IRC log of #zope3-dev for Monday, 2005-12-26

*** projekt01 has joined #zope3-dev00:20
*** projekt01 has quit IRC00:53
*** jenner has quit IRC01:17
*** dobee has quit IRC02:09
*** dobee has joined #zope3-dev02:10
*** jinty has joined #zope3-dev02:59
*** yota has quit IRC03:02
*** romanofski is now known as rom|zzZZ03:09
*** stub has joined #zope3-dev03:15
*** stub has quit IRC03:17
*** stub has joined #zope3-dev03:31
*** dobee has quit IRC03:39
*** J1m has joined #zope3-dev04:06
*** J1m has quit IRC04:12
*** jinty has quit IRC04:13
*** tahara has joined #zope3-dev08:30
*** yota has joined #zope3-dev08:37
*** VladDrac has joined #zope3-dev08:41
*** dobee has joined #zope3-dev09:34
*** dobee has joined #zope3-dev09:36
*** natea has quit IRC10:16
*** SteveA has quit IRC11:49
*** rom|zzZZ is now known as romanofski11:50
*** romanofski has quit IRC12:15
*** dobee has quit IRC12:18
*** dobee has joined #zope3-dev12:19
*** dobee has quit IRC12:33
*** BjornT has joined #zope3-dev12:36
*** dobee has joined #zope3-dev12:37
*** efge has joined #zope3-dev12:56
*** philiKON has joined #zope3-dev13:36
yotahum, good question on zope3 users13:42
*** jenner has joined #zope3-dev13:54
*** dobee has quit IRC14:23
*** jack-e has joined #zope3-dev14:46
*** dobee has joined #zope3-dev14:58
*** ruda_porto has joined #zope3-dev15:12
*** dobee has quit IRC15:16
*** Aiste has quit IRC15:17
*** dobee has joined #zope3-dev15:55
*** alga has joined #zope3-dev16:15
jennerHm, can I remove an object withouth __parent__ somehow? zope/app/container/ tries to object.__parent__ which doesn't exists, so I get an AttributeError and thus cannot remove the object anymore16:17
* jenner replaces the Data.fs with a clean one16:19
efgejenner: an object without __parent__ is not "contained", so you shouldn't use the container framework to remove it16:27
jennerefge: I just tried to remove a broken instance of an IContainer from ZMI16:28
efgeah it's broken... hm I guess that use case has been forgotten :)16:28
jennercan I set log level to debug somehow? It seems that an exception from one of my views is swallowed and I get a blank page16:31
*** drzoltron_ has joined #zope3-dev16:56
*** dobee has quit IRC17:10
jenner*sigh*, setting up like absolute_url as adapter for a cusomt interface is a PITA17:24
drzoltron_pita ?
jennerI have some types implementing the same base interface, now I'd like to display breadcrumbs (/folder1/folder2/item) of those types, but the default adapter (app.traversion.browser.absoluteurl.AbsoluteURL) adds only the __name__ of the object as info and I'd need a bit more17:37
jennerSo I thought I could just do the same as AbsoluteURL, but apparently I either fail to register the adpater properly or something's wrong with my types - if I register the adpater only for my.custom.Interface I get a ComponentLookupError, if I register them for "*" I get AttributeError's for __parent__17:39
*** tiredbones has left #zope3-dev17:45
*** stub has quit IRC17:58
*** drzoltron_ has quit IRC18:37
*** tiredbones has joined #zope3-dev18:54
*** danfairs has joined #zope3-dev19:13
* jenner has got his first ZODB conflict error with z319:14
*** alga has quit IRC19:15
danfairsjenner: happy christmas! :)19:15
*** andrew_m has quit IRC19:15
* danfairs has just got his very first content item working in z319:15
jennerdanfairs: :))19:16
* jenner is slowly approaching towards a full blown photo gallery19:16
jennerflickr style baby19:16
danfairsI'm not looking forward to debugging zcml though :S19:16
danfairsah well, the z3 schema stuff sure knocks Archetypes into a cocked hat ;)19:17
jennerdanfairs: I found out that zcml is not that bad after all, just don't put everything in one big configure.zcml, use <include package=.. /> for your browser and probably browser/skin packages and if you've defined your own roles/permissions - put them in security.zcml and include it using <include file=.. />19:19
