IRC log of #zope3-dev for Friday, 2005-11-04

*** efge has joined #zope3-dev00:07
*** tarek has joined #zope3-dev00:11
*** GaryPoster has quit IRC00:34
*** lunatik has joined #zope3-dev00:41
*** lunatik has left #zope3-dev00:43
*** projekt01 has joined #zope3-dev00:47
*** srichter has joined #zope3-dev00:49
*** ChanServ sets mode: +o srichter00:50
*** J1m has left #zope3-dev01:04
*** SteveA_ has quit IRC01:59
*** einheit_ has joined #zope3-dev01:59
*** stub has joined #zope3-dev02:02
*** einheit_2 has joined #zope3-dev02:24
*** zbir has joined #zope3-dev02:28
*** einheit_ has quit IRC02:32
*** zbir has quit IRC02:41
*** bradb_ has joined #zope3-dev02:51
*** bradb_ has quit IRC02:59
*** projekt01 has quit IRC03:05
*** bradb has quit IRC03:06
*** einheit_2 is now known as SteveA03:12
*** sm has quit IRC03:27
*** BjornT has quit IRC03:30
*** GaryPoster has joined #zope3-dev03:45
*** SteveA has quit IRC04:28
*** zbir has joined #zope3-dev04:39
*** stub has quit IRC04:39
*** Aiste has quit IRC05:10
*** __mac__ has quit IRC05:10
*** tiredbones has quit IRC05:10
*** vinsci has quit IRC05:10
*** Hellfried has quit IRC05:10
*** GaryPoster has quit IRC05:11
*** peaceman has quit IRC05:17
*** andres_ has quit IRC05:17
*** xenru has quit IRC05:17
*** bob2 has quit IRC05:17
*** peaceman has joined #zope3-dev05:29
*** MJ has quit IRC05:35
*** Aiste has joined #zope3-dev05:35
*** andres_ has joined #zope3-dev05:36
*** xenru has joined #zope3-dev05:36
*** bob2 has joined #zope3-dev05:36
*** MJ has joined #zope3-dev05:39
*** __mac__ has joined #zope3-dev05:42
*** tiredbones has joined #zope3-dev05:42
*** vinsci has joined #zope3-dev05:42
*** Hellfried has joined #zope3-dev05:43
*** __mac__ has quit IRC05:53
*** vinsci has quit IRC05:53
*** tiredbones has quit IRC05:53
*** vinsci has joined #zope3-dev05:53
*** __mac__ has joined #zope3-dev05:54
*** tiredbones has joined #zope3-dev05:54
*** newpers has joined #zope3-dev05:54
*** sm has joined #zope3-dev07:04
*** sm has quit IRC08:10
*** ignas has quit IRC08:24
*** zagy has quit IRC08:29
*** zagy has joined #zope3-dev08:34
*** newpers has quit IRC08:44
*** sashav has quit IRC08:48
*** dlk has joined #zope3-dev08:51
*** dlk has quit IRC08:52
*** dlk has joined #zope3-dev09:01
*** jinty has joined #zope3-dev10:17
*** sashav has joined #zope3-dev10:19
*** MJ has quit IRC10:37
jintyhoi srichter, more zpkg issues for z3 trunk from the schooltool nightly build...10:45
*** xenru has quit IRC10:54
*** ignas has joined #zope3-dev11:06
*** andres_ is now known as andres11:16
*** MJ has joined #zope3-dev11:24
*** mgedmin has joined #zope3-dev12:07
*** mkerrin has joined #zope3-dev12:14
*** efge has quit IRC12:14
*** fermigier has joined #zope3-dev12:24
*** philiKON has joined #zope3-dev12:53
*** xenru has joined #zope3-dev12:58
*** MrTopf has joined #zope3-dev13:10
*** efge has joined #zope3-dev13:28
__mac__hi, is it possible to use the catalog to query all objects which implement a given interface?14:29
VladDracit is14:31
*** Aiste has quit IRC14:31
VladDracI think14:32
VladDraccan't remember if I actually managed to do so14:32
*** Jim7J1AJH has quit IRC14:34
*** xenru|clone has joined #zope3-dev14:38
VladDracsorry can't really confirm that it works14:39
zagynot cool ;)14:39
VladDracbut you should be able to introspect what interfaces you implement and index them14:39
zagyprobably as dotted names, right?14:40
zagyI mean indexing the instances would be silly14:40
srichterphiliKON: so are you going to finish the zpkgtools work you started in a branch?14:40
*** zbir has quit IRC14:41
VladDraczagy don't know if that's silly14:41
