IRC log of #zope3-dev for Monday, 2005-06-20

*** bskahan has joined #zope3-dev00:13
*** MiUlEr has joined #zope3-dev00:13
*** timte has quit IRC00:26
*** bskahan has quit IRC00:31
*** andrew_m has quit IRC01:08
J1msrichter, I wish you hadn't posted the windows tar ball.01:11
J1mPeople are trying to use it. :(01:12
projekt01J1m, of corse they do, windows is the most used OS ;-)01:13
*** bskahan has joined #zope3-dev01:13
J1mRight, but that isn't the release.01:13
J1mIt's raw input that Tim needs to generate the release.01:13
projekt01Ah, uhh I see, isn't this working with Tim's *.pyd files01:14
J1mno01:14
*** J1m has quit IRC01:15
*** projekt01 has quit IRC01:48
*** bskahan has quit IRC01:50
*** nturner has joined #zope3-dev02:18
*** yota has quit IRC02:41
*** deo has quit IRC02:47
*** deo has joined #zope3-dev02:47
*** stub has joined #zope3-dev02:58
*** bskahan has joined #zope3-dev03:23
*** MiUlEr has quit IRC03:51
*** RaFromBRC has joined #zope3-dev04:28
*** RaFromBR1 has joined #zope3-dev04:41
*** RaFromBRC has quit IRC04:53
*** bskahan has quit IRC05:01
*** stub has quit IRC06:51
*** SureshZ has left #zope3-dev07:39
*** RaFromBR1 has quit IRC07:59
*** stub has joined #zope3-dev08:29
*** Aiste has joined #zope3-dev09:04
*** guido_g has joined #zope3-dev09:20
*** timte has joined #zope3-dev09:34
*** yota has joined #zope3-dev09:37
*** Aiste has quit IRC09:37
*** sashav has joined #zope3-dev09:58
*** hdima has joined #zope3-dev10:03
*** __gotcha has quit IRC10:05
*** sashav has quit IRC10:14
*** sashav has joined #zope3-dev10:17
*** jhauser has joined #zope3-dev10:23
*** jhauser has joined #zope3-dev10:24
*** jhauser_ has joined #zope3-dev11:10
*** BjornT has quit IRC11:13
*** anguenot has left #zope3-dev11:14
*** jhauser has quit IRC11:17
*** BjornT has joined #zope3-dev11:19
*** philiKON has joined #zope3-dev11:23
*** BjornT has quit IRC11:24
*** BjornT has joined #zope3-dev11:34
*** lunatik has joined #zope3-dev11:51
*** bob2 has joined #zope3-dev11:51
*** philiKON has quit IRC11:51
*** lunatik has quit IRC12:01
*** apoirier has joined #zope3-dev12:01
*** andrew_m has joined #zope3-dev12:01
*** BjornT has quit IRC12:03
*** lunatik has joined #zope3-dev12:04
*** Aiste has joined #zope3-dev12:07
*** BjornT has joined #zope3-dev12:14
*** _drzoltron has joined #zope3-dev12:17
*** BjornT has quit IRC12:19
*** BjornT has joined #zope3-dev12:28
*** BjornT has quit IRC12:29
*** lunatik has left #zope3-dev12:58
*** anguenot has joined #zope3-dev13:09
*** anguenot has left #zope3-dev13:11
*** anguenot has joined #zope3-dev13:11
*** __gotcha has joined #zope3-dev13:13
*** J1m has joined #zope3-dev13:18
andrew_mdoes anyone know the trick for adding an entry to the "quick add" box of the boston skin? somehow it seems different from the rotterdam skin and only shows "hello world"13:38
*** zagy has quit IRC13:41
*** zagy has joined #zope3-dev13:47
*** debugger has joined #zope3-dev13:58
debuggerhi13:58
*** _drzoltron has quit IRC13:59
debuggerI just saw that zope 3.1 is out :)14:01
*** J1m has quit IRC14:04
*** bskahan has joined #zope3-dev14:07
*** BjornT has joined #zope3-dev14:11
*** philipp has joined #zope3-dev14:24
debuggeroh, zope3.1 doesn't seem to work on windows out of the box :(14:26
debuggerstrange... some files are named: .py0000644 :|14:27
*** projekt01 has joined #zope3-dev14:28
d2mdebugger: wait for the windows binary14:31
debuggerd2m, I've build it myself, but it doesn't seem to work.14:32
d2mah ok14:32
debuggerits just a: python install.py bdist_wininst14:33
srichterandrew_m: best ask projekt0114:33
srichterdebugger: I think you have to also run a fix script14:34
