IRC log of #zope3-dev for Friday, 2006-11-03

*** benji has quit IRC00:14
*** Londo has joined #zope3-dev00:16
*** J1m has quit IRC00:22
*** gumpa has quit IRC00:38
*** dobee has quit IRC00:41
*** yota has quit IRC00:59
*** jinty has quit IRC01:11
*** timte has quit IRC01:26
*** projekt01 has joined #zope3-dev01:27
*** marcin_ant has joined #zope3-dev01:29
*** marcin_ant has quit IRC01:30
*** ignas has joined #zope3-dev01:41
*** RaFromBRC is now known as RaFromBRC|away01:57
*** harobed_ has quit IRC02:01
*** whit has quit IRC02:30
*** rocky is now known as rocky|Zzz02:30
*** jinty has joined #zope3-dev02:38
*** philiKON has quit IRC02:59
*** RaFromBRC|away has quit IRC03:13
*** reco has joined #zope3-dev03:27
*** ignas has quit IRC03:57
*** ofer has joined #zope3-dev05:22
*** alecm has quit IRC05:22
*** ofer has quit IRC05:38
*** stub has joined #zope3-dev05:42
*** projekt01 has quit IRC06:17
*** reco has quit IRC06:31
*** jinty has quit IRC06:35
*** Theuni has quit IRC06:49
*** stub has quit IRC06:56
*** stub has joined #zope3-dev06:57
*** __mac__ has quit IRC06:58
*** stu1 has joined #zope3-dev07:28
*** ktwilight_ has joined #zope3-dev07:28
*** ktwilight has quit IRC07:43
*** stub has quit IRC07:46
*** dunny has quit IRC07:58
*** romanofski has quit IRC08:03
*** d2m has left #zope3-dev08:07
*** Theuni has joined #zope3-dev08:23
*** __mac__ has joined #zope3-dev08:25
*** eins has joined #zope3-dev08:25
*** yota has joined #zope3-dev08:35
*** BjornT has quit IRC08:42
*** d2m has joined #zope3-dev08:47
*** BjornT has joined #zope3-dev08:47
*** philiKON has joined #zope3-dev08:52
*** timte has joined #zope3-dev08:52
*** batlogg has joined #zope3-dev09:11
*** dlk has joined #zope3-dev09:21
*** dlk has left #zope3-dev09:22
*** hdima has joined #zope3-dev09:31
*** wrobel has joined #zope3-dev09:32
*** philiKON has quit IRC09:39
*** dlk has joined #zope3-dev09:39
*** romanofski has joined #zope3-dev09:39
*** flox has quit IRC09:45
*** kobold has joined #zope3-dev09:51
*** Aiste has quit IRC10:01
*** ofer has joined #zope3-dev10:04
*** harobed has joined #zope3-dev10:31
*** dobee has joined #zope3-dev10:32
*** ofer has quit IRC10:53
*** ktwilight_ is now known as ktwilight11:08
*** baijum has joined #zope3-dev11:20
*** MJ has joined #zope3-dev11:48
*** jinty has joined #zope3-dev11:58
*** baijum_ has joined #zope3-dev12:11
*** stu1 is now known as stub12:13
*** baijum has quit IRC12:20
*** Aiste has joined #zope3-dev12:42
*** mgedmin has joined #zope3-dev12:54
*** dobee has quit IRC13:01
*** Aiste has quit IRC13:01
*** Aiste has joined #zope3-dev13:02
*** mkerrin has joined #zope3-dev13:04
*** henri_ has joined #zope3-dev13:10
*** vlado has joined #zope3-dev13:13
*** kobold has quit IRC13:30
*** J1m has joined #zope3-dev13:44
*** batlogg has quit IRC14:14
*** ChrisW has joined #zope3-dev14:18
*** dobee has joined #zope3-dev14:19
*** xenru has quit IRC14:32
*** flox has joined #zope3-dev14:36
*** hazmat has quit IRC14:38
*** torkel_ has joined #zope3-dev14:40
*** stub has quit IRC14:41
*** dobee has quit IRC14:51
*** dobee has joined #zope3-dev14:55
*** flox has quit IRC14:56
*** baijum_ has quit IRC15:00
*** SteveA has quit IRC15:08
*** philiKON has joined #zope3-dev15:10
*** benji has joined #zope3-dev15:12
*** SteveA has joined #zope3-dev15:32
*** radix has left #zope3-dev15:43
*** mgedmin has quit IRC15:43
*** dlk has left #zope3-dev15:47
*** hazmat has joined #zope3-dev15:48
*** ChanServ sets mode: +o hazmat15:48
*** MJ has quit IRC15:51
*** hdima has quit IRC16:25
*** torkel_ has quit IRC16:32
*** vlado has quit IRC16:40
*** vlado_ has joined #zope3-dev16:40
*** ChrisW has left #zope3-dev16:46
*** __mac__ has quit IRC16:49
*** eins has quit IRC16:55
*** rocky|Zzz is now known as rocky16:59
*** whit has joined #zope3-dev17:07
*** whit has quit IRC17:08
*** whit has joined #zope3-dev17:12
*** azazel has joined #zope3-dev17:38
*** tonico is now known as tonico|away17:39
azazelhi all, i'm approaching now zope3 technologies as i'm porting some old code to plone2.5 and zope 2.9.5, i've already used some bits of them like views and adaptors, but now i'm facing with a more extensive adaptors usage snd i have some questions thart i hope you guys can /will answer17:41