VladDracas long as the catalog/index can index/compare them14:42
* zagy ponders14:42
zagywe'll try something14:48
zagythanks VladDrac :)14:48
VladDrac    def getType(self):14:52
VladDrac        """ return those interfaces that extend ICubicContent """14:52
VladDrac        return [inf.unwrap() for inf in providedBy(self.context) if inf.extends(ICubicContent)]14:52
VladDracthat's what I did I guess ;)14:52
VladDracwith the following index14:53
VladDrac          ('type', 'getType', ISearchableType, 1, FieldIndex),14:53
*** xenru has quit IRC14:53
VladDracso it does work, hurray :)14:55
zagywe'll try that, thank you :)14:57
*** alga has joined #zope3-dev14:57
*** Aiste has joined #zope3-dev15:01
*** MJ has quit IRC15:23
*** MJ has joined #zope3-dev15:25
*** bradb_ has joined #zope3-dev15:26
*** gumpa has joined #zope3-dev15:40
*** mgedmin has quit IRC15:44
__mac__VladDrac: from where do you import providedBy?15:44
*** alga has quit IRC15:49
*** zbir has joined #zope3-dev15:59
__mac__thx srichter16:05
*** SteveA has joined #zope3-dev16:05
__mac__VladDrac: your example almost worked: we had to use providedBy(self.context).interfaces() and inf.getName()16:06
VladDracmac the code I showed is a couple of months old16:10
VladDracdunno if anything in zope3 changed, but I assume it once worked for me this way as well :)16:10
VladDracwith indexing pure interfaces, not strings16:11
*** anguenot has joined #zope3-dev16:16
*** J1m_ has joined #zope3-dev16:22
*** benji has joined #zope3-dev16:24
*** d2m has joined #zope3-dev16:28
*** mgedmin has joined #zope3-dev16:31
*** Egon__ has joined #zope3-dev16:39
*** Egon__ has quit IRC16:42
*** stub has joined #zope3-dev16:46
*** BjornT has joined #zope3-dev16:48
*** mgedmin has quit IRC16:53
*** sm has joined #zope3-dev16:56
*** SteveA has quit IRC16:57
*** dlk has quit IRC17:07
*** sashav has quit IRC17:14
*** mgedmin has joined #zope3-dev17:18
*** SteveA has joined #zope3-dev17:19
*** d2m has quit IRC17:23
*** d2m_ has joined #zope3-dev17:23
*** d2m_ is now known as d2m17:23
*** andres_ has joined #zope3-dev17:28
*** zagy has quit IRC17:37
*** __mac__ has quit IRC17:43
*** andres has quit IRC17:46
*** Aiste has quit IRC18:12
*** andres_ is now known as andres18:20
*** strichter has joined #zope3-dev18:25
*** srichter has quit IRC18:27
*** strichter is now known as srichter18:28
*** ChanServ sets mode: +o srichter18:28
*** zagy has joined #zope3-dev18:28
*** zagy has quit IRC18:29
*** Aiste has joined #zope3-dev18:34
*** andres has quit IRC18:41
*** sashav has joined #zope3-dev18:46
*** gumpa has quit IRC18:48
*** tiredbones has quit IRC18:59
*** mkerrin has quit IRC19:00
*** mkerrin has joined #zope3-dev19:03
*** sashav has joined #zope3-dev19:12
*** sashav has quit IRC19:18
*** sashav has joined #zope3-dev19:19
*** peaceman has quit IRC19:21
*** agroszer has joined #zope3-dev19:25
* mgedmin stares in wonder at @apply in zope.testbrowser.browser19:26
srichteryeah, I am not sure I like this use of decorators either; it seems dubious19:27
srichtermuch more effort and less readability for using a new feature19:28
srichterI do like @property for read-only properties though19:28
*** mkerrin has quit IRC19:30
*** MJ has quit IRC19:31
*** MrTopf has quit IRC19:31
*** mkerrin has joined #zope3-dev19:33
benjisrichter, I'd be interested in discussing the use of decorators in testbrowser if you wish (or ignore you if you don't :)19:35
benjimgedmin, do you stare because of confusion or awe?  :)19:35
srichterbenji: I just realize that I would not prefer it this way19:36
srichterbenji: I accept your choice there, otherwise I would have changed it already ;-)19:36