debuggerit even borked the imports of "zope".    the install named the directory "Zope", I've renamed it to "zope", and that problem was gone... but other appear and I'm stuck.14:34
srichterpython install.py bdist_wininst --install-script Zope-win-fixup14:34
debuggerah :D14:34
debuggerlemme try. :)14:34
srichterat least that's what Jim told me; I will put those instructions in the final release notes14:34
projekt01andrew_m, I just accessed the IRC, what do you like to ask?14:35
srichter<andrew_m> does anyone know the trick for adding an entry to the "quick add" box of the boston skin? somehow it seems different from the rotterdam skin and only shows "hello world"14:36
andrew_mprojekt01: i think it was only an authentication problem14:36
projekt01Really? It shouldn't, it should use the same IAdding14:36
projekt01Ah, Ok14:37
andrew_msometimes when i click a folder firefox kinda logs me out14:37
debuggersrichter, humm, that srcipt is inside the directory Zope-win, should I do a: python install.py bdist_wininst --install-script Zope-win\Zope-win-fixup.py ?14:38
debuggersrichter, oh, it doesn't seem to work :|14:39
projekt01andrew_m, really? What view do you get on this logout? @@contents.html?14:39
debugger"""error: install_script 'Zope-win\Zope-win-fixup.py' not found in scripts"""14:39
srichterI dunno14:39
srichterI have only tried the tar ball on Linux14:40
srichterI personally cannot support Windows, because I don't run it14:40
andrew_mprojekt01: that is ok (add menu full) but when i click a folder in the navigation tree i often end up as 'Unauthenticated User'14:40
debuggersrichter, the "Zope-win" directory has some other files, *.cfg, eg: SETUP.cfg;  what are those?14:41
andrew_mprojekt01: it looks like i have to click "login" for every folder or so..14:41
andrew_mprojekt01: maybe a firefox thing or a caching issue14:42
projekt01Hm, do you use the "zope.Manager" user? This user should have access on all objects14:42
projekt01Otherwise you can run into a unauthorized error if you expand the tree (and you don't have the permission for a child)14:42
debuggersrichter, oh, I've peeked inside the Zope-win-fixup.py, and the only thing it does is creating a batch file.  so, this will not fix my problems hehe14:43
andrew_mprojekt01: yes, i do. it also works, but if i log in for a folder, then click another folder it looks like i'm loged out again14:43
debuggeroh, well, I'll w8 for the windows bins.14:43
*** regebro has joined #zope3-dev14:44
projekt01Andrew_m, Oh, well caching could be a problem, but shouldn't. Did you set keep-alive to True?14:44
andrew_mprojekt01: simply clicking login again, logs me in again without asking for user/pw14:44
srichterdebugger: the cfg files are used by zpkgtools to create the tar ball14:44
projekt01andrew_m, Yes, this happens because the sessions knows about your credentials. I guess.14:44
srichterthis is a fairly sophisticated process actually, since we are not distributing the entire Zope3 source tree, but only a stable subset that we are willing to support14:45
srichterif someone sponsors a cheap Windows box, I'll certainly do Windows releases and test them too :-)14:46
andrew_mprojekt01: where can i check that keep-alive?14:47
debuggersrichter, hehe14:49
debuggersrichter, doesn't zop.eorg have a donation page or something like that?14:50
projekt01Andrew_m, type about:config in the address of your mozilla, you can see the config page then14:51
projekt01Type keep-alive into the Filter field14:52
andrew_mprojekt01: ah.. for the browser, yes that's set14:52
projekt01Hm, does this happen only on your content types or on default Zope3 objects like the file?14:53
