IRC log of #zope3-dev for Tuesday, 2006-02-14

*** J1m has joined #zope3-dev00:02
*** natea has joined #zope3-dev00:06
*** dobee has quit IRC00:13
*** J1m has left #zope3-dev00:14
*** SiggyF has quit IRC00:26
*** edgordon has quit IRC00:36
*** zbir has quit IRC00:40
*** _anguenot has joined #zope3-dev00:42
*** Aiste has joined #zope3-dev00:47
*** doko_ is now known as doko00:58
*** wrobel has quit IRC01:08
*** GaryPoster has quit IRC01:15
*** benji has quit IRC01:29
*** sashav has quit IRC01:30
*** MrTopf has quit IRC01:32
*** SiggyF has joined #zope3-dev01:33
*** natea has quit IRC01:33
*** natea has joined #zope3-dev01:34
*** stu1 has joined #zope3-dev02:08
*** stub has quit IRC02:28
*** russf has quit IRC02:34
*** russf has joined #zope3-dev02:35
*** vlado_ has joined #zope3-dev02:37
*** vlado has quit IRC02:37
*** zbir has joined #zope3-dev02:39
*** philiKON has quit IRC02:40
*** SiggyF has quit IRC02:41
*** vlado_ has quit IRC02:49
*** vlado_ has joined #zope3-dev02:49
*** stu1 is now known as stub02:51
*** J1m has joined #zope3-dev02:57
*** zbir has quit IRC03:03
*** GaryPoster has joined #zope3-dev03:09
*** TresEquis has joined #zope3-dev03:35
*** niemeyer has quit IRC03:44
*** deo has quit IRC03:49
*** J1m has quit IRC03:50
*** newpers has joined #zope3-dev03:59
*** natea has quit IRC04:30
*** natea has joined #zope3-dev04:31
*** russf has quit IRC04:43
*** russf has joined #zope3-dev04:43
*** RockyBurt is now known as RockyBurt|Zzz04:58
*** vlado__ has joined #zope3-dev05:05
*** tiredbones has left #zope3-dev05:05
*** vlado_ has quit IRC05:06
*** newpers has quit IRC05:24
*** zbir has joined #zope3-dev05:30
*** yota has quit IRC05:36
*** natea is now known as natea|zZz05:36
*** vlado__ has quit IRC05:37
*** vlado__ has joined #zope3-dev05:37
*** vlado__ has quit IRC05:47
*** vlado__ has joined #zope3-dev05:47
*** ChanServ sets mode: +o hazmat05:47
*** sm has quit IRC06:12
*** Theuni has quit IRC06:13
*** vlado_ has joined #zope3-dev06:14
*** vlado__ has quit IRC06:14
*** TresEquis has quit IRC06:15
*** vlado__ has joined #zope3-dev06:25
*** vlado_ has quit IRC06:27
*** kamalgill has joined #zope3-dev06:53
*** vlado_ has joined #zope3-dev07:06
*** vlado__ has quit IRC07:06
*** vlado_ has quit IRC07:27
*** vlado__ has joined #zope3-dev07:27
*** dobee has joined #zope3-dev07:40
*** eins has joined #zope3-dev08:18
einshi08:18
*** vlado_ has joined #zope3-dev08:19
*** vlado__ has quit IRC08:19
*** kamalgill has quit IRC08:20
*** vlado__ has joined #zope3-dev08:30
*** vlado_ has quit IRC08:30
*** jinty has joined #zope3-dev08:39
*** vlado__ has quit IRC08:51
*** vlado__ has joined #zope3-dev08:51
*** dlk has joined #zope3-dev08:58
*** zagy has joined #zope3-dev09:08
*** hdima has joined #zope3-dev09:16
*** j-w has joined #zope3-dev09:17
*** wrobel has joined #zope3-dev09:19
*** agroszer has joined #zope3-dev09:24
*** Aiste_ has joined #zope3-dev09:28
*** Aiste has quit IRC09:32
*** jinty has quit IRC09:33
*** romanofski has joined #zope3-dev09:39
romanofskimoin09:40
agroszermoin09:40
*** Aiste_ is now known as Aiste09:46
*** vlado_ has joined #zope3-dev09:53
*** MJ has quit IRC09:55
*** vlado__ has quit IRC09:55
*** mgedmin has joined #zope3-dev09:57
*** andres has quit IRC10:04
*** andres_ has joined #zope3-dev10:04
*** agroszer has quit IRC10:14
*** __mac__ has joined #zope3-dev10:15
*** xenru has quit IRC10:15
*** vlado_ has quit IRC10:18
*** vlado_ has joined #zope3-dev10:19
*** vlado_ has quit IRC10:22
*** andrew_m has left #zope3-dev10:28
*** Aiste has quit IRC10:36
*** jinty has joined #zope3-dev10:43
*** sashav has joined #zope3-dev10:44
*** tarek has quit IRC10:46
*** tarek has joined #zope3-dev10:47
*** marchino has joined #zope3-dev10:47
*** Aiste has joined #zope3-dev11:00
*** MrTopf has joined #zope3-dev11:00
*** jinty has quit IRC11:03
*** russf has quit IRC11:03
*** russf has joined #zope3-dev11:05
*** MJ has joined #zope3-dev11:14
*** romanofski has quit IRC11:16
*** romanofski has joined #zope3-dev11:16
*** russf has