mgedminbenji, I think it is either insanely great, or insanely crazy19:36
mgedminI'm not sure which :-)19:36
benjimgedmin, I'll take either19:36
tarekgreatly crazy19:36
J1m_The @apply hack is better than the alternative,19:36
* mgedmin just hooked up mechanize and zope.testbrowser.browser to functionally test a roundup instance19:36
srichterwell, I like the alternative a little bit better19:37
srichterI guess the advantage of @ apply is that it does not leave useless methods around in the class19:37
*** J1m_ has quit IRC19:38
*** J1m has joined #zope3-dev19:38
smnice, mgedmin19:39
benjimgedmin, I'm confused about what you have done; testbrowser already uses mechanize and can access arbitrary web sites (see over_the_wire.txt)19:39
mgedminall I have done is learned how to use what already exists19:41
*** Egon__ has joined #zope3-dev19:41
mgedmin(without depending on zope.schema/zope.interface/other chunks of zope)19:42
benjimgedmin, so you followed over_the_wire.txt?19:42
mgedminnot quite19:45
benjiwhy not?19:46
mgedminI needed a script to find a bunch of *.txt files, load them, muck with pythonpath to make mechanize available, start the roundup demo server in the background, run the doctests with unittest, then kill the background server19:46
mgedminso I wrote one19:46
mgedminand since I liked zope.testbrowser.browser.Browser better than mechanize.Browser (nice properties, etc.)19:47
mgedminso I had to figure out how to get that one without importing half of Zope 3 into our local project repository19:47
mgedmin(which was: comment out all mentions of interfaces)19:48
mgedmintherefore the reading of zope/testbrowser/ and goggling at @apply decorators19:48
mgedminbrightened up my day19:48
benjithere is a doctest that explains how to do that without touching the testbrowser code (testbrowser is useable without the rest of Zope 3)19:48
mgedminis it?  it imports zope.interface unconditionally19:49
benjinext time you need brightening I have more code you can ready19:49
mgedminI also liked the any() function ;-)19:49
mgedminis there an obfuscated Python code contest?19:49
benjiyep, see... wait for it.... over_the_wire.txt19:50
mgedmin(I saw it)19:50
mgedminI think we're misunderstanding each other19:50
mgedminby "without the rest of Zope 3" I meant "without having modules in my PYTHONPATH"19:51
mgedminnot "without using the Z3 publisher infrastructure directly"19:51
*** BjornT has quit IRC19:51
benjimgedmin, ok that makes more sense, I've been wanting to make a simple distribution with the required parts (zope.interfaces, etc.) included20:00
*** gumpa has joined #zope3-dev20:01
* mgedmin dreams about zpkgtools and python eggs converging and becoming automatically built Ubuntu packages20:01
srichterbenji: it's a matter of 15 mins, if DEPENDENCIES.cfg and SETUP.cfg are setup ;-)20:01
benjisrichter, 15 for someone who knows how to do it (which isn't me) :)20:02
srichtermgedmin: I think there will be some drive to create eggs using zpkgtools20:02
benjiI saw where someone already started on that, so that's good20:02
srichterbenji: just copy the ZopeInterface release directry, change some of the info and you should be done20:02
srichtermgedmin: isn't jinty also working on this type of stuff?20:03
* mgedmin doesn't know20:04
jintyin my dreams;) if I had the time20:04
*** BjornT has joined #zope3-dev20:05
*** anguenot has quit IRC20:14
*** mp has joined #zope3-dev20:29
*** BjornT has quit IRC20:33
*** bradb_ has quit IRC20:48
*** stub has quit IRC20:50
*** jinty has quit IRC20:57
*** ignas has quit IRC21:04
*** efge has quit IRC21:08
*** fermigier has quit IRC21:11
tareksrichter, ayt ?21:25