philiKONazazel, just ask your questinos17:42
philiKONdon't ask questions about questions :)17:42
azazelyes:)... was just an intro:)17:43
philiKONno intros necessary :)17:43
azazelok so i have some fragments of a plone ui which are calculated in different way than the standard plone , For exemple i have declare a IPortletCalculator  which works in pair with a DefaultPortletCalulator that adapts (*, IMySite) to IPortletCalculator and supplies portlets data for the generic plone content. The main view MySite asks for an adaptor and get it17:49
philiKONazazel, you write too long lines17:51
philiKONplease split them up17:52
philiKONand perhaps ask specific questions17:52
azazelNow i  need to declare a more "specific" adaptor for just one interface so my question is if i declare that adaptor to be more specific  like for="IMyspecificInterface \n IMySite, what will be the behavior of zope, when IMySite ask for an adaptor for a content object which implements IMySpecificInterface?17:54
azazelwill i get both adaptor classes? or just the more specific one?17:55
benjiyou get the most specific adapter17:56
azazelmmm, so it does the right thing! nice to know!17:58
azazel:) thanks benji17:59
azazelboth the next question is: what happens when two adaptors are equally specific18:01
azazelbut it's enough for now:)18:02
benjithe most recently registered one wins18:02
* rocky schemes on how to ensure his adapters are always the most recently registered ones... and decides to prefix all of this package names with zzzzz just to be sure .... :)18:03
benjiactually, if you use ZCML, you'll get a conflict, but that's a ZCML thing not a component registry thing18:04
*** ChrisW has joined #zope3-dev18:04
ChrisWphiliKON: ping?18:04
azazelor i can use override.zcml to rule things a bit?18:04
benjiazazel: yep18:06
azazelgood, many thanks18:07
* azazel set up emacs to open .zcml with nxml-mode:)18:08
*** alecm has joined #zope3-dev18:09
*** vlado_ has quit IRC18:09
*** whit has quit IRC18:11
*** whit has joined #zope3-dev18:12
philiKONJ1m, ayt?18:38
*** ChrisW has left #zope3-dev18:38
philiKONso i brainstormed on "sites" a bit18:39
philiKONnowadays we usually call a site manager a component registry18:39
philiKONcoz that's what it does, it registeres (and looks up) components18:39
*** ChrisW has joined #zope3-dev18:39
philiKONso a site is just a place where we can have such a component registry18:39
ChrisW*sigh* how come trailing dots are no longer supported in zcml dotted paths?18:40
philiKONChrisW, use +18:40
philiKONpeople usually associate something else with the term "site", though18:40
philiKONthey think "website"18:40
philiKONe.g. like "CMF Site"18:40
philiKON"Plone SIte" etc.18:40
philiKONso, i think we need a better name for sites18:40
ChrisWI thought that was the zope 3 idea of a site too?18:40
philiKONz3's idea of sites is just a place where you can make local component registrations18:41
ChrisWhmm, well, it used to be ;-)18:41
philiKONthey are, as i called it in seattle during a dinner conversations, the quantum leaps of acquisition in z3 :)18:41
J1mz3's notion is pretty close to CMF Site.18:41
*** romanofski has quit IRC18:41
J1mwhich is probably why I used that terminology.18:41
philiKONthis might sound crazy18:42
philiKONbut at the seattle sprint18:42
philiKONgodefroid came to me with a propblem18:42
ChrisWJ1m: yup, except you can't successfully nest CMF Sites, although I've known people to try ;-)18:42
philiKONthey wanted a *view* to be notified of arbitrary events that occurred during the view's execution time18:42
J1mChrisW, I've seen it done.18:42
philiKONthe use case was ajax18:43
philiKONtheir ajax framework uses z3 events18:43
ChrisWclass=".SomePyFile+" don't work :-(18:43
ChrisWshould it?18:43
philiKONno idea. just spell it out :/18:43