quit IRC11:16
*** andres_ is now known as andres11:52
*** RockyBurt|Zzz has quit IRC11:56
*** RockyBurt has joined #zope3-dev11:57
*** dunny has quit IRC12:07
*** marchino has quit IRC12:16
*** mkerrin has joined #zope3-dev12:33
*** philiKON has joined #zope3-dev12:34
*** alga has joined #zope3-dev12:46
*** ignas has joined #zope3-dev12:50
*** povbot has joined #zope3-dev13:03
*** povbot` has quit IRC13:04
*** niemeyer has joined #zope3-dev13:10
*** russf has joined #zope3-dev13:42
*** MrTopf has quit IRC13:48
*** MrTopf has joined #zope3-dev13:48
*** mgedmin has quit IRC13:55
*** jinty has joined #zope3-dev14:04
*** mgedmin has joined #zope3-dev14:30
*** tiredbones has joined #zope3-dev14:42
*** natea|zZz is now known as natea14:51
*** jinty has quit IRC14:54
*** Borax has joined #Zope3-dev14:56
*** zbir has quit IRC14:57
BoraxHi :-)14:57
*** yota has joined #zope3-dev15:07
*** edgordon has joined #zope3-dev15:17
*** tiredbones has quit IRC15:18
*** tiredbones has joined #zope3-dev15:22
sashavin zope.app.catalog.README.txt there is an example how to create a keyword index but there isn't any ready one as the TextIndex which uses zope.index.text , any special reason that you need to implement your own KeywordIndex?15:22
*** mgedmin has quit IRC15:23
*** AJC has joined #zope3-dev15:24
philiKONsashav, what do you mean, there's zope.index.keyword.index15:24
philiKONactually, zope.index.keyword.KeywordIndex15:24
sashavyes but can zope.app.catalog.text.TextIndex uses zope.indext.text.TextIndex to make it ICatalogIndex... so I figured you need to a KeywordIndex that implements ICatalogIndex too15:25
sashav-can15:26
philiKONICatalogIndex seems overrated :)15:29
philiKONwell, perhaps not15:29
philiKONstill it'd be trivial to make a catalog-compatible keyword index15:29
sashavyes it's trivial.. just copy the README example.. I just wondered if there was a reason for not having one there?15:30
philiKONwell, the README example is even more complicated than necessary15:30
philiKONsimply use this:15:31
philiKON    >>> class KeywordIndex(zope.app.catalog.attribute.AttributeIndex,15:31
philiKON    ...                    BaseKeywordIndex,15:31
philiKON    ...                    zope.app.container.contained.Contained,15:31
philiKON    ...                    ):15:31
philiKON    ...    zope.interface.implements(zope.app.catalog.interfaces.ICatalogIndex)15:31
philiKONexcept BaseKeywordIndex would be zope.index.keyword.KeywordIndex15:31
sashavexactly15:31
philiKONfeel free :)15:31
niemeyerMy commit messages are being sent to the checkins list with the wrong email. Is anyone here able to fix it?15:49
*** benji_york has joined #zope3-dev15:49
philiKONniemeyer, changing the info in your zope.org account should do it15:49
*** doko has quit IRC15:49
*** doko has joined #zope3-dev15:49
* philiKON wonders what niemeyer is committing :)15:49
*** benji_york is now known as benji15:49
benjimurder?15:49
niemeyerphiliKON: Thanks, will have a look at it. My email changed recently.15:49
philiKONbenji, ah, yes. and you're next :)15:50
benjilol!15:50
philiKONhey benji15:50
niemeyerphiliKON: I'm changing ZCML to parse files twice as fast..15:50
benjihi15:50
philiKONniemeyer, weee15:50
benjiooh, I like it!15:50
philiKONhow do you do that?15:50
niemeyerLying15:51
philiKONi knew it15:51
* benji cries fowl!15:52
*** zbir has joined #zope3-dev15:59
*** mgedmin has joined #zope3-dev16:03
philiKONbenji, you comin to pycon?16:03
*** andres_ has joined #zope3-dev16:05
benjiunfortunately not16:06
benjiif it were in DC I would have16:06
*** MJ has quit IRC16:08
*** edgordon has quit IRC16:14
*** jhauser has quit IRC16:20
*** edgordon has joined #zope3-dev16:22
*** andres has quit IRC16:23
*** mexiKON has joined #zope3-dev16:30
*** hdima has quit IRC16:30
*** philiKON has quit IRC16:38
*** sashav has quit IRC16:41
*** jhauser has joined #zope3-dev16:49
*** mexiKON is now known as philiKON17:02
*** tonico has joined #zope3-dev17:08
*** sawdog has joined #zope3-dev17:10
AJChmmm, did you guys fix doctest for code coverage analysis in SVN?17:26
