IRC log of #zope3-dev for Tuesday, 2005-12-06

*** sashav has quit IRC00:00
*** sashav has joined #zope3-dev00:01
*** sashav has quit IRC00:02
J1manyone wanna test the source release?00:22
lifelessJ1m: no, but I have a small white paper I'd like you to read ;000:25
J1muh, ok00:26
lifelessits on testing, and test suite speed etc00:26
J1mThe requested URL /~robertc/interfaceverfication.txt was not found on this server.00:27
lifelessmissing i00:28
*** jfmoxley has joined #zope3-dev00:31
*** J1m has quit IRC01:03
*** niemeyer has quit IRC01:06
*** jhauser has quit IRC01:18
*** z3noob has joined #zope3-dev01:34
z3noobhi, new five user here... got a basic interfaces question:  how do i have my code ask for an implementation of an interface and not care which implementation it gets? or to put it another way - where do I put the responsibility for creating some actual instance that implements my interface?01:39
z3noobsome kind of factory?01:40
lifelessgetUtility is a call that can perform service location by interface01:40
z3noobi'll look that up01:40
z3noobis there a better channel for newbie stuff? or is this it?01:41
z3noobhmm i don't think utilities are what i want - these objects i'm working with are typically used only for one request, then thrown away01:47
z3nooband i don't care about their name01:48
*** _anguenot has joined #zope3-dev02:03
*** MiUlEr has joined #zope3-dev02:03
z3noobcan i register an adapter for python builtin types?  e.g. given a string, adapt it to my interface02:05
dman13I would think you should be able to do that02:11
*** dman13 has quit IRC02:11
z3noobhmm i would think so too, but i note he didn't say how :-\02:16
z3noobi tried using <adapter for="str" but doesn't seem to work02:22
z3noobi guess the value of "for" must be an interface02:23
z3nooball the examples in docs do that, anyway02:24
z3noobah, my zcml is borken anyway ;-)02:30
WebMavensound like a bad movie: 'Zope3: The Borkening'02:34
z3noobok my zcml is no longer borken... at least i get no errors on startup... but in test code i can't adapt a string to my interface.  rats.02:36
*** tiredbones has joined #zope3-dev02:41
z3nooboh well, try again later.02:45
*** z3noob has left #zope3-dev02:45
*** tiredbones has quit IRC02:47
*** yota has quit IRC02:57
*** sashav has joined #zope3-dev03:06
*** tiredbones has joined #zope3-dev03:43
*** deo has quit IRC03:55
*** stub has joined #zope3-dev04:01
tiredbonesIf I ever understand zope 3 technology, I will be blessed.It will be awhile before this puzzle is complete.04:04
*** hazmat has joined #zope3-dev04:22
*** ChanServ sets mode: +o hazmat04:22
*** natea|calsprint has quit IRC04:26
*** MiUlEr has quit IRC05:05
*** dobee has joined #zope3-dev07:10
*** Volpe has joined #zope3-dev07:20
VolpeHi, I was wondering if anyone could tell me how to define roles on Principles stored in a BTreePrincipleSource.  I don't quite get it... :\07:21
VolpeOr mention the source file that would give me a hint. :\07:22
*** dobee has quit IRC07:30
VolpeNo body around? or no body know?07:34
*** dobee has joined #zope3-dev07:45
*** d2m has joined #zope3-dev07:47
VolpeCan you grant permissions outside of zcml? :\07:48
lifelesszcml is just a config language, its all python at the code07:52
VolpeI'm just figuring that out now.07:52
VolpeWhere's the api for this stuff? :\07:53
Volpeor is it that ++apidoc++ thing.07:53
VolpeAh I think I get it, the role and permission stuff are just zope utilities.08:01
VolpeThat are initialized from zcml.08:02