philiKONanyways, their ajax framework works with standard object events18:43
ChrisWit works if I do class=".SomePyFile.SomePyFile"18:43
J1mChrisW, I don't remember what the trailing + was supposed to do.18:43
J1mso use that18:44
philiKONso when an object changes, e.g. the title, all the appropriate bits and pieces in the UI are refreshed18:44
philiKONit's a very neat idea18:44
philiKONso, what we came up with is to have local component registrations on the view18:44
ChrisWwell, the trailing . used to be so you didn't have to type lots of ".MyClass.MyClass"18:44
philiKONso the view is a site18:44
ChrisWanyone know why it's now verboten?18:44
philiKONthat obviously confused some people18:44
philiKONChrisW, well, you shouldn't be creating CamelCase modules in the first place18:44
philiKONChrisW, and i have no idea why it's not workign anymore18:45
*** dobee has quit IRC18:45
philiKONi think we converted it into a trailing + at some point18:45
ChrisWit's been explicitly removed, since the trailing . raises an error18:45
philiKONbut perhaps we dind't18:45
ChrisWbut it looks like the trailing + doesn't look like it's there18:45
ChrisWoh well, ne'ermind18:45
philiKONanyways, while perhaps crazy or not, views can very well be ISites and it made perfect sense for their use case18:46
J1mphiliKON, thinking about a view being s site is too much work. :)18:46
philiKONit was very little work, in fact :)18:46
J1mFor their use case, they needed to somehow have a registry of subscribers on a view (or a view class).18:47
J1mI suspect that there is a lighter way to do it.18:47
philiKONmy point is just that zope 2 allowed you to acquire things from anywhere18:47
J1mI also suspect that are lots of particulars in this story that I don't have time to know. :)18:48
philiKONin z3 you have to make things a site in order to be able to acquire from it18:48
J1manyway, pretending that this is a reasonable thing to do ....18:48
philiKONsites don't always have to be folders18:48
J1mwhere do you think this leads?18:48
philiKONa new understanding/terminology for sites18:48
J1mno, you can acquire from anything above you in the containment hoerarchy18:48
philiKONwell, i meant like "component architecture" acquisition18:49
J1mI'm not terribly fond of the word site.18:49
J1mRight, you can'rt acquire components from something that isn't a component registry.18:49
philiKONme neither18:49
philiKONor, to be exact:18:49
philiKONyou can't acquire components form something that doesn't *have* a component registry18:49
philiKONcurrently we call sucha  thing a site18:49
philiKONbut potentially anything could have a component registry18:50
J1malthough, given feedback recently, I'm not fond of changing things unless there is a *really* compelling reason to. :)18:50
philiKONnot just things people generally  understand as "sites"18:50
philiKONi realize that also18:50
*** henri_ has quit IRC18:50
philiKONi feel the same way18:50
philiKONanyways, i really have nowhere to go with this yet18:51
philiKONjust wanted to run my thoughts by you18:51
philiKONthe views-need-specific-subscribers thing isn't a terribly good example, i admit18:51
philiKONbut i can see the use in more things having their own registries18:52
J1mkeep in mind that registries are not lightweight.18:52
J1mThey can consume a fair bit of memory due to their caches.18:53
philiKONthat's a good point18:53
J1mIt might be possible in the future to allow registries to us eless memory by being a little more selective about what they will cache.18:54
J1mBut that will take some engineering.18:54
J1mAnd my list of engineering projects is long and my time to work on them is short.18:54
philiKONhehe, sure18:54
philiKONwell, i'm not saying we should encourage registry proliferation18:54
philiKONbut in grok, for example, we thought of a simple way to say: "make this view available there" where "there" is some object18:55
J1mI'm not saying it wouldn't be nice if we could. :)18:55
philiKONit just seems that the word "site" is misleading in many ways18:55
