IRC log of #zope3-dev for Tuesday, 2005-09-27

*** palmTree has quit IRC00:05
*** whit has quit IRC00:06
*** whit has joined #zope3-dev00:19
*** FrankBerger has left #zope3-dev00:20
*** deo has quit IRC00:34
*** ignas has joined #zope3-dev00:36
*** yota has quit IRC00:38
*** projekt01 has quit IRC00:42
*** whit has quit IRC00:52
*** anguenot has joined #zope3-dev00:53
*** whit has joined #zope3-dev00:54
*** benji_york has quit IRC01:28
*** projekt01 has joined #zope3-dev01:31
*** whit has quit IRC02:22
*** whit has joined #zope3-dev02:24
*** whit has quit IRC02:24
*** ignas has quit IRC02:44
*** d2m has quit IRC03:02
*** d2m_ has joined #zope3-dev03:02
*** d2m_ is now known as d2m03:02
*** niemeyer has quit IRC03:28
*** stub has joined #zope3-dev03:29
*** batok has joined #zope3-dev03:42
*** projekt01 has quit IRC03:45
*** batok has quit IRC03:52
*** niemeyer has joined #zope3-dev03:53
*** jinty has quit IRC04:16
*** bradb has quit IRC06:05
*** tvon has quit IRC06:14
*** tvon has joined #zope3-dev06:22
*** MacYET has joined #zope3-dev07:09
*** MacYET has left #zope3-dev07:09
*** dobee has joined #zope3-dev07:58
*** zagy has joined #zope3-dev08:14
*** niemeyer has quit IRC08:20
*** BjornT has quit IRC08:59
*** BjornT has joined #zope3-dev09:03
*** stub has quit IRC09:53
*** projekt01 has joined #zope3-dev09:58
*** SureshZ has left #zope3-dev10:12
*** tekNico has joined #zope3-dev10:12
*** MacYET has joined #zope3-dev10:23
*** stub has joined #zope3-dev10:37
*** tarek has joined #zope3-dev10:39
*** philiKON has joined #zope3-dev11:06
MacYETalready in SHG?11:22
MacYETdon't litter11:24
MacYETdon't spit11:24
philiKONyou confuse this with singapore11:24
philiKONnow you got it :011:24
MacYETdon#t spit at the chinese wall11:25
philiKONhaha, ok11:25
*** faassen has joined #zope3-dev11:35
*** sashav has joined #zope3-dev11:39
*** MJ has quit IRC11:44
*** d2m has quit IRC12:05
*** Aiste has quit IRC12:21
*** Aiste has joined #zope3-dev12:22
tekNicophiliKON: just raised your book's price from 27 to 38£. :-(12:26
*** kaczordek has joined #zope3-dev12:27
*** jinty has joined #zope3-dev12:35
*** mgedmin has joined #zope3-dev12:45
philiKONtekNico, ouch13:01
*** roym has joined #zope3-dev13:02
philiKONtekNico, i'm sorry, i don't do the pricing13:03
tekNicophiliKON: I'd like to put the book on my Palm. Are you going to let buyers have an electronic copy somehow? :-)13:03
philiKONi can't decide that.13:03
tekNicophiliKON: who can, then?13:04
philiKONspringer has the exclusive rights to publish13:04
philiKONspringer is the publisher13:04
roymfolks: why are view classes untrusted, if all they have to do (and can do) is use removeSecurityProxy on the proxied object?13:04
tekNicophiliKON: does your agreement explicitly forbid such possibility?13:04
philiKONtekNico, yes13:05
*** ignas has joined #zope3-dev13:05
tekNicophiliKON: I see. :-(13:05
philiKONit's not unusual13:05
philiKONroym, because you simply shouldn't use removeSecurityProxy13:05
philiKONif you use removeSecurityProxy, you're throwing away the zope 3 security system; you're on your own then.13:06
roymso it is almost always better design to move such code into the content class?13:06
philiKONno. never use removeSecurityProxy in your application code13:07
philiKONif you're tempted to use it, you're doing somethign wrong13:07
philiKONcontent classes should generally be dull13:07
philiKONthey know how to store and retrieve data. that's it13:07
roymhmm, if I need to do an update on content classes, then you're saying that this forces one to do this only thru the published interfaces (w/security checking).13:08
philiKONif by "publihsed interfaces" you mean something like <require permission="zope.ManageContent" interface=".mymodule.interfaces.IMyClasssInterface" />, then yes13:09