*** WebMaven has quit IRC08:26
*** zagy has joined #zope3-dev08:47
*** eins has joined #zope3-dev09:04
*** sashav has quit IRC09:07
*** jhauser has joined #zope3-dev09:10
*** rom|zZZ has quit IRC09:11
*** agroszer has joined #zope3-dev09:13
*** hdima has joined #zope3-dev09:25
*** stu1 has joined #zope3-dev09:28
*** hazmat has quit IRC09:31
*** romanofski has joined #zope3-dev09:42
*** stub has quit IRC09:45
*** hazmat has joined #zope3-dev09:47
*** Volpe has quit IRC09:54
*** Theuni has quit IRC09:55
*** WebMaven has joined #zope3-dev09:56
*** stu1 has quit IRC10:04
*** WebMaven has quit IRC10:31
*** WebMaven has joined #zope3-dev10:32
*** sashav has joined #zope3-dev10:33
*** hazmat has quit IRC10:40
*** hazmat has joined #zope3-dev10:43
*** WebMaven has quit IRC11:04
*** WebMaven has joined #zope3-dev11:05
*** WebMaven_ has joined #zope3-dev11:15
*** stub has joined #zope3-dev11:29
*** WebMaven has quit IRC11:37
*** ignas has joined #zope3-dev11:55
romanofskidoes something similar exist like verbose security or is there a way to track, why some permissions seems not to work?11:59
romanofskia ++debug++security or something would be neat11:59
*** Aiste has quit IRC12:16
*** ignas is now known as parsifal12:24
*** parsifal is now known as ignas12:24
*** Aiste has joined #zope3-dev12:35
*** philiKON has joined #zope3-dev12:37
*** efge has joined #zope3-dev12:38
*** andres has joined #zope3-dev12:38
*** mgedmin has joined #zope3-dev12:38
*** niemeyer has joined #zope3-dev12:50
*** Hellfried has quit IRC12:56
*** Hellfrie2 has joined #zope3-dev12:56
*** Hellfried has joined #zope3-dev12:57
*** projekt01 has joined #zope3-dev12:58
projekt01srichter, ayt?12:58
andresprojekt01, btw, your calendar widget is working. I didnt include the css which was necessary for it to work.13:06
projekt01andres, Ah, cool, do you use it in a project?13:09
andresBut just an inhouse project.13:10
projekt01cool, if I find out how to package all this libraries with zpkg, I will release it.13:12
*** yota has joined #zope3-dev13:29
*** markus_ has joined #zope3-dev13:32
*** mleist has joined #zope3-dev13:34
*** faassen has joined #zope3-dev13:43
*** mkerrin has joined #zope3-dev13:45
SteveAfaassen: aargh.  python's optimisation of calling __len__ when doing list(foo) keeps tripping me up!13:51
mgedminSteveA, what do you do?13:51
SteveAfor example, sqlobject's ResultSet has a __len__ that makes an SQL count query13:52
faassenSteveA: I didn't do it!13:52
SteveAfaassen: i know!  but you're a good recipient of rants!  you have a weblog!13:52
faassenSteveA: heh. :)13:52
faassenSteveA: that means I'm a good generator of rants. :)13:52
mgedminwhile True: yield random.choice(self.rants)13:53
SteveAso, i want to do $work on __len__ of an object13:53
SteveAand the object is iterable, so it makes sense to say list(object), for example, from a tal:repeat in TALES13:54
SteveAbut, tal:repeat will call __len__, even though it doesn't really want to13:54
SteveAthere is no way from the object in question to tell if __len__ is a legitimate call worth expending effort on, or an optimisation from list()13:54
SteveAi did manage to look up to the next frame, look in its names and bytecode13:55
SteveAand try to infer if the global list() callable was called last13:55
SteveAbut that's kinda flaky13:55
SteveAi think this is a mistake in python development.  python should have used a separate slot for "optimize me for list()"13:55
SteveAlike __listlen__13:56
*** MrTopf has joined #zope3-dev13:58
faassenwrite it on your blog. :)13:59
mgedminget a blog, *then* write it13:59
SteveAi have been saying for 18 months "i should have a blog"14:00
SteveAand then i never do anything about it14:00
SteveAi am so lazy14:00
SteveAah... but i do have a fresh zope3 tree now, so i can look into that HTTP Headers bug that might have been fixed14:00