mgedminAJC, I would gladly take a look at the problem, if I had a test case17:27
SteveAmgedmin: when you did schooltool work recently, did you notive any changes in how resources are looked up in zope3.2 ?17:28
AJCwell, the doctest file in SVN is not the same as the standard python distribution, and the standard one does not work... plus the guy who wrote coverage.py points to your SVN server and says you seemed to have fixed it.17:28
AJCi had a version from early january lying around, and that seemed to report too little coverage for a file i had extensively tested, but the HEAD revision now gives me 100% for everything...17:29
mgedminSteveA, no17:30
mgedminAJC, what is "my SVN server"?  source.schooltool.org?17:31
mgedminI've seen various unit test coverage breakages with schooltool17:31
AJCno, no :-)  the zope3 svn17:31
mgedminfor various reasons17:31
mgedminah, ok17:31
mgedminin that case "you" did not refer to me personally17:32
mgedminI have seen a comment mentioning code coverage in zope.testing.doctest17:32
mgedminI have fixed some code coverage problems in a closed-source Zope3-based project17:32
mgedminby replacing import doctest with import zope.testing.doctest, and fixing an infinite recursive loop in one __getattr__ method17:32
mgedminI do not know if replacing 'import doctest' was really necessary17:33
mgedminbecause it did not fix the problem17:33
mgedminbut the __getattr__ loop did fix the problem17:33
AJCwell, i just plopped the current zope.testing.doctest into my python Lib directory...17:33
mgedminand I was too lazy to revert import changes and see if the problem reoccurred17:33
AJCi have no looping problems, just erroneous reports...17:33
mgedminthe looping was not a problem17:33
mgedminbut it had a side effect that caused erroneous reports17:34
AJClooking at the coverage files, it seems the format has changed.  now they print line numbers for each line of code instead of >>>> to mark unused lines... so i presume the current parser fails17:34
mgedminno17:34
mgedminthe numbers indicate how many times each line was executed17:34
mgedmininstead of 0 it shows ">>>"17:34
AJChmmm, there are no >>>> in my files anymore...17:35
mgedminyou might want to grab http://source.schooltool.org/svn/trunk/schooltool/bin/coverage_reports.py17:35
AJCwhat does it do?17:35
mgedminand generate nice HTML repors like on http://source.schooltool.org/coverage/17:35
mgedmin(you will need to hack the source and remove the 'ignore everything outside schooltool' filter17:35
mgedmin)17:35
AJCwell, once i have valid reports then HTML would make sense, but everything is 100% at the moment!17:36
mgedminis that bad?17:36
mgedmin:)17:36
AJCheh, yes because i know it's wrong17:36
mgedminif you have a function you suspect is never called during the run of unit tests17:36
mgedminand the coverage report shows it is being called17:37
AJCi just wrote one on purpose.17:37
AJCthe point is, i don't see any >>> marks in front of it17:37
mgedminwhat happens if you insert 'raise NotImplementedError' there, and rerun the tests?17:37
*** MiUlEr has joined #zope3-dev17:37
AJC(as i did before using Zope's doctest)17:37
AJCtests still run, no exceptions17:38
mgedminso you want to say Zope's doctest makes the reports accurate17:39
mgedminand the stdlib's doctest doesn't?17:39
AJCwell, Zope's doctest fixed the problems with code coverage not even working, but the latest head revision reports 100% everywhere17:39
mgedminare you running coverage reports on Zope 3 sources, or your own code?17:40
AJCmy own code17:40
mgedminis it secret code?17:41
AJCnot really, but it's company code, closed source...17:42
AJChow do i run the zope coverage?17:42
AJCi guess the same problems are there17:42
mgedminlast time I tried test.py --coverage on a package (I forgot which one I tried), it produced accurate results17:54
mgedminAJC, those lines that are shown as covered17:54
mgedminis there a number next to them, or nothing at all?17:54