philiKONbecause that "there" doesn't really need to have the notion of a site that we typically do18:56
*** mgedmin has joined #zope3-dev18:56
philiKONanyways, we're agreeing mostly18:56
*** harobed has quit IRC19:03
*** hazmat has quit IRC19:05
*** ChrisW has quit IRC19:29
*** Aiste has quit IRC19:37
*** mgedmin has quit IRC19:38
*** RaFromBRC has joined #zope3-dev19:56
*** guruz has joined #zope3-dev20:04
guruzhow can i get line numbers etc when the TAL parser can't evaluate something and throws for example a KeyError?20:04
philiKONthe traceback should contain that info20:05
guruzit does not20:06
*** mkerrin has quit IRC20:12
*** whit has quit IRC20:19
*** hazmat has joined #zope3-dev20:24
*** ChanServ sets mode: +o hazmat20:24
*** whit has joined #zope3-dev20:29
*** whit has quit IRC20:31
*** whit has joined #zope3-dev20:32
*** whit has quit IRC20:34
*** batlogg has joined #zope3-dev20:45
*** whit has joined #zope3-dev20:49
*** projekt01 has joined #zope3-dev20:59
*** radix has joined #zope3-dev21:04
*** stub has joined #zope3-dev21:05
*** dunny has joined #zope3-dev21:06
*** ofer has joined #zope3-dev21:11
*** hazmat has quit IRC21:18
*** WebMaven has joined #zope3-dev21:20
timteIf I want prev and next available as browser pages, do I need on browser:page for each page or can I specify an interface and all its methods will be browser pages?21:32
philiKONyou need individual browser:page decl21:35
philiKONthoough you can save some typing with browser:pages21:35
timte<browser:pages interface="..." /> would be nice, or are there drawbacks?21:36
philiKONhow would you specify edit.html?21:38
philiKONor similar views21:38
philiKONi don't really see the advantage21:38
philiKONyou'd have to type the interface21:38
timteless typing21:38
philiKONso you can just aas well type the zcml21:38
timteI have an interface anyways21:39
philiKONeither way, interfaces typically document python apis21:39
philiKONnot everything in a browser view API has to be meant to be publishable21:39
timteLet's say you have a IMusicPlayer with stop, record, rewind, fast forward and all that stuff. I go to for stopping the music or have some ajax stuff that uses that url. That's many browser pages. I thought there might be a shortcut.21:42
*** projekt01 has quit IRC21:43
timteor is it like in zope 2 that an object's method can be triggered by the url without any browser pages?21:46
philiKONnot really21:46
philiKONyou probably want a <browser:view> with subpages21:47
philiKONother than that i'm afraid you'll have to type :(21:47
philiKONanyways, bbl21:47
radixSo I think I need to implement something like a HiddenSequence formlib widget. Has anyone done anything like that yet?21:47
timteok, thanks21:47
philiKONradix, hidden?21:48
philiKONradix, as in: widget.hidden() ?21:48
radixphiliKON: I need to put in a bunch of <input type="hidden" name="constant" value="variable" /> into some forms21:48
philiKONyes, i believe existing widgets can already to that21:48
philiKONthey have a hidden  method21:48
philiKONincl. the sequence widget21:48
philiKONgotta go, though21:49
radixhmm ok, I'll take a look. thanks.21:49
*** jukart has joined #zope3-dev22:23
*** jinty has quit IRC22:26
*** jukart has quit IRC22:34
*** jukart has joined #zope3-dev22:36
*** hazmat has joined #zope3-dev22:37
*** ChanServ sets mode: +o hazmat22:37
*** marcin_ant has joined #zope3-dev22:46
*** hazmat has quit IRC23:08
*** hazmat has joined #zope3-dev23:11
*** ChanServ sets mode: +o hazmat23:11
*** hazmat has quit IRC23:16
*** RaFromBRC is now known as RaFromBRC|lunch23:17
*** hazmat has joined #zope3-dev23:22
*** ChanServ sets mode: +o hazmat23:22
*** jukart has quit IRC23:23
*** nebo has joined #zope3-dev23:26
*** hazmat has quit IRC23:36
*** _tarek has joined #zope3-dev23:36
*** tarek has quit IRC23:41
*** nebo has quit IRC23:47
*** _tarek has quit IRC23:55
*** tarek has joined #zope3-dev23:56
*** BjornT has quit IRC23:58
*** BjornT has joined #zope3-dev23:59

Generated by 2.15.1 by Marius Gedminas - find it at!