*** andres_ has joined #zope3-dev14:01
*** andres has quit IRC14:03
*** andres_ is now known as andres14:03
*** mleist has quit IRC14:03
lifelesswhy does tal:repeat call list(object) rather than iter(object)?14:06
philiKONlifeless, because it's probably been around longer than iter() has14:07
philiKONlifeless, feel free to change it14:07
romanofskidoes something similar exist like verbose security or is there a way to track, why some permissions seems not to work?14:08
romanofskia ++debug++security or something would be neat14:08
*** srichter has quit IRC14:09
philiKONromanofski, yeah, the Debug skin has a view for the Unauthorized exception that tells you the permission it's lacking14:09
* romanofski tries that14:10
romanofskihm.. it works with the debug skin, but not with my own skin14:12
* projekt01 there is also a debug namespace ++debug++tal or ++debug++source14:12
romanofskiright projekt01, but you can't debug why a user isn't allowed to access a particular page14:13
lifelessphiliKON: ok that helps then ;)14:14
SteveAlifeless: i'll look into that too, of course14:14
romanofskisomething like verboseSecurity in the old zope2 days14:14
projekt01romanofski, yes, that's true14:16
romanofskiphiliKON: how does the skin suppose to work, if a user is unauthorized to view an object? are there little tricks or buttons I've to press?14:17
projekt01romanofski, you can click on the Error view where is reporting unauthorized errors if you enable to traceback unauthorized errors.14:17
* romanofski looks for a big red button labeled: "press here and I do what you wan't" *G14:17
philiKONromanofski, the debug skin simply has a view for the Unauthorized exception14:18
philiKONa special view14:18
philiKONthat exposes security details14:18
romanofskiokey... i try to look it up14:18
andresromanofski, you can make the debug skin part of your skin.14:18
romanofskijust include it in the layers right?14:19
lifelessSteveA: naturally ;)14:19
andresromanofski, yes, i did that, and it works well.14:19
philiKONreminds me of writing that skins+layers = 1 proposal14:20
romanofskisounds like a very good idea andres - thanks :)14:20
romanofskithanks philiKON and projekt01 :)14:21
*** andres_ has joined #zope3-dev14:25
*** andres has quit IRC14:28
*** andres_ is now known as andres14:28
*** mleist has joined #zope3-dev14:28
*** andres_ has joined #zope3-dev14:40
*** mkerrin has quit IRC14:41
*** natea has joined #zope3-dev14:44
*** natea is now known as natea|zZz14:47
*** mgedmin has quit IRC14:50
*** andres has quit IRC14:57
*** andres_ is now known as andres14:58
*** J1m has joined #zope3-dev15:07
*** srichter has joined #zope3-dev15:13
*** ChanServ sets mode: +o srichter15:13
*** natea|zZz has quit IRC15:15
philiKONJ1m, just FYI, i'm updating zope 2.9 branch to the zope 3.2.0b1 tag15:22
J1mI really prefer using revision number in general, but the use of the tag here is convenient.15:23
* philiKON agrees15:23
*** mkerrin has joined #zope3-dev15:38
*** eins has quit IRC15:52
*** srichter has quit IRC16:00
*** natea has joined #zope3-dev16:01
*** srichter has joined #zope3-dev16:01
*** ChanServ sets mode: +o srichter16:01
d2mJ1m: are the files at private on purpose ?16:08
d2mok, thanks16:09
J1mI'll publish them when I'm ready to announce the release.16:09
*** agroszer has quit IRC16:16
*** agroszer has joined #zope3-dev16:17
*** hdima has quit IRC16:29
*** natea is now known as natea|zzz16:39
efgeI'm trying to get a list of subscribers for an event interface, but without "running" them. How would I go about doing that?16:53
SteveAyou'd need to dig into the component registries16:54