danfairsjenner: ok, good tip. I've got Philipp's book (which I cunningly lent to a colleague before xmas) which I've read19:20
danfairsI'm currently referring to Stephan's book online19:20
danfairsjust got to the bit you mention ;)19:20
danfairsthe reason I mentioned it was this:
danfairsi got a little scared reading it. 80 declarations, eek!19:22
jennerdanfairs: I've learned from Stephans book (online edition on, though the most enlightenment came from reading the API docs, sources and last but not least - bugging people on #zope3-dev :)19:22
danfairsjenner: well here I am ;)19:22
danfairsyeah, reading source is how I learned Zope 2 and CMF... no stranger to that.19:22
danfairsunfortunately my day job is pure zope 2.19:23
danfairsLearning Zope 3 is a spare time thing right now.19:24
jennersame here19:24
jenneractually I should be writing my diploma thesis... coding java :)19:24
danfairsheh, could you get away with jython? ;)19:24
danfairs"...but it runs on the VM!"19:24
jennerNah, java rocks for bigger SOA stuff, especially the Spring framework, so I'm sticking to native java code :)19:25
jennerAnd 1.5 (aka tiger) is pretty nifty too19:26
jennerI just need some time to switch from javas type-safety to pythons "do-whatever-you-like" :)19:27
* danfairs hasn't touched Java since uni19:27
danfairsyeah, I can see the need for a 'python brain' and a 'java brain'19:27
danfairsthe next language I encounter is likely to be C# actually. So Java again ;)19:28
danfairsone of my colleagues like to freak the c# programmers out by demonstrating python code that changes a class' base classes on the fly.19:29
danfairshe's not allowed near the production code with that of course ;)19:29
jennerhe should freak out lisp or perl programmers by demonstrating the speed of pythons list comprehension and string operations :)19:30
danfairsheh, we don't really have any lisp or perl programmers - we're mostly a windows shop, hence the c#19:30
* danfairs has edited .emacs and doesn't want to go any closer to lisp than that19:31
danfairsof course, the windows people will then point to the benchmarks of IronPython vs. CPython ;)19:32
jennerNever heard of IronPython until now (just looked it up on google)19:33
philiKONoh, it's cool19:33
philiKONpython interpreter on top of .NET19:34
philiKONlike jython, just .NET19:34
jennerYeah, just read then19:34
philiKONi'm more excited about pypy, though19:34
jennerThat will most likely mean, that it will lag behind "real" python19:34
philiKONif someone wrote a C# backend to that...19:34
* danfairs googles pypy19:34
jennerpython interpreter which interprets python :919:34
philiKONthat's also what cpython is19:35
philiKONpypy is an implementation of the python interpreter in python itself19:35
jennerthat's what I meant :)19:35
philiKONthe nice thing is that you can use it to translate python code into an abstract object flow graph19:36
philiKONand from there to basically anything19:36
philiKONfor example C19:36
philiKONso, pypy can translate itself to C19:36
philiKONso you end up with a C-based python interpreter again19:36
philiKONbut it's source is pypy, which is python and hence much nicer to develop19:36
philiKONand you could also write a backend that generated Java or C#19:36
philiKONhence you'd have the equivalent of jython and ironpython19:37
jennerphiliKON: btw, in case you come to refactoring the code - wouldn't it be cool to ask the object for some info first (like IZopeDublinCore(context).title) and if that fails trying to getattr(context, '__name__', None)?19:38
philiKONwhat has dublin core got to do with urls???19:38
jennerjust as an example, it should rather be like (non-existent) IContextInfo19:39
philiKONzapi.getName() uses that19:39
philiKONperhaps it should use that19:39
philiKONit=the absolute url view19:39
jenneryes, but that doesn't offer any meta infos, if there would be an interface _just_ for the case you'd like to have a title or something else in your breadcrumbs, you could implement that interfaces19:40
