IRC log of #schooltool for Wednesday, 2014-10-08

*** menesis has quit IRC01:40
*** replaceafill has quit IRC01:49
*** mgedmin has quit IRC03:39
*** mgedmin has joined #schooltool03:42
*** mobert has quit IRC07:53
*** yvl has joined #schooltool08:04
*** mgedmin has quit IRC08:12
*** mgedmin has joined #schooltool08:12
*** khildin has joined #schooltool08:48
*** khildin has quit IRC11:36
*** menesis has joined #schooltool11:39
*** khildin has joined #schooltool11:49
*** khildin has quit IRC12:11
*** menesis has quit IRC14:16
*** menesis has joined #schooltool14:57
*** replaceafill has joined #schooltool17:26
th1ahi replaceafill.17:33
th1aHow long did that take you?17:33
replaceafillhey th1a17:33
replaceafill~2.5 hours17:33
replaceafillthe header was the tricky part17:33
replaceafilli didn't want to use a "manual" approach17:33
replaceafillbecause i wanted to know how far table formatters can get17:33
th1aSo... what is the difference in the two approaches?17:35
*** replaceafill has quit IRC17:38
*** replaceafill has joined #schooltool17:39
replaceafillsorry th1a17:39
replaceafillgot disconnected17:39
replaceafillit took me a while to notice :(17:39
th1aSo... what is the difference in the two approaches?17:40
replaceafillwell, table formatters give you sorting, column handling etc17:40
replaceafillfor "manual" tables17:40
replaceafillyou usually end up duplicating that17:41
replaceafilljust to get the template a little easier to handle17:41
th1a"Manual" would just be writing <table><tr>, etc.?17:41
replaceafillbut i just set three rows for the header in the formatter17:41
replaceafilland used styles to merge the cells17:42
replaceafillone thing i did notice17:42
replaceafillis that we should move every table to use paragraphs17:42
replaceafillwhen a simple cell17:42
replaceafilllike <td>foo</td>17:42
replaceafillis rendered17:42
replaceafillyou always have the risk that it'll overflow17:43
replaceafillif the content doesn't fit17:43
replaceafilli think that's something we can safely land in the core tables17:43
replaceafillwith a couple of paraStyle definitions17:43
th1aThat makes sense.17:44
replaceafilli mean, i don't see any advantage of using <td>content</td> over <td><para>content</para></td>17:44
replaceafilli also fixed a couple of issues in core tables17:45
replaceafilllike some fonts getting bigger17:45
replaceafillin some cases17:45
replaceafilland fixed a feature where you can register a table using just strings in python code17:45
replaceafilli tried to use zcml to set the same attribute17:45
replaceafillbut zcml sets unicode17:46
replaceafillnot a big deal, but it's something nice you can do17:46
replaceafillone thing i did notice today when i was generating the sample to send you17:46
replaceafillis that Rick's report has two columns with *17:46
replaceafillNeeds attention columns17:47
replaceafilland they use an area at the bottom of the page for that17:47
replaceafill"* Please clarify all checks in columns marked with an asterisk"17:47
replaceafillwe just need a custom page template to put that in every page, nothing major17:48
th1aAt this point, it is up to your mood if you want to do that right now.17:50
replaceafilli also worked on the journal, fixed the bug on the absences by day report and added a new test for it17:50
replaceafillah ok17:50
replaceafillyeah, shouldn't take long17:50
th1aJust remember this is a favor.  ;-)17:50
replaceafilli started the last report, absences for range of dates17:50
replaceafillbut haven't finished it yet17:50
replaceafillwell, it's good training i must say17:50
replaceafillreal (weird) cases :D17:51
replaceafilli'm not scared of headers anymore ;)17:51
replaceafilli plan to start on packaging today17:51
replaceafillhope to get it done17:52
replaceafillthat's it from me17:52
th1aYeah, I'm hoping these will start to feel routine.17:52
th1aMaybe we can even post some prices.17:52
replaceafilli'd say most of these reports have taken ~2h each17:54
replaceafillsome more, some less obviously17:54
replaceafillbut that's a good estimate at this point17:54
th1aThe big thing is counting task switching, setup and teardown.17:54
th1aI mean, that's what I want to make sure we include that.17:54
th1aSo assuming the skeleton of a package exists,17:54
th1aIt is still about 2 hours.  You could do three or four in a full day?17:55
replaceafillthree i'd say17:55
replaceafilli mean, at a decent speed17:55
replaceafillwe do need the skeleton17:56
replaceafillit's easy to copy schooltool.cando and regexp strings, etc17:57
replaceafillbut something that does that would be really nice to have17:57
replaceafillincluding a /debian directory17:57
th1aWell, I'd like to document that too.17:59
th1aSo if you automated it, it would be a lot easier to document.17:59