andrew_mprojekt01: for the standard folders14:53
*** philiKON has joined #zope3-dev14:54
projekt01Hm, and you use the default user with the role zope.Manager, hm, I think also this is a caching problem.14:54
andrew_mprojekt01: yes, think so too. it actually only happens if the folders were visited before as unauthenticated user14:55
projekt01Perhaps there is a issue already on the bugtracker?14:56
projekt01If you can, it would be nice to re-install the firefox and if it is still a problem post it on the bugtracker.14:57
andrew_mprojekt01: ok, i'll try that14:57
projekt01Perhaps it's a bug and has to do with header infos from the server or some session issue with the PAU?14:58
projekt01Do you use a local PAU?14:58
andrew_mprojekt01: uh.. no idea. i didn't change anything with authentication14:59
projekt01Ok, do you use the z3 server on a local computer or is the z3 server on a different machine?15:00
andrew_mlocal15:00
srichterandrew_m: that's the problem; zope.org is not a non-profit org yet, so donating is not regulated15:17
srichterprojekt01: andrew_m: I think it is a browser caching problem, not a Zope one, since I have never observed that, but my caches are all disabled15:18
andrew_msrichter: i'm planning to disable the cache on server side if possible (using meta tags in the page template if necessary)15:20
andrew_msrichter: so the user doesn't have to reconfig his browser15:21
andrew_msrichter: disabling the cache in firefox did not help in my case, btw15:22
*** philipp has quit IRC15:31
*** philipp_ has joined #zope3-dev15:31
*** bigkevmcd has joined #zope3-dev15:32
*** MiUlEr has joined #zope3-dev15:34
*** zagy has quit IRC15:36
srichterfor all passive readers: Andrew and I have resolved the problem privately; when using basic HTTP auth, you are only authenticated for the folder and its subtree you are doing the authentication from15:39
*** zagy has joined #zope3-dev15:43
*** sashav has quit IRC15:55
*** MiUlEr has quit IRC15:56
*** hdima has quit IRC16:24
*** stub has quit IRC16:25
*** J1m has joined #zope3-dev16:26
regebro<In cartman voice> 3.1beta! Sweeeeeet.16:27
*** gintas has joined #zope3-dev16:39
*** bradb-afk is now known as bradb16:40
*** SureshZ has joined #zope3-dev16:40
*** bskahan has quit IRC16:41
*** niemeyer has joined #zope3-dev16:59
*** bigkevmcd has quit IRC17:19
*** philiKON has quit IRC17:21
*** lunatik has joined #zope3-dev17:22
*** guido_g has quit IRC17:27
*** tarek has joined #zope3-dev17:35
*** gsbain has joined #zope3-dev17:48
*** MiUlEr has joined #zope3-dev17:48
*** faassen has joined #zope3-dev17:57
VladDrachttp://opensource.amazesolutions.com/projects/zope3/cubic/ <- if anyone feels adventurous17:59
*** lunatik has left #zope3-dev17:59
J1minteresting18:00
*** bskahan has joined #zope3-dev18:04
VladDrac(I know it needs more tests, it needs alot more of everything :)18:07
VladDrac(just trying to release something before ep2005)18:07
J1m:)18:07
bob2hm, so, where can I find documentation on what's hiding inside "context"?18:18
bob2e.g. I only found out context.user existed by googling for "context user zope3"18:18
J1mnothing is hiding inside context.18:19
J1mYou can't count on anything.18:19
J1mWhat you can count on depends on the type of context.18:19
bob2hmm, ok18:19
bob2this is confusing enough that I'm not even sure how to frame my question18:21
*** lunatik has joined #zope3-dev18:21
*** lunatik has left #zope3-dev18:21
J1mhave you read the tutorial or one of the books?18:21
bob2yeah, I'm reading srichter's book18:22
bob2it doesn't seem to mention dealing with user-specific data at runtime, though18:22