philiKONif you're setting stuff on a content object from the "outside" (e.g. an untrusted view component), it needs to be on attributes that have security declarations13:10
mgedminphiliKON, sometimes removeSecurityProxy is inevitable13:10
philiKONotherwise you get a ForbiddenAttributeError13:10
mgedmin... unless you have a trusted adapter13:10
philiKONmgedmin, those cases are *Extremely* rare13:10
philiKONexactly, most of the times it can be fixed with a trusted adapter13:10
* mgedmin is enlightened13:11
roymCan a view be declared trusted? after all, isn't it an adapter.13:11
philiKONno, we don't have such a mechanism yet AFAIK13:12
philiKONroym, maybe you can shed some light as to what you're trying to do?13:12
roymI have a folder iterator; it needs to store a user choice (from a question).13:13
*** andres has joined #zope3-dev13:13
roymThe user sees a view (of a readonly content object - question).13:14
andresSorry, for bothering again, but i simply cant get an addform,which is created via code, to work.13:14
roymHowever, I find that to present the view, I need to access the question attributes, and find myself circumventing zope security since I am trying to do this from a view.13:14
*** MacYET has left #zope3-dev13:15
*** jinty has quit IRC13:15
philiKONroym, sounds like you're missing security declarations for that question content object13:15
philiKONandres, what's the problem?13:15
roymphiliKON: let me dig in deeper into that.. thanks.13:16
roym...and incidentally, is there an easy way to find out the permissions that the current principal has at runtime13:16
andresphiliKON, im not even sure, if i try to do it the right way. Currently i inherit from editview, is that correct?13:16
roymI find checker.get_permissions very handy13:17
philiKONroym, not sure if this is the "blessed" way, but maybe canAccess and canWrite is already enough for you?13:17
roymthats good enough for moi.13:18
andresphiliKON, in __init__ i set self.schema to the Interface for which the addfrom should be displayed. In this form: self.schema= abc.interfaces.IPerson13:20
andresbut then i always get type errors13:21
philiKONaha. maybe you can post the traceback somewhere... i might take a look at it13:23
andresIm doing this atm:
roymAm i right in assuming that if I have a PersistentDict/PersistentList attribute, then I don't need to set _p_changed on the container to have the changes persisted?13:30
andresphiliKON, Others, Any Idea13:30
mgedminroym, you are right13:31
mgedminandres, does your content object provide IPerson?13:31
mgedminis it adaptable to IPerson?13:31
mgedminI guess not -- at least that's what the traceback says13:32
andresmgedmin, It should.13:32
mgedminprovide? or be adaptable?13:32
*** MJ has joined #zope3-dev13:35
mgedminso, klinik.Klinik implements(IPerson)13:36
philiKONroym, yep13:36
mgedminis there only one IPerson?13:36
philiKONroym, yep 2 persistentlist/persistentdict13:36
*** srichter has quit IRC13:38
andresmgedmin, args, thanks, i think i got it.13:39
*** regebro has joined #zope3-dev13:56
andresYes, it really worked.14:06
andresAlthough for now only with existing IPerson's14:06
*** sashav has quit IRC14:10
*** anguenot has quit IRC14:22
*** srichter has joined #zope3-dev14:33
*** niemeyer has joined #zope3-dev15:04
*** kaczordek has quit IRC15:08
*** ChanServ sets mode: +o srichter15:10
*** sashav has joined #zope3-dev15:24
*** benji_york has joined #zope3-dev15:33
*** MrTopf has joined #zope3-dev15:50
*** mgedmin has quit IRC15:52
*** anguenot has joined #zope3-dev15:57
*** tvon has quit IRC16:03
*** MacYET has joined #zope3-dev16:04
*** mkerrin has joined #zope3-dev16:06
*** sashav has quit IRC16:19
*** niemeyer_ has joined #zope3-dev16:26
*** J1m has joined #zope3-dev16:28
*** deo has joined #zope3-dev16:28
*** VladDrac has quit IRC16:29
*** J1m has quit IRC16:30
*** deo has quit IRC16:37
*** niemeyer has quit IRC16:41
*** niemeyer_ is now known as niemeyer16:41
*** efge has joined #zope3-dev17:15
*** faassen has quit IRC17:21
*** _projekt01 has joined #zope3-dev17:22
*** _torsti76_ has joined #zope3-dev17:31
_torsti76_hi @all!17:35
