvlbut that is a lot of pain17:40
yvl(well, AFAIK possible)17:40
th1aI'm sure it is possible.17:41
yvlth1a: I kind of expected a similar problem, but I hoped that it did not show it's ugly head for a year or so17:41
th1aThis is the problem with having users.17:42
yvlby the way, this is a good example of bugs I like to fix in one bunch17:43
yvltweak FCK editor options, add filter for RML, update the widget to deal with disabled JS, etc.17:44
th1aHm... I see it has spell check.17:45
th1aThat in itself is pretty huge for schools... not sending home comments with spelling errors is a big deal.17:45
th1a(ckeditor has it)17:46
yvlneed to check if we want to user some newer version of CK editor17:47
yvlth1a: maybe you wan't to log a bug?17:47
th1aProbably we do... where do we get it from?17:47
th1aA package?17:47
yvlI think so17:49
yvllooking for it...17:49
yvlthis one17:54
yvlit's latest version is
yvlwe're kind of terribly horribly outdated :)17:56
yvl2.0.0 uses CK editor that "seems to be less f than fck editor" (c) Ignas17:57
th1aOK, looks like we can manage this now that we've got a good look at the data.18:00
ignasyvl, hmm, i think I have an html cleanup function that you might find useful by the way18:05
yvlcan you mail it to me?18:06
th1aI don't want to break aelkner's flow...18:06
yvlthat's ok18:06
yvlignas: cleaning html is always a beach :)18:07
yvl(especially I need to revisit cleaning up for RML)18:08
ignasyvl, sent18:08
yvlah, the nifty lxml cleaner18:10
yvlthank you, thank you18:10
yvl    cleaner = Cleaner(18:10
yvl        scripts = True,18:10
yvl        javascript = True,18:10
yvl        annoying_tags = True,18:10
yvlI especially like the sound of this parameter :)18:10
ignas<blink>Why would you want to clean that?</blink>18:11
yvlnobody appreciates the beauty of blinking any more...18:13
yvlth1a: can we move the rest of this meeting to Monday, if there's something more to discuss?18:13
yvlto summarize, I'd be happy if Alan fixed first18:14
yvland this one last:
yvlthis one second:
*** dlobo has quit IRC18:15
yvlthis one third:
yvlgot the email.... what a mess!18:23
yvlthanks, this will be very, very useful for testing.18:23
th1aIt has, however, screwed up our planning meeting...18:24
yvlwell, these things happen from time to time18:24
* yvl is quite happy to get such testing cases18:26
yvlth1a: so what's the plan?18:33
yvl(for today, at least)18:33
th1aAgh... I guess we'll finish this project.  We're SO CLOSE...18:33
th1aSo you can go home.18:33
th1aPerhaps we still need to meet Monday though.18:33
yvlwell, I think interventions are a priority now18:34
th1aYes, we can discuss sequence on Mondayu.18:35
yvlsorry that I got to go18:36
th1aNo problem.18:36
th1aSorry about the distraction.18:36
yvlno problem18:36
yvlgood luck with the fix!18:36
*** yvl has quit IRC18:36
*** dlobo has joined #schooltool19:41
*** pcardune has joined #schooltool20:07
*** ignas has quit IRC20:14
*** pcardune has quit IRC20:15
*** jelkner has quit IRC20:31
*** dlobo has quit IRC20:47
*** jelkner has joined #schooltool20:55
*** dlobo has joined #schooltool20:56
*** jelkner has quit IRC20:56
*** jelkner has joined #schooltool20:57
*** dlobo has quit IRC20:58
*** mgedmin has quit IRC21:05
*** jelkner has quit IRC21:11
*** pcardune has joined #schooltool21:22
*** dlobo has joined #schooltool22:33
*** dlobo has quit IRC23:04
*** aelkner has quit IRC23:45
replaceafillmenesis, are you there?23:50

Generated by 2.15.1 by Marius Gedminas - find it at!