*** MJ has joined #zope3-dev21:30
*** mkerrin has quit IRC21:30
*** tiredbones has joined #zope3-dev21:32
srichtertarek: yep21:32
*** zagy has joined #zope3-dev21:37
*** SteveA has quit IRC21:37
*** stub has joined #zope3-dev21:40
*** mp has quit IRC21:41
*** mp has joined #zope3-dev21:41
*** SteveA has joined #zope3-dev21:41
tareksrichter, yes so i located the problem of #481, it's an issue with I need to dig it more, to learn how BuffedStream works but I have a workaround for
tarekcalling readlines() instead of read() works fine21:45
srichterok, so that;s a twisted bug, we should add it to their bug collector and use read() for now21:45
tarekread() always sends '' when using readExactly api21:46
agroszerare local utilities persistent?21:47
agroszerbut globals not?21:47
agroszerI'm asking, because I'm a little bit confused about how zope.wfmc and ecm.workflow are cooperating21:48
srichtersprry, I know nothing about ecm.workflow21:49
agroszerI'm also just trying to understand21:50
agroszerit does a "provideUtility(pd, IProcessDefinition,"21:50
agroszerso that should be a global utility21:51
*** mexiKON has joined #zope3-dev21:51
srichteryou do this to register the process definition as a utility (global that is(21:51
srichterI think some ZCML directive does this automatically21:51
agroszernow I'm wondering, because after I shut+restart Z3, the definition is still there21:52
agroszerlist(getAllUtilitiesRegisteredFor(IProcessDefinition)) returns them21:52
srichterwhere is the provideUtility call made?21:53
agroszerjust a sec21:53
tareksrichter, the problem is, the zope.publisher.xmlrpc doctest passes just fine with the bug (??) we have a problem there to properly simulate rpc calls21:53
srichterwell, note that functional tests do not use the server21:54
agroszerwill be a lengthy description...21:54
srichterthey send their stuff to the publisher directly21:54
*** xenru|clone is now known as xenru21:55
agroszerI have an Application class (something like schoolbell's), it implements IReadContainer, IPossibleSite, IAttributeAnnotatable21:55
tarekso i guess i'll just go for that fix, w/o adding a test21:56
agroszerthere is an event subscriber for .interfaces.IApplication
srichtertarek: yeah, in this case it is ok I think21:57
srichteragroszer: note that the provideUtility call above is usually done from ZCML21:57
agroszerin the subscriber method I convert the app object to site, then add the WorkflowUtility, which contains the ProcessDefinitionRegistry21:57
srichterI bet you one of the directives you have ssets it up already21:57
agroszerstill in the subscriber I do addProcessDefinitionFromXpdl('z:/ecmdemo.xpdl')21:58
srichterok, that is a local utility call21:58
agroszerI see21:59
agroszerso that's why it is persistent21:59
*** mgedmin has quit IRC22:00
agroszerwait a sec22:00
agroszerin zope.component.provideUtility it calls for getGlobalSiteManager()22:01
srichterdoes addProcessDefinitionFromXpdl call provideUtility? I doubt it22:02
agroszerbut it does22:03
srichterok, then I don;t know22:03
srichteryou have to ask one of the ECM developers22:03
srichterlike Jean-Marc or Roger22:03
agroszerin fact Julien wrote in a mail that I can't keep the process def's persistent...22:04
agroszerI'll reply him22:04
agroszeranyway, thanks for your help22:04
*** philiKON has quit IRC22:07
agroszeris there any way to un-provide a utility?22:09
srichternot globally22:10
srichterrestart should work22:10
srichterlocally you can delete the ZODB of course, to get a new slate22:11
agroszerI'm thinking about how to manage the process definitions... I mean add/modify/delete22:12
srichtervia ZCML22:12