benji_yorkhi there17:35
_torsti76_stefan, i set up a closed channel for our NeckarSprint at #zope3-neckar17:35
philiKON_torsti76_, can i join?17:36
MacYET_torsti76_: why is the channel protected?17:37
efgewhy closed ?17:37
philiKONright, why closed17:37
* philiKON didn't even bother to think about that17:37
_torsti76_hmm, maybe it's too long since i used irc for the last time17:38
_torsti76_but then it was "in" for some kiddies to occupy opened channels17:38
efgethis isn't undernet :)17:38
philiKON_torsti76_, no problem with this on freenode17:39
philiKON_torsti76_, or do you see kiddies spamming this channel?17:39
_torsti76_nope - so just for your relief - the channel is open now ;o)17:39
*** projekt01 has quit IRC17:40
_torsti76_you're welcome17:40
*** _projekt01 is now known as projekt0117:40
*** bradb_ has joined #zope3-dev17:41
*** jvts has joined #zope3-dev17:44
*** faassen has joined #zope3-dev17:47
*** bradb_ has quit IRC17:48
*** _torsti76_ has left #zope3-dev17:50
*** zagy has quit IRC17:50
*** bradb has joined #zope3-dev17:52
*** projekt01 has quit IRC17:57
*** whit has joined #zope3-dev17:57
*** vlado has joined #zope3-dev18:00
*** whit has quit IRC18:09
*** zagy has joined #zope3-dev18:11
*** d2m has joined #zope3-dev18:12
* MacYET is away: Blicket auf zum Retterblick, Alle reuig Zarten, Euch zu sel'gem GlŸck, Dankend umzuarten!18:23
*** MacYET has left #zope3-dev18:24
vladothere is use of 2.4 style decorators in zope.formlib doctests but i just can't use them in the doctest of my simple script18:25
vladowhat am i doing wrong ?18:26
vladoi'm getting SyntaxError: unexpected EOF while parsing18:26
vladoanyone ?18:27
benji_yorkvlado, do you have an encoding marker in your source file?18:27
benji_yorkhow long is your script?18:27
benji_york(i.e. is is short enough to post here?)18:28
*** alga has joined #zope3-dev18:29
vlado>>> @d18:29
vlado>>> def f(): pass18:29
vladodef d(f): pass18:29
vladoif __name__ == '__main__':18:29
vlado    import doctest18:29
vlado    doctest.testmod()18:29
benji_yorkvlado, the second "prompt" in your doctest should be "..." because you're continuing the defenition of the function18:31
benji_yorkwhen writing doctests each >>> starts a new "file" which will be compiled and run, hence the EOF error18:32
vladoi wouldn't guess it18:34
benji_yorknow you will  :)18:34
*** Alef has joined #zope3-dev18:38
*** jinty has joined #zope3-dev18:39
*** niemeyer is now known as nie_lunch18:48
vladoso we can't decorate classes because in current implementation the decorators are part of func's definition...18:52
benji_yorkvlado (if I understand you correctly): right, Python doesn't have "class decorators"18:53
vladomaybe pje's bracked-based impl has...18:56
vladoor maybe they are yagni ;-)18:56
SteveAthe class advice we use for implements() and other things is equivalent19:00
vladobut the route is different ;-)19:02
*** dobee has quit IRC19:04
*** stub has quit IRC19:13
*** projekt01 has joined #zope3-dev19:14
*** MrTopf has quit IRC19:17
*** nie_lunch is now known as niemeyer19:42
*** deo has joined #zope3-dev19:45
*** philiKON has quit IRC19:50
*** J1m has joined #zope3-dev19:51
*** alga has quit IRC19:57
*** tvon has joined #zope3-dev20:03
J1manguenot, efge, regebro ayt?20:06
*** faassen has quit IRC20:10
*** J1m has quit IRC20:11
anguenotHi Jim20:14