SteveAi don't believe there is a good API for getting things by interface16:54
J1mUse the subscriptions method on the adapter registry.16:54
J1mIf you get hold of the current site, it has an adapters attr which is the registry.16:55
J1mSteveA, for the June release, I'd like to follow up on a suggestion of yours.16:56
J1mI'd like to make explcit objects with methods for component lookup.16:56
J1m(In addition to what we have now.)16:56
philiKONJ1m, can you elaborate on that? what are "explicit objects with methods"?16:57
J1mThis way, an app could be a bit more explicit about how/where it looks things up.16:57
J1mso, now we use component package methods.16:57
J1mThese hide a policy of having thread-specific "sites",16:58
J1mI think it's fine to have such a default policy, but I think it would be good to have the ability to say: "use this particular component source".16:59
J1mIn certain contexts you want to use sources other than the default.16:59
philiKONi guess so16:59
SteveAthread-specific sites?16:59
philiKONwe already have that, no?16:59
J1mFor example, when processing ZCML, you want to use a ZCML-provides source of components.17:00
J1mphiliKON, we have bits of this, but it's not as well thought out as it should be.17:00
SteveAthis would make the metazcml registry of components more explicit17:00
J1mSteveA, the component methods use a "site".17:00
J1mThe definition of site is pluggable.17:01
J1mZope plugs in a mechanism that uses separate sites for each thread.17:01
SteveAas well as allow various nefarious plans i'd like to try out, like having a source of components for related to something about a particular interaction17:01
J1mwhich is exactly what Zope does now.17:01
J1mIn setting up a request, we set the thread-local site to one traversed to in the request.17:02
J1mAt the end of the request, we set the thread-local site back to the global one.17:02
efgeJ1m: ok I found my subscribers, thanks17:04
SteveAbut i think this "component source" would be less than the current zope3 concept of a "site"17:06
*** stub has quit IRC17:12
*** J1m has quit IRC17:14
benjiagreed, SteveA; if not, we need a better name then "site" for them17:16
SteveAsites carry a lot of extra stuff now17:16
SteveAlike, the expectation of a URL traversal to them17:16
SteveAzodb stuff17:16
SteveAthe expectation of a UI17:16
SteveAof mutability17:16
SteveAof having local code17:16
SteveAon launchpad i've had no particular use for the zope3 "site" concept as such, but i do have uses for something that says from where i look up adapters and utilities17:17
SteveAmaybe this is a case like with TAL/TALES vs DTML.  a less powerful but clearer concept takes you farther.17:18
*** dman13 has joined #zope3-dev17:24
*** philiKON has quit IRC17:27
*** philiKON has joined #zope3-dev17:29
zbirI'd like to contribute another zopectl command that I have to allow for Zope2-like 'zopectl run' to allow for easy client scripting. I've got it working here. For stuff like zopeskel binaries, is it just enough to have it and the batch file working?17:29
zbirJ1m suggested working on his branch, since it *is* a new feature17:30
zbirjust wanted to get some other input.17:30
zbiralso, I'm curious why the very slight extra effort isn't made in zopectl debug to put the root container in the locals - I've got a version that does. Saves having to do 'root = debugger.root()' every. single. time.17:32
benji+1 zbir17:33
*** J1m has joined #zope3-dev17:33
*** sashav has quit IRC17:34
philiKONJ1m, btw, would you care to explain why exactly we need 'Zope' to be in the tag name?17:38
J1mBecause of a bug/missfeature in zpkg.17:39