bob2(for example)18:23
J1mI suggest going through the tutorial before reading a book.18:24
J1mI also suggest skipping over the testing material on a first read.18:24
*** lunatik has joined #zope3-dev18:25
bob2whee can I find the tutorial?18:25
*** lunatik has left #zope3-dev18:25
J1mhttp://www.zope.org/Wikis/DevSite/Projects/ComponentArchitecture/ProgrammerTutorial18:26
bob2ahh, thank you18:26
*** __gotcha has quit IRC18:30
*** th1a has joined #zope3-dev18:36
*** Aiste has quit IRC18:54
*** th1a is now known as th1a|unch19:00
*** zagy_ has joined #zope3-dev19:11
*** zagy has quit IRC19:13
*** th1a|unch has quit IRC19:13
*** apoirier has quit IRC19:15
*** SureshZ has quit IRC19:22
*** niemeyer has quit IRC19:24
*** MiUlEr has joined #zope3-dev19:27
*** philipp_ has quit IRC19:28
*** niemeyer has joined #zope3-dev19:30
*** philiKON has joined #zope3-dev19:41
*** projekt01 has quit IRC19:43
*** BjornT_ has joined #zope3-dev19:48
*** faassen has quit IRC19:48
*** yota has quit IRC19:51
*** BjornT has quit IRC19:57
*** BjornT has joined #zope3-dev20:01
*** th1a has joined #zope3-dev20:01
*** gintas has quit IRC20:11
*** BjornT_ has quit IRC20:15
*** bradb is now known as bradb-lunch20:19
*** SureshZ has joined #zope3-dev20:22
*** projekt01 has joined #zope3-dev20:26
*** yota has joined #zope3-dev20:27
*** efge has joined #zope3-dev20:31
SteveAJ1m: do you recall why we don't use action="" in forms?20:36
SteveAa colleague just pointed out section 4.2 of http://www.ietf.org/rfc/rfc2396.txt20:36
SteveA4.2. Same-document References20:36
SteveA   A URI reference that does not contain a URI is a reference to the20:36
SteveA   current document.  In other words, an empty URI reference within a20:36
SteveA   document is interpreted as a reference to the start of that document,20:36
SteveA   and a reference containing only a fragment identifier is a reference20:36
SteveA   to the identified fragment of that document.  Traversal of such a20:36
SteveA   reference should not result in an additional retrieval action.20:36
SteveA   However, if the URI reference occurs in a context that is always20:37
SteveA   intended to result in a new request, as in the case of HTML's FORM20:37
SteveA   element, then an empty URI reference represents the base URI of the20:37
SteveA   current document and should be replaced by that URI when transformed20:37
SteveA   into a request.20:37
*** th1a has left #zope3-dev20:37
J1mwe never decided not to use action="" in forms.20:40
J1mI don't think we ever considered it.20:40
SteveAi don't know if it works in real browsers20:41
efgeHi20:42
efgeI'd like to hook into the interface adaptation mechanism to have some kind of "catchall" adapter that would know to what desired interface it's trying to be adapted. Is there a way to do that ?20:43
efgeSo this factory would not only receive a context but also an interface.20:44
SteveAwalk up the stack...20:44
SteveAefge: i've wanted this too a couple of times.20:44
efgeSteveA: yep that was what I was hoping to avoid :)20:44
SteveAnot for a catch-all adapter, though20:44
SteveAi think because it was for some particular niche, i just did some niche thing for it20:45
SteveAyou could do so by using the adapter registry directly i expect20:45
efgeI expect to use this for proxy-base repositories, so the proxies in content space would have a special adapter that decides how to redirect20:45
SteveAbut you'd need your own getAdapter replacement20:46
SteveAi've also wanted an adapter that knows its own name20:46
efgeyes I can see the need for that too20:46