philiKONmeta infos?19:41
jennertitle = meta inf19:41
philiKONwe use the term metadata19:41
jennerok, metadata then19:41
philiKONand IZopeDublinCore is an example of metadata19:41
philiKONbut it has little to do with urls19:41
jennerthe point why I'm sying this is, I had to reimplement IAbsoluteUrl for my own content types, just so I'm able to show the title of an item/folder in the breadcrumbs19:42
danfairshm, Zope3Book still talks about __used_for__ in some examples...19:43
philiKONdanfairs, yes, zope3book isn't updated, unfortunately19:43
jennerif AbsoluteURL.breadcrumbs() would query the object for an abstract interface instead of just getting its __name__, I could simply implement the interface instead of creating a new adapter19:43
danfairsshame :/19:43
philiKONjenner, i see.. well, first, your reimplementation might break the IAbsoluteURL contract...19:44
danfairsoh - great job on your book btw, very readable and understandable19:44
philiKONdanfairs, thanks19:44
philiKONdanfairs, there will definitely be updates there19:44
danfairsphiliKON: that's great news ;)19:44
jennerphiliKON: how so? it's registered only for="my.custom.IType"19:44
jenneroh, sorry19:44
jenneryou mean the interface to be queried19:44
philiKONjenner, well, if i get the IAbsoluteURL view, i expect view.breadcrumbs() to return names, not titles...19:45
philiKONi would say19:45
philiKONperhaps not19:45
philiKONdunno what IAbsoluteURL says :)19:45
philiKONjenner, it might be feasible to extend the IAbsoluteURL interface and the views to allow them to optionally include IZopeDublinCore(context).title for convenience19:46
jenner"Name is the name to display for that segment of the breadcrumbs."19:46
philiKONhrm19:46 can be pretty ugly sometimes :)19:47
*** jukart has joined #zope3-dev19:47
*** jukart has left #zope3-dev19:49
jennerAnyway, I learned (again) a lot while reimplementing IAbsoluteURL about adapters, so I shouldn't complain :)19:50
philiKONyes, you should :)19:50
jennerI will :)19:54
* jenner waves19:54
*** jenner has quit IRC19:54
*** danfairs is now known as danfairs|away20:02
*** natea has joined #zope3-dev20:14
*** alga has joined #zope3-dev20:17
*** philiKON has quit IRC20:29
*** danfairs|away is now known as danfairs20:32
*** danfairs is now known as danfairs|away20:34
*** danfairs|away is now known as danfairs20:46
danfairshi again. I've got a custom Task object which is contained within a TaskContainer.21:05
danfairsTasks themselves can contain other tasks21:05
danfairsIn order to successfully view the 'Contents' tab for a task object, I've had to make security declarations in configure.zcml for attributes __len__ and items.21:06
danfairsThis doesn't feel right to me ;)21:06
danfairsWhat am I doing wrong? Should I define these methods in ITask so that the security assertion based on the interface 'takes'?21:06
danfairsThat also doesn't feel very clean...21:06
*** efge has quit IRC21:29
*** danfairs is now known as danfairs|imbibin21:41
*** regebro has joined #zope3-dev21:51
regebroMerry Xmas!21:52
*** alga has quit IRC21:55
regebroWas there any skin/layers changes between 3.0 and 3.2?21:56
regebroCMFonFive / CMFs skin registration no longer works under 2.9... :-/21:56
*** alga has joined #zope3-dev21:59
*** danfairs|imbibin is now known as danfairs22:01
dman13regebro: yes.  there was a thread recently about how multi-adapter lookup was fixed and that caused the priority of certain registrations to change.  (there's a thread on the mailing list about it)22:09
*** alga has quit IRC22:18
*** BjornT has quit IRC22:33
*** danfairs has quit IRC22:45
*** danfairs has joined #zope3-dev22:45
*** alga has joined #zope3-dev23:04
*** jack-e has quit IRC23:06
*** danfairs has quit IRC23:11
regebrodman13: Thanks, I'll check it out,23:38

Generated by 2.15.1 by Marius Gedminas - find it at!