IRC log of #schooltool for Friday, 2005-01-28

*** tvon has left #schooltool00:03
bskahananyone know why schooltool.main doesn't get a coverage report00:17
bskahanth1a: you around?00:17
*** hazmat has quit IRC00:20
*** tvon has joined #schooltool00:22
bskahanI'm stumped, this is the last piece of migration, and it should be trivial00:25
jintyadeiu, hasta manana!00:32
*** jinty has quit IRC00:32
*** hazmat has joined #schooltool00:34
th1aHad a little nap.00:51
th1aThat dhtml scheduling hack is nice.00:52
bskahanit was similar to a thought I had for the calendar UI00:53
bskahanfor changing event time/day00:54
th1aMore dragging and dropping would be nice.00:54
bskahangot the problem down to one pretty simple example bug01:19
bskahangoing to dinner, write a test case for the bug after01:19
*** hazmat has quit IRC01:23
*** hazmat has joined #schooltool01:23
*** hazmat has quit IRC02:14
*** d2m has quit IRC02:18
*** d2m has joined #schooltool09:29
*** th1a has quit IRC09:40
*** bskahan has quit IRC09:54
*** Aiste has joined #schooltool10:31
*** Aiste has quit IRC11:40
*** Aiste has joined #schooltool11:48
*** mgedmin has joined #schooltool11:53
* mgedmin looking at
* mgedmin installing trang11:59
mgedminone important thing before the 0.9 release12:17
mgedmindo the README and debian/copyright explicitly acknowledge what schoolbell contains some icons from GNOME?12:17
mgedminIf you just define __eq__ but not __ne__ in your class, will Python handle != correctly?12:21
* mgedmin tries12:21
* mgedmin looks for jinty but does not see him12:22
*** tvon has quit IRC12:31
*** SteveA has quit IRC12:31
*** tvon has joined #schooltool12:31
mgedmintvon, where do the nice schoolbell icons come from?12:35
mgedmincould you update the copyright section in the README (and debian/copyright) to explicitly mention them?12:39
mgedmincurrently the README claims that all files in src/schooltool (except some in src/schooltool/translation) are (c) shuttleworth foundation12:40
*** SteveA has joined #schooltool12:46
*** mgedmin has quit IRC13:03
*** Aiste has quit IRC13:22
*** Aiste has joined #schooltool13:38
*** SteveA has quit IRC14:04
*** SteveA has joined #schooltool14:06
*** gintas has joined #schooltool14:09
*** mgedmin has joined #schooltool14:56
*** Aiste has quit IRC15:17
*** alga has joined #SchoolTool15:34
*** th1a has joined #schooltool16:31
*** jinty has joined #schooltool18:26
*** SteveA_ has joined #schooltool18:30
*** SteveA has quit IRC18:31
*** SteveA_ is now known as SteveA18:31
* jinty just read the irc logs18:31
jintymgedmin: __eq__ was only for the unit tests, it has absolutely no use outside that and is not in any interfaces. Do i need to bother to define __ne__? everything seems to work18:36
jintymaybe I should anyway.18:38
*** Aiste has joined #schooltool18:46
* jinty considers committing an imaginary timetabling client with illusionary tests under src/timetabling18:47
mgedminjinty, a simple ``def __ne__(self, other): return not self.__eq__(other)`` would be nice19:26
jintywill do19:33
*** gintas has quit IRC19:49
tvonmgedmin: theyre GNOME, I'll update it now20:07
* mgedmin tries the db conversion script20:10
mgedminno exceptions20:10
mgedminI'm not sure what the expected contents of the database should be, though20:11
* jinty is going to create a trunk/packaging directory20:16
*** bskahan has joined #schooltool20:48
* tvon debates moving all the gnome icons into a subdir21:13
th1aI'd prefer that.21:13
bskahanshould put up a wiki page with the icons we want to replace21:14
tvonwww/gnome/ ? since they are all from gnome21:14
th1aThere aren't many, right?  If we're getting rid of the little "event" and "resource" icons.21:14
bskahanmgedmin: I can't get the same error to occur in a test21:15
tvonth1a: there are 12 I believe21:16
th1aOK.  I guess I know where to find them now.21:16
mgedminbskahan, are you talking about a regular unit test, or the db conversion test script?21:16
mgedminwhere do you see the error then?21:16
bskahanboth, the same thing works fine in the unit test21:16
mgedminI ran the test script today21:17
mgedminI saw no errors21:17