SteveAwhich is useful when you think about using things other than strings as names20:46
efgeis any part of the adapters coded in C ?20:47
SteveAsome optimisations20:47
SteveAwhich can be removed, but i don't know if the python versions are tested20:47
efgeif not I can look at extending the registry with flags for such special adapters20:47
efgeI'll walk the stack to prototype, first20:48
*** bradb-lunch is now known as bradb20:48
efgethanks20:49
J1mI suggest exploring other alternatives.20:49
SteveAyou mean alternative designs?20:50
J1mProxies are evil20:50
J1mThey are more evil the more powerful they are.20:50
J1myes20:50
efgeJ1m: alternatives to proxy-based repositories you mean ?20:50
J1mIf you use a proxy design you need to make deep changes to the component architecture, then I'd think hard about the proxy design.20:51
J1mefge, possibly20:51
efgeI see your point, but OTOH having slightly more intelligent adapters is not really a fundamental change20:51
J1mI don't find "proxy-based repository" to be a bit vague.20:51
SteveAJ1m: do you think there's a case for having adapters that can know what question they are given as an answer to?20:51
J1mI do find "proxy-based repository" to be a bit vague.20:52
J1mSteveA, we don't need anything like this for methods.20:52
efgeJ1m: I'll detail that in my presentation for EuroPython, I was being terse here20:52
J1mMakes me wonder if we really need it for adapters.20:52
J1mefge, I think I'm somewhat familiar with CPS's respository approach, which is what I think you are trying to port.20:53
J1mI need to think harder about it.20:53
efgeJ1m: yes, port or find the best way to do in a Z3 setting20:53
J1mThe fact that you want adapters like this makes alarms go off in my head. :)20:54
efgeIt may very well be a dead-end or too complex, but I want to try them and see by myself the complexity of it20:54
J1mI'd really like to have some time to discuss your repository based approach with you some time.20:56
J1mI'd like to brainstorm some twists on your approach that might ne a bit simpler.20:56
J1mI'd like to brainstorm some twists on your approach that might be a bit simpler.20:57
J1mas far as adapters knowing more about how they are called, we have some precedent for this.20:57
J1mWe sometimes set veiw __name__ attributes, I believe.20:57
*** bcsaller has joined #zope3-dev20:58
J1mI suppose that it could be argued that if an adapter factory is an object that implements a certain interface, then the adapter machinery might set some attributes on it.20:58
J1mBut that would be after the factory is called.20:58
J1mSimilarly, I suppose you could look for special meta data on the factory.20:59
efgein my case I'd need it before the call, yes20:59
J1mOf course, all of this will tend to slow things down.20:59
J1mand I want to make adapters go faster, not slower.21:00
efgeyes there's the speed problem21:00
efgeAs for the brainstorm, too bad you can't be at EP, but I hope we'll see each other around soon21:00
J1mme too21:00
srichteryou can always brainstorm now :-)21:01
J1mToo hard over irc21:01
efgewell face to face is so much faster ... :)21:01
srichterwe shoudl have a sprint in the US before the summer is over21:02
J1mIf we had a zen of zope, like the zen of python, it would have a stanza like: "Proxies are evil.  Sometimes necessary, but to be avoided when possible."21:03
efgeI agree in general that the more magic there is, the more evil it is, and that proxies are often quite magic21:04
bob2hm, what bits of zope should I be looking at if I want a form that affects multiple objects?21:05
SteveAthe experimental zc.page21:05