*** eins has quit IRC17:55
AJChmm, well it works for zope :-)   yes, there are some numbers...17:56
mgedminhmm17:58
mgedminI am mystified17:58
AJCi'm wondering what i'm doing differently!17:58
mgedmindid you try the 'raise NotImplementedError' trick?17:58
AJCyep17:58
mgedmindid it raise an exception?17:59
AJCbasically, i have my own tests.py which creates a bunch of doctest suites, and adds other test suites depending on parameterss17:59
AJCno exception, code coverage 100% ;)17:59
mgedmindid you remove *.cover before trying?17:59
AJCyep18:00
AJCshould i just use Zope's test runner?18:01
AJCdo you have a custom tests.py file?18:01
mgedminI tend to use schooltool's test.py18:01
mgedminbut it doesn't support the new zope 3 test runner features such as layers18:02
mgedminthe intention is to merge whatever enhancements schooltool's test.py had into Zope 318:02
mgedmintest.py --coverage did not have any enhancements, but the code itself is different18:02
mgedminif anybody finds any differences between z3 test.py --coverage and ST test.py --coverage, I'd like to hear about that18:03
benjiAJC, IIRC the zope.testing testrunner does quite a bit of work to do coverage properly18:03
AJCok, so i should just reuse that :-)18:03
benjior do what mgedmin suggests and improve zope's with your enhancements and then use that18:04
AJCwell, i guess they already support what i need, it's pretty basic18:04
*** wrobel has quit IRC18:06
*** wrobel has joined #zope3-dev18:07
SteveAbenji: do you know if Jim is around today?18:10
benjiHe's about 5 feet from me... does that help?18:11
*** J1m has joined #zope3-dev18:11
SteveAhi jim18:11
J1mhey18:11
SteveAdo you have a few minutes to discuss an issue with changes in how views and interfaces work together in zope 3.2?18:12
J1muh18:12
J1mnot really, but ok18:12
SteveAok, thanks18:12
SteveAseems that views (pages, in particular) are now registered to provide Interface18:13
J1mI don't think that was new in 3.2.18:13
J1mI don't think it was new in 3.1.18:13
SteveAtraversal presentation components provide IPublishTraverse / IBrowserPublisher18:13
J1myup18:14
SteveAwhen a view is looked up, such as the "resource locator view", the class view from namespace.py is used18:14
J1mBTW, I want to define a more specific interface for pages to provide.18:14
SteveA        view = zope.component.queryMultiAdapter((self.context, self.request),18:14
SteveA                                                name=name)18:14
SteveAusing this code18:14
SteveAnow, if I have a root object, and it is supposed to have a page (view) that is a resource container registered with name=''18:14
SteveAand i also have a custom URL traverser registered for it, with name=''18:15
SteveAbecause the resource locator provides Interface, and the url traverser provides IPublishTraverse18:15
SteveAthe url traverser is looked up for the no-name view, rather than the resource locator18:15
J1mthat should not be the case.18:16
SteveAboth components are registered for the same (context, request-type)18:16
SteveAand the same name18:16
SteveAand differ only in the interface they provide18:16
J1mwell, the specificity of the required interfaces takes precedence.18:17
J1mIn that case, the one closest to the requested provided interface should win.18:17
SteveAthe view-looker-upper call to queryMultiAdapter doesn't specify a requested provided interface18:17
SteveAso i need to look up what that comes out as...18:17
J1mBut I think it defaults to Interface.18:17
SteveAyes18:18
SteveAdef queryMultiAdapter(objects, interface=Interface, name=u'', default=None,18:18
SteveA                      context=None):18:18
J1mAgain. I'm not happy that pages are registered as providing Interface.18:18
J1mI'd love to fix that for 3.3.18:18
SteveAdo you think this might be also a bug in interfaces?18:18
J1mIt's possible.18:19
SteveAi mean, is it behaving as designed that although Interface was requested, given the choice between a component that provides Interface and one that provides something more specific,18:19