*** andres has quit IRC20:19
*** yota has joined #zope3-dev20:26
*** srichter has quit IRC20:28
*** deo has quit IRC20:41
*** vlado has quit IRC20:41
*** mgedmin has joined #zope3-dev20:41
*** niemeyer_ has joined #zope3-dev20:50
benji_yorkmgedmin, I mislead you yesterday about the contact info for planet.python.org20:51
benji_yorkI still don't know who the correct contact is, so I just emailed, and fdrake replied :)20:51
*** niemeyer has quit IRC20:53
mgedminbenji_york, the mail you gave me was the one which is prominently displayed on the front page of planetpython.org21:00
mgedminI'm surprised it was wrong21:00
*** tarek has quit IRC21:02
benji_yorkmgedmin, the email I gave you *is* the contact for, but that site has been made "unofficial" by and the maintainer of (who responded to me) said he didn't know how to contact the maintainer of planet.python.org21:03
benji_yorkFred was hopeful that one of the people that get the webmaster@ email would be able help21:06
benji_york(but I haven't heard anything back yet)21:06
*** andres has joined #zope3-dev21:06
andresHi again.21:06
mgedminbenji_york, your blog is already aggregated on the python planet21:06
benji_yorkon the "unofficial", but not (they use different lists of feeds)21:07
* mgedmin tries to remember which one he aggregates21:08
*** mkerrin has quit IRC21:09
*** tekNico has left #zope3-dev21:13
*** anguenot has quit IRC21:25
*** SureshZ has joined #zope3-dev21:28
andresSorry for bothering again. My Problem: I want different autogen. form's on one page.  What i did: I subclassed Editview. With this i can generate a editform for one object. This works, as long as only on formular is included, but as soon two are included i get an "Unautorized: You're not allowed in here."21:30
andresWith no further information than the Object + Address of the current container.21:31
andresAny idea how i could get more information about that? Or what the problem could be?21:33
andresBoth forms, i try to display at once work, when i display them alone.21:33
*** efge has quit IRC21:35
*** niemeyer_ is now known as niemeyer21:36
benji_yorksorry, andres, nothing off of the top of my head and can't dig into it at the moment21:39
andresbenji_york, its okay ;-)21:39
andresthanks anyway21:40
*** projekt01 has quit IRC22:05
*** projekt01 has joined #zope3-dev22:05
*** tvon has quit IRC22:16
*** Alef has quit IRC22:19
*** FrankBerger has joined #zope3-dev22:25
FrankBergerhello everyone, can someone help me to get tcpwatch and dochttp running correctly? it seems that dochttp does not include all requests into the textfile and die responses are missing at all :-(22:26
mgedminFrankBerger, try the built-in request recorder22:27
*** mgedmin has quit IRC22:29
FrankBergermgedmin: you have developed that recorder?22:30
FrankBergeri am currently working on something similar to that (thats why i am looking after tcpwatch and dochttp) :-)22:31
benji_yorkFrankBerger, note that Zope 3.2 will include testbrowser which is a friendlier way of doing functional testing22:33
FrankBergerwow, cool, may be some day i can contribute to that solution22:34
FrankBergerone of my targets at the moment is to seperate testscripts from testdata22:36
FrankBergerand i am looking to get a way from an HTTP based solution towards an "web-browser" simulation like mechanize to get the possibility to get javascript support some day22:37
benji_yorkthat's what testbrowser is (a software "web-browser"), but it doesn't currently support JavaScript (and I don't think it ever should)22:38
FrankBergerwow, are you developing that solution?22:39
benji_yorkyep, let me find the Subersion URL for you to check it out...22:40
FrankBergeralready found it in google :-)22:40
FrankBerger thats it?22:41
benji_yorkgo to the Zope 3 "src/zope" directory and check out svn:// and you can try it out22:42
FrankBergerok i will do that, and look about the differences22:43