mgedmincan you commit a version of the script that shows the error?21:17
bskahanone sec21:17
bskahancommitted with 3 commented lines21:23
bskahanuncomment them to break migration21:23
* mgedmin looks21:24
bskahansomething about accessing link.source in anyway stops the method21:25
*** Aiste has quit IRC21:31
mgedmindid not fail for me21:31
bskahantry hitting the site21:33
bskahanafter its running in the db-convert test21:34
bskahanshould throw TraversalError: 'community'21:34
bskahanon the front page21:34
bskahanbecause the migration never actually happens21:34
mgedminI expected an exception21:35
bskahanit fails silently21:35
mgedminwtf?  try: ... except: pass !21:35
mgedminok, I got the error now21:36
mgedmin'except: pass' is evil!21:36
* bskahan agrees21:36
th1aDon't be evil.21:36
bskahanit relates to the test stubs21:36
bskahanwill fix21:37
* mgedmin is mystified21:42
bskahanme too21:43
* mgedmin hates pdb21:50
mgedminhmm, I found the pickled version of that Link object22:20
mgedminthe string 'source' does not appear in the pickle22:20
bskahanshouldn't it fail to provide the interface in that case?22:22
mgedminif you call verifyObject on it, I expect it would fail22:29
mgedminok, I'm going to check out 0.8 and try looking inside that Data.fs with the old version22:30
*** bskahan has quit IRC22:33
*** bskahan has joined #schooltool22:34
mgedminsomething strange22:38
mgedminthere is only one person ("Manager") in the database22:38
bskahanat what point?22:38
* mgedmin pauses22:38
bskahanthere should be manager and bskahan22:39
mgedmind'oh! mv Data.fs-0.8 Data.fs22:39
* bskahan nods22:39
mgedminI get the same error with 0.822:42
mgedminif I insert a loop that prints all sources of all links inside prepareDatabase22:43
mgedminkinda strange, isn't it?22:43
bskahanthe only thing I could think of was something not being setup yet22:43
bskahanbut what gets me is that link.reltype is correct22:44
mgedminall of the attributes seem to be correct22:44
bskahanunless they need source22:44
bskahanthe sane link objects do have a source if you test it in other places22:45
mgedminI found three links on persons, and they all show the error22:46
*** hazmat has joined #schooltool22:53
* mgedmin slaps himself22:55
mgedminlinks did not have 'source' or 'target' in 0.822:56
mgedminthey had '__parent__' and 'traverse()'22:56
mgedminyou can replace '' with 'link.relationship.traverse(link).__parent__' and it should work22:59
bskahanyeah thats it23:00
mgedmingood luck23:01
mgedminI'll be going then23:01
mgedminhave a nice weekend23:01
*** mgedmin has quit IRC23:01
th1aThat was a tricky one.23:17
* bskahan nods23:27
bskahanno uselful traceback23:27
bskahanjust failed silently23:27
bskahanwhich is still wierd23:28
* jinty hates plone23:32
th1aAre you trying to do structured text with it?23:32
th1aSo are we doing an rc now or are we just going to put out 0.9?23:33
bskahanshould still do an RC23:33
bskahanjust for people to report bugs23:33
bskahanI'm not sure how many people are downloading RCs and reporting bugs, but its good to give them the chance23:34
jintyI'm in between, but lean to the way we normally do it. i.e. RC23:34
th1aI think it is good for us finding bugs.23:35
th1aThe fact of the matter is that it doesn't make much difference.23:35
th1aI'd lean toward doing an rc.23:35
jintyIn fact, yes, I think the policy should be: for RC's most bugs can be fixed23:35
jintyand after, security only23:35
* bskahan agrees23:36
* jinty thinks about writing up a release policy23:36
jintyso, bskahan, can I go and branch the trees now?23:37
* jinty taps his foot impatiently23:39
bskahannot yet sorry23:39
bskahanI have to do the relationship migration without any of the normal relationship tools23:40
bskahansince they expect Links to have the current attributes23:40
th1aWhich relationships?23:40
bskahanURIMembership and URISubscriptuion23:40
bskahannormally its easy23:41
jintyin theory everything is simple23:41
* jinty jumps up and down, screams at plone, then decides brute force is the best option23:52

Generated by 2.15.1 by Marius Gedminas - find it at!