J1mIt constructs the tag name by combining the collection name and the release name.17:39
J1mI didn't find a way to get around that and I didn't want to delay things to fix it.17:40
philiKONJ1m, intersting, because MacYET reports that a '2.9.0b1' tag name works fine on zope 217:45
philiKON(no Zope- prefix)17:45
srichtersee Andreas' last checkin17:52
srichterLog message for revision 40597:17:52
srichter  we need to pass -r tagname to zpkg17:52
philiKONbut that needs to be done anyways, right17:53
srichteroh, ytes, right17:53
J1mI'm not allowed to register package info for zope17:57
*** jhauser has quit IRC17:58
*** MrTopf has quit IRC18:11
*** projekt01 has quit IRC18:21
J1msrichter, OK, I've done release steps 1-8,5.18:23
J1mI sent a note to zope3-dev.18:23
J1mI'd be inclined to wait a day to send the message elsewhere.18:23
J1mCan you follow up and send it to ither lists and outlets?18:24
J1mAnd I think you have to register it with pypi.18:24
*** sashav has joined #zope3-dev18:30
*** dobee has quit IRC18:33
agroszerhelp, I'm stuck with Pluggable Authentication18:33
agroszerI can't create a working PAU, principal folder, group folder by code18:33
agroszerthey are created, registered but don't seem to work18:33
agroszeris there any package/project doing something like this?18:33
zbirdefine "don't seem to work"18:35
*** philiKON has quit IRC18:52
andresIf its interesting: Zope3.2b1 is passing all manual and automated testing here.18:52
andresI mean with my application ;-)18:52
andresExcept one thing i had fixed manually in zope3.1 but forgot to report. Im not sure if this is my failure or zope's. What is the preferred way to report a (supposedly) fix of 3 lines? Mail, IRC, or Bugtracker?18:55
J1mThe collector18:56
*** sashav has quit IRC18:59
*** sashav has joined #zope3-dev19:01
*** _anguenot has quit IRC19:02
srichterJ1m: ok, will do19:03
*** _anguenot has joined #zope3-dev19:05
*** mleist has quit IRC19:06
*** sashav has quit IRC19:11
*** mkerrin has left #zope3-dev19:16
*** zagy has quit IRC19:36
*** drzoltron_ has joined #zope3-dev19:46
*** zagy has joined #zope3-dev19:47
*** mgedmin has joined #zope3-dev19:54
*** srichter has quit IRC20:00
*** romanofski is now known as rom|aw20:03
*** agroszer has quit IRC20:03
*** drzoltron_ has quit IRC20:36
*** faassen has quit IRC20:37
*** andres has quit IRC20:40
*** hazmat has quit IRC20:52
*** alga has joined #zope3-dev20:52
*** BjornT has quit IRC20:53
*** WebMaven_ is now known as WebMaven20:55
*** BjornT has joined #zope3-dev20:58
*** natea|zzz is now known as natea21:02
*** timte has joined #zope3-dev21:13
*** efge has quit IRC21:19
*** andres has joined #zope3-dev21:26
*** andres_ has joined #zope3-dev21:38
*** andres__ has joined #zope3-dev21:40
*** deo has joined #zope3-dev21:46
*** andres has quit IRC21:54
*** andres_ has quit IRC21:57
*** sashav has joined #zope3-dev22:04
*** natea is now known as natea|away22:18
*** ignas has quit IRC22:22
*** natea|away is now known as natea22:34
*** mgedmin has quit IRC22:39
*** Aiste has quit IRC22:40
*** BjornT has quit IRC22:51
*** BjornT has joined #zope3-dev22:53
*** pcardune has joined #zope3-dev23:02
*** alga has quit IRC23:04
*** srichter has joined #zope3-dev23:11
*** ChanServ sets mode: +o srichter23:11
*** timte has quit IRC23:16
*** slinkP has joined #zope3-dev23:19
*** slinkP has left #zope3-dev23:20
*** sashav has quit IRC23:22
*** mp has quit IRC23:23
*** mp has joined #zope3-dev23:25
*** sashav has joined #zope3-dev23:30
*** mp has quit IRC23:45
*** mp has joined #zope3-dev23:45
*** Aiste has joined #zope3-dev23:48

Generated by 2.15.1 by Marius Gedminas - find it at!