FrankBergerdo you know tools like canoo webtest?22:43
benji_yorkI haven't seen/used that one22:44
FrankBergerwell, it is a java based solution, made as an ANT plugin22:45
*** ignas has quit IRC22:46
FrankBergergood thing, they have an XML format which i wanted to try to use22:46
*** tvon has joined #zope3-dev22:53
benji_yorkreading about canoo now...22:54
* benji_york call's "inapropriate use of XML" on canoo22:55
benji_yorkI'm not real keen on their total dismissal of record/playback; we have a prototype testbrowser doctest and selenium test recorder that doesn't have the problems they characterize record/playback with having (because it records at the user gesture level, not at the HTTP level)22:57
FrankBergeryes, thats why i try to do something inbetween like you22:58
FrankBergerwhat do i have to do to get testbrowser running in my zope instance?22:59
benji_yorkjust cd to "src/zope" and so "svn co svn://" and you should be able to import it and run the tests23:04
benji_york(you may also have to install one or more of its dependencies, but try without them first)23:04
FrankBergerah, ok, so currently you do not have a webinterface23:05
benji_yorkhmm, no... what would a web interface do?23:05
FrankBergerthe webpages? (or parts of it)23:06
benji_yorkI'm not understanding.  Are you asking if testbrowser can access non-zope3 web pages?23:07
FrankBergerno, well, i try to guide the user throu the process of creating the test-script23:07
FrankBergere.g. you have a zope-from say (open_url)23:08
FrankBergerthe user can enter an URL23:08
FrankBergerand afterwards testbrowser uses mechanize to get that URL23:08
benji_yorkoh, ok like an interactive test case builder23:09
FrankBergerbut it is not as easy at is sounds like23:09
benji_yorkyou'd probably be interested in our testrecorder code, but it's not been released yet23:09
FrankBergerup to what i have don sofar23:09
FrankBergerwell, if i have something to show, i can send you some proof-of-concept code23:11
FrankBergermay be we can join some of it (or the ideas behind it only)23:11
FrankBergercurrently i am HTTP based and not browser based23:12
FrankBergerbecause i have problems, to get mechanize running inside zope :-)23:12
benji_yorkfor mechanize in zope, look at testbrowser.  It uses mechanize as its back-end23:13
benji_yorksee testbrowser.testing for how to wire mechanize directly into the Zope 3 publisher (to bypass going onto the network, and also so you don't have to have an open port for Zope)23:14
FrankBergerah, well, but my problem was a little bit different, I wanted to run mechanize for instance by calling its functions via a web-frontend23:15
FrankBergerproblem is here to get an persitant state for that mechanize instance23:16
benji_yorkoh, ok23:16
benji_yorkyou *might* be able to put the mechanize browser object into a principal annotation (but they must be pickleable, and I don't know if they are or not)23:17
FrankBergeras far as what i have seen they are not and that was one of my problems23:18
FrankBergertherefore my tool looks very similar to your request-recorder23:18
FrankBergeri additionally have some functions to seperate data like form-data/cookies/session-ids from the requests23:22
benji_yorkwell, I will warn you that we decided that it was impossible to make HTTP-level recording work well enough to be usable, but good luck anyway FrankBerger :)23:30
FrankBergeryea, i know that, that is one of the reasons way i wanted to go to mechanize in the long-term :-)23:31
FrankBergerbut i have to finish a release-death-line on Friday and i had to postpone the switch to mechanize due to the problems i had with it inside zope :-)23:33
*** jinty has quit IRC23:50
*** SureshZ has left #zope3-dev23:51

Generated by 2.15.1 by Marius Gedminas - find it at!