bob2e.g. a table with a checkbox for each row, each row showing some info for an object21:05
SteveAor the ObjectField (is that what it's called?)21:06
*** regebro has quit IRC21:06
bob2thanks again, SteveA21:06
efgehmm ztapi.provideAdapter and zope.component.provideAdapter have a different order of arguments21:11
bob2hrm, a little discouraging that ObjectField is only mentioned in the zopex3 3.0.0 pot files21:15
bob2InterfaceField, perhaps?21:16
bob2ah, "object fields"21:17
*** efge has quit IRC21:18
*** RaFromBRC has joined #zope3-dev21:27
bob2hm, does anyone know of any example code using this technique?21:37
*** SureshZ has quit IRC21:46
srichterI would suggest using only the widget API and write the view from scratch21:47
*** bradb has quit IRC21:51
J1mI'd like to phase out ztapi21:53
*** lunatik has joined #zope3-dev21:53
J1mthe provide functions in zope.component should be used rather than the ztapi calls.21:54
*** lunatik has left #zope3-dev21:57
*** RaFromBRC is now known as RaFromBRC|afk22:00
*** bradb has joined #zope3-dev22:08
*** anguenot has left #zope3-dev22:12
*** nederhoed has joined #zope3-dev22:25
VladDrachoedje22:28
VladDracnederhoed: check out opensource.amazesolutions.com/projects/zope3/cubic22:28
srichterJ1m: wanna phase out ztapi for 3.1? :-)22:28
srichterVladDrac: do you have a demo online?22:29
*** niemeyer has quit IRC22:29
*** niemeyer has joined #zope3-dev22:31
J1msrichter, no22:32
srichterVladDrac: suggestion: your dcmap should be an adapter or even better have a special DCProperty, like you have FieldProperty properties22:32
srichterJ1m: ok22:32
*** RaFromBRC|afk is now known as RaFromBRC22:33
J1mwe need to get 3.1 out. Feature freeze for 3.2 in a little over 4 months. :)22:33
srichterhow much do we plan for 3.222:34
srichterthe big changes so far seem to be twisted integration and lxml22:35
srichteranything else?22:35
philiKONlxml is going to make it into z3?22:35
srichteryes22:35
J1mI certanly hope so22:35
J1mI certainly hope so22:35
philiKONyay22:35
philiKONthis is great22:35
J1mWe'll have whatever we have Nov 1.22:35
J1mIt depends on what people actually do between now and then.22:36
philiKONi've started using it, expecting to make bug reports any minute now ;)22:36
J1mOf course, new features actually need to be mature long before Nov 1.22:36
srichterdo you expect wfmc to mature enough?22:37
J1mNew features really should be complete by Oct 1.22:37
srichterwhat about zc.page?22:37
J1mI'm not sure if wfmc is of wide enough interest to be included.22:37
J1mI would hope that zc page (probably renamed) in 3.2.22:38
srichterok22:38
srichtersounds good22:38
nederhoedah I will22:38
J1mSomething I'd really love to do with 3.2 is to start defining a small core.22:42
J1mA truly minimal app server and make almost everything else optional.22:42
VladDracsrichter: z3alot (.org) is supposed to become the demo22:43
VladDracsrichter I'm prematurely releasing stuff so people can have a look before Ep200522:44
VladDracsrichter I'll write down your remark and look at it later - tx22:44
srichterJ1m: that would be basically the CA + ZODB + transactions22:54
srichterJ1m: not even the publisher would be part of this22:54
SteveAsrichter: i'd hope the zodb wouldn't be part of the core23:00
SteveACA + transactions + security23:00
J1mI suppose it depends on your perspective.23:05
J1mI'm not sure that *anything* is core universally.23:06
J1mI would say that both the publisher and ZODB are core to the Zope Web Application server.23:06
*** SureshZ has joined #zope3-dev23:34
*** bskahan has quit IRC23:46
*** anguenot has joined #zope3-dev23:56

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