srichteryou should not allow users to edit workflows22:12
agroszersorry, no way, it must be user defined22:13
srichterthen you are guaranteed that your system will not work for a day22:13
agroszerI cannot pre-define the workflows22:13
srichterin this case you have to write persistent workflow definitions from scratch22:14
srichter(including workflow definitions)22:14
srichter(including all the edit screens)22:14
srichterthis is a lot of work and you should really re-evaluate your requirements22:15
srichteryou could , of course, also allow people to just upload XPDL files to the database and have a subscriber register them globally at startup22:16
agroszerI know that it is hell lot, that's why I wanted to use ecm.workflow/zope.wfmc22:16
srichterhowever, I think this is a recipe for disaster22:16
srichterbecause you interact with the workflow via applications22:16
agroszeryes, this is a still a big hole in my plans22:16
srichterthose applications have to be written in Python and if you cannot predefine the workflwo, you can also only hardly define the applications22:17
agroszermy app is a document management app22:17
srichterI really do not know what ecm.workflow does, so I cannot say anything about them22:17
agroszerthat's why the workflow should be user defined22:18
srichterthe workflow in a given company is almost always well-defined22:18
srichterI disagree22:18
agroszerthey decide on site what to do, and there are not always gurus there to hack zcml22:18
srichtera workflow is a very complex data structure and there is no way the average Joe can modify this22:18
srichteryou will corrupt the entire application in no-time22:19
srichterediting ZCML is a lot easier than editing a workflow22:19
srichteryou could only provide a very limited set of operations for users22:20
srichterbut that means a lot of code development22:20
agroszersurely, for the first time they won't get the mercedes :-)22:20
srichterok, so you need to start writing local process definition utilities first22:21
srichterthen edit screens for them22:21
srichterand then specific edit screens for your use case22:21
agroszerin fact everything what's there in zope.wfmc22:22
agroszerbut instead of using (local) utilities I have to provide a sort of process registry22:22
agroszerwhich is of course persistently stored22:23
agroszersomething like that22:24
srichterno you need to make process definitions persistent22:24
srichterthe process definition registry is just the local utility resgistry22:24
agroszeryes, of course22:24
agroszermissed that22:24
agroszerdoes Roger and/or the guys from z3ecm come online here on IRC?22:26
srichterroger is projekt0122:27
srichterthe Nuxeo guys are resularly here as well22:27
tarekjulien and florent have left a while ago22:27
tarekjulien == anguenot, florent == efge22:28
agroszerI see, I'll try to catch them22:28
agroszerthank you very much22:28
srichteryou are welcome22:29
*** ballen has joined #zope3-dev22:32
mphow do I specify default field date to be now?22:36
srichterI think you have to do this manually22:37
*** agroszer has left #zope3-dev22:41
*** agroszer has joined #zope3-dev22:42
*** agroszer has left #zope3-dev22:46
*** SteveA has quit IRC22:50
*** SteveA has joined #zope3-dev22:51
*** BjornT has joined #zope3-dev22:51
*** agroszer has joined #zope3-dev22:54
*** agroszer has left #zope3-dev22:55
*** agroszer has joined #zope3-dev22:55
*** agroszer has quit IRC22:55
*** bradb_ has joined #zope3-dev22:57
*** Jim7J1AJH has joined #zope3-dev23:02
*** Egon__ has left #zope3-dev23:03
*** gumpa has quit IRC23:05
*** projekt01 has joined #zope3-dev23:27

Generated by 2.15.1 by Marius Gedminas - find it at!