th1aWe can definitely put that on the to-do list after the School page.17:59
th1aSo we'll release these as 2.8.2?18:01
replaceafilland 2.8.1 for the journal iirc18:01
replaceafilland i just used 2.8.0 for schooltool.encinitas18:02
replaceafillah question18:03
replaceafillcan i create a schooltool-owners ppa to put encinitas in?18:03
replaceafillor do you want me to use replaceafill for that?18:03
th1aYou can use owners.18:03
th1aWe're doing it as official ST work.18:04
th1aIn appreciation for some good examples of reports we need.18:04
replaceafillone thing i'd like to try for the journal release18:05
replaceafillis Gediminas instructions on how to update translations18:05
replaceafillhe documented that really well18:05
replaceafilland i'd like to test it18:05
th1aThere is no screaming rush.18:06
replaceafillcool, thanks18:06
th1aAnd I know people have been working on them.18:06
replaceafillah yes, someone is updated the Catallan translation18:07
replaceafilli'd love this photo:18:09
replaceafilli wonder if it's "real" :)18:09
replaceafilli love*18:09
replaceafillthat translation makes no sense :D18:10
replaceafillth1a, why the suicide + python tweet?18:11
th1aOK, I picked up on that.18:11
replaceafillis it because you're getting back to programming?18:11
replaceafillor because what you're discovering on the data?18:11
th1aSo the number crunching is on survey data from students, specifically on depression and suicide.18:11
th1aAs it turns out, our high flying charter schools also lead the state in depression and reported suicide attempts in middle schoolers.18:12
th1aSo, that's a pretty big deal, especially since I don't think any other state collects and publishes that data every year at the school level.18:13
th1aIt was just in a bunch of pdf's so you couldn't rank them.18:13
replaceafillwere you able to parse them?18:13
replaceafilli mean, to get all the data18:13
th1aYeah, like I said, the tables I couldn't do, but luckily what I really needed was not in tables.18:13
replaceafillgot it18:14
th1aI just can't go down the rabbit hole of looking for correlations with all the other data, which is probably a good thing.18:14
th1aI'm pretty much done with getting the data now.18:14
th1aTook me a while to get the school names sanitized so I could get it in a tidy  worksheet.18:15
th1aJust school names are a nightmare here.18:15
th1aWe keep renaming, merging and splitting schools.18:15
replaceafillsounds like an interesting research18:16
th1aIt is going to be a pretty big deal, actually.18:16
th1aThis is something that's just been anecdotal up to this point.18:17
th1aAnd we really have years of consistent survey data sitting there.18:17
th1aAnyhow, I got through the last of the programming last night.18:19
th1aI took me a while to figure out that I had some "ghost" data because even though I'd deleted some CSV files there were still gedit temp files copies of them lying around.18:19
th1aThat drove me INSANE for about a half hour.18:19
th1a"BUT I DELETED THAT!"18:20
th1aThe parsing is more or less just like this:18:21
th1a            for count in range(len(l)):18:21
th1a                if l[count][:6] == "so sad":18:21
th1a                    depressed = l[count+17].rstrip()18:21
th1aFind this string and jump down 17 lines.18:21
replaceafillyou should try pandas18:22
th1aYou should be encouraging me to stop fiddling.  ;-)18:22
th1aOK.  Thanks replaceafill.18:22
replaceafillthanks th1a18:22
*** replaceafill has quit IRC19:18
*** replaceafill has joined #schooltool19:18
*** menesis has quit IRC22:57
th1areplaceafill:  Did you see Glenda's email?23:05
replaceafillbut haven't checked anything yet23:06
th1aOK.  Try to have a look today if you can.23:06
replaceafilldo you want me to switch to it?23:06
replaceafillon it...23:06
replaceafillth1a, i was able to reproduce it locally23:16
replaceafillit's the future again :(23:16
replaceafillsetting a student as Pending in a future date23:16
replaceafillbreaks the gradebook23:16
th1aThe whole gradebook. OK.23:17
replaceafillyeah, it's gradebook specific23:17
replaceafill(not cando)23:17
replaceafilli'll fix it23:17
th1aOK.  Cool.23:17
replaceafillthe others errors seem just unauthorized pages asking for password23:18
th1areplaceafill:  Is there some reason CanDo reports wouldn't work for past years?  I know in general they would work.23:26
replaceafillnot that i can think of th1a23:27
replaceafilli mean23:27
replaceafillmost of them work on the right contexts23:27
replaceafilllike terms or years23:27
replaceafillor sections23:27
*** menesis has joined #schooltool23:36
*** menesis has quit IRC23:37
*** menesis has joined #schooltool23:37

Generated by 2.15.1 by Marius Gedminas - find it at!