SteveAthe one providing something more specific wins?18:19
SteveAi'm not at all sure what the right thing would be here18:19
SteveAmaybe tossing a coin? ;)18:19
J1mIf two adapters are registered for the same required types, and different provided types, and you do a lookup for the more general provided type, you should get the corresponding adapter.18:19
J1mNo, the one providing the less specific interface is supposed to win.18:20
SteveAokay, i'll put together a test case for this18:20
SteveAi think the wrong one is winning18:20
J1mk18:20
SteveAthanks for the chat, Jim18:20
J1mnp18:20
J1mI look forward to high-bandwidth coimmunication at pycon. :)18:20
SteveAyeah, me too18:20
*** dlk has quit IRC18:24
*** j-w has left #zope3-dev18:24
*** J1m has left #zope3-dev18:25
*** zagy has quit IRC18:38
*** sawdog has quit IRC18:38
*** sashav has joined #zope3-dev18:44
*** sawdog has joined #zope3-dev18:45
*** sm has joined #zope3-dev18:52
*** dobee has quit IRC18:54
*** MiUlEr has quit IRC19:03
*** AJC has quit IRC19:09
*** romanofski has quit IRC19:10
*** edgordon is now known as ed_lunch19:21
*** russf has quit IRC19:30
*** russf has joined #zope3-dev19:31
whithowdy russf19:31
whithey yall, is the boston skin the best example of viewlets?19:32
mgedminhmm, now I'm unable to make 'test.py -pv -u -s mypackage --coverage=coverdir' produce any output19:36
mgedminit doesn't even mkdir coverdir19:37
*** xenru has joined #zope3-dev19:38
*** ruda_porto has joined #zope3-dev19:41
mgedminis it constrained to ignore packages outside the 'zope' namespace?19:42
mgedminor does it not like my symlinks?19:42
*** russf_ has joined #zope3-dev19:47
srichtermgedmin: it's probably the symlinks19:48
*** russf has quit IRC19:48
srichterwhit: boston skin is a good example; SchoolTool has some usages too19:49
srichter( the boston skin is quiet formal about them though19:49
* whit nods19:49
russf_whit:howdy!19:58
russf_srichter: hey Stephan. Your book should arrive on my doorstep tomorrow!20:00
*** jinty has joined #zope3-dev20:04
srichterrussf_: great! :-)20:08
russf_srichter: another wheel on your maserati!20:09
srichterhe he20:11
*** dunny has joined #zope3-dev20:11
smlol20:14
*** ignas has quit IRC20:15
*** whit has quit IRC20:42
*** andres_ is now known as andres20:45
*** whit has joined #zope3-dev20:45
sashavis there a concept as meta_type/portal_type in zope3 or is all that identified by interface?20:59
RockyBurtsashav: identified by the interface which extends IContentType21:03
RockyBurtwell, in the case of portal_type anyhow21:03
sashavok, and how is a catalog query from zope2 like: catalog( { 'portal_type' : ¨CMFDocument' } ) done in zope3? a InterfaceIndex?21:05
*** edgordon has joined #zope3-dev21:05
RockyBurtsashav: honestly i've never done catalog queries using zope 3.x -- i assume you're talking about deploying on zope 3.x and not zope 2.x21:07
RockyBurtsince you can't really use zope3 content types on zope 2.x21:08
*** ed_lunch has quit IRC21:08
sashavyes catalog in zope321:08
sashavyou can't? even with five?21:08
sashaveh never mind zope2 :)21:09
RockyBurtzope2 content types still require all the cmf stuff including a valid portal_type, even if you use five/zope321:09
*** whit has quit IRC21:10
*** benji has quit IRC21:17
*** kamalgill has joined #zope3-dev21:22
*** agroszer has joined #zope3-dev21:24
*** sashav has quit IRC21:33
*** benji has joined #zope3-dev21:38
*** agroszer has quit IRC21:38
*** mgedmin has quit IRC21:40
*** dunny has left #zope3-dev21:49
*** alga has quit IRC21:53
*** Aiste has quit IRC22:01
*** hazmat has quit IRC22:33
*** Aiste has joined #zope3-dev22:43
*** mkerrin has quit IRC22:44
*** deo has joined #zope3-dev22:45
*** dobee has joined #zope3-dev23:03
*** dobee has quit IRC23:08
*** jinty has quit IRC23:24
*** andres has quit IRC23:25
*** whit has joined #zope3-dev23:40
*** sawdog has left #zope3-dev23:45
*** marchino has joined #zope3-dev23:52

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!