- Zope - Zope lists

OSX users looking to test a Mac GUI for Zope?
by David Shaw 25 Aug '01

25 Aug '01

I have been working on a Mac OS X package for Zope based off of the excellent binary distribution for MacOS X on zope.org. What makes mine different is (a) it includes ZEO, and (b) is has a native Mac GUI configuration tool front-end. You can see some screen shots here: http://homepage.mac.com/dshaw/ If you are a MacOS X user and are interested in helping me develop or test, just let me know via email. Since the distribution includes a full copy of Zope and python, it's not small, so bear that in mind.

2 1

Fw: gvib kinterbasdb firebird linux
by W. Robert Kellock 25 Aug '01

25 Aug '01

Hello again, Linux newbie problem resolved now (why are the Linux newbie books so silent about shell programming?) means I have successfully compiled gvibBase.so under Python 1.5.2. However this hasn't got me out of the woods yet, since Zope 2.4.1b uses Python2.1 and gvibBase.c code contains Py_Realloc which I think has become PyMem_Realloc. Unless somebody has already rewritten gvibBase.c to compile under Python 2.1 and would care to share it, I will Bash on and post the result when I am done. Cheers. ------------------------------------------------------------------------------------------------------------- W.Robert Kellock Ph: +64 3 326 6115 Sales Manager Fax: +64 3 326 6115 Credit Systems Control Ltd Web: www.creditscore.co.nz "making decisions easy" ------------------------------------------------------------------------------------------------------------- ----- Original Message ----- From: W. Robert Kellock To: zope(a)zope.org Sent: Friday, 24 August 2001 20:02 Subject: gvib kinterbasdb firebird linux I have for some months now been utilising, very successfully, gvibDA on Windows 2000 with Zope 2.3.2 and Firebird. Unfortunately I cannot say the same for RedHat 7.1 (very much a Linux newbie so be gentle ;- ). I have installed the latest Zope 2.4.1b (which uses Python2.1) and Firebird binaries and they are working fine. However when I compile gviBase.so and attempt to use it within Zope or Python (both 1.5 rpm and 2.1.1 compiled from source) I get Import error: undefined symbol isc_dsql_execute I assume this means that the compile process cannot see the Firebird shared library libgds.so, but I can't see why. I have tryed to edit the Setup.in file to import /usr/lib/libgds.so instead of the more usual -lgds but it makes no difference. Can anybody help me out? I have also tryed kinterbasdb, but when I import it I get Import error: undefined symbol crypt I know I'm supposed to editing kinterbasdb's Makefile to include -lcrypt but I'm not sure where. gvib is my main priority, but a copy of an appropriate Makefile for kinterbasdb would also be much appreciated. Thanks. ------------------------------------------------------------------------------------------------------------- W.Robert Kellock Ph: +64 3 326 6115 Sales Manager Fax: +64 3 326 6115 Credit Systems Control Ltd Web: www.creditscore.co.nz "making decisions easy" -------------------------------------------------------------------------------------------------------------

1 0

RE: [Zope] Ruling from the bench regarding 'Hotfixes' :)
by sean.upton＠uniontrib.com 25 Aug '01

25 Aug '01

Jargon: why not "coolfix" instead? It connotes: 1 - No severity (cool vs. hot) 2 - Novelty (cool==neat) 3 - It does indeed make some changes to the core, but nothing crucial 4 - It doesn't require you to feed it bananas ;) Sean -----Original Message----- From: Brian Lloyd [mailto:brian.lloyd@zope.com] Sent: Friday, August 24, 2001 12:04 PM To: marc lindahl; zope(a)zope.org Subject: RE: [Zope] Ruling from the bench regarding 'Hotfixes' :) > I disagree -- there needs to be a clearer distinction between > products that > add new functionality and those that change existing behavior. > > We need > some clearly defined terms for these areas. That much is obvious > - already > the user (product building) community has interpreted the word 'hotfix' > differently than ZC has intended... > > Some pithy names would go a long way towards this goal -- HotPatch - very > evocative! My argument is that that is a documentation issue. If people strongly disagree and think that new jargon needs to be invented, that's ok, but the new jargon should not be "Hotxxx" or anything that could be confused with "hotfix", for jargon-backward-compatibility reasons. > I for one would like to be able to do Product searches and have a list for > Core Enhancements as separate from Added Functionality... I don't really understand the distinction as put here. To me, a product just lets me do something that I wouldn't be able to do without it. Core Session Tracking is both a core enhancement and added functionality. I suspect you mean "products that add new objects" and "products that modify existing objects". I suspect that is not a primary criteria most people would use to look for products (though this is separate from the jargon issue - it is really a categorization issue). >From the jargon standpoint, I could see some piece of jargon that equals what we currently call "monkey patch" being useful as a shorthand amongst product developers (but it should not be a word that can be confused with hotfix). Brian Lloyd brian(a)zope.com Zope Corporation www.zope.com _______________________________________________ Zope maillist - Zope(a)zope.org http://lists.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope-dev )

6 7

[Note] "absolute_url" strips any context
by Dieter Maurer 25 Aug '01

25 Aug '01

I just recognized (and was astonished) that "absolute_url" strips any context from the URL. Thus, e.g., if you have the following structure: / F/ D -> absolute_url en/ de/ fr/ then "GET('/en/F/D')" will return "/F/D". This is because, "absolute_url" is based on "getPhysicalPath". Probably it is this behaviour that should be indicated by the "absolute" in "absolute_url" (I never understood it so far). Nevertheless, I suggest to make this behaviour more explicit in the documentation. Dieter

2 1

REQUEST syntax
by Montagne, Michael 24 Aug '01

24 Aug '01

I'm attempting to loop through a database and create users. I can successfully print a table with all the proper info but the users are not being added. No errors but no users. If I plug in specific strings (ie <dtml-call "REQUEST.set('name','bob')">), it works ok. This is my code: <dtml-call "REQUEST.set('name','<dtml-var Login>')"> <dtml-call "REQUEST.set('password','<dtml-var password>')"> <dtml-call "REQUEST.set('confirm','<dtml-var password>')"> <dtml-call "REQUEST.set('domains',[])"> <dtml-call "REQUEST.set('roles',['Staff' 'YihawAuthor'])"> <dtml-call "acl_users.manage_users('Add',REQUEST,RESPONSE)"> What is wrong?

4 3

Permissions and roles
by Todd Loomis 24 Aug '01

24 Aug '01

Hi: I'm trying to set roles and permissions to folders as they are created by the users the a form. However I get this error: Error Type: TypeError Error Value: too many arguments; expected 5, got 6 Here's my code, can anybody assist me? <dtml-if expr="_.SecurityCheckPermission('Add Folders', this())"> <dtml-with "folders"> <dtml-call expr="manage_addFolder(id=REQUEST['FolderName'], title=REQUEST['FolderName'])"> </dtml-with> <dtml-with "REQUEST['FolderName']"> <dtml-call "manage_addProperty('View','Access contents information',['PAO','Manager'],1,REQUEST)"> </dtml-with> </dtml-if> <dtml-call "RESPONSE.redirect('./index_html')"> Thanks, Todd

2 1

ZOracle SP does not commit ?
by Ernst Eeldert 24 Aug '01

24 Aug '01

Hi, when I use the Z Oracle Stored Procedure object to execute a package procedure which inserts stuff into the database, the transaction does not get committed. I use a Python Script to execute the SP. I checked the source code of the ZOracle SP, and it seems that the commit gets done in the _finish() method. How can I tell if this method is called ? Thanks. Ernst

2 1

Traceback error
by oliver 24 Aug '01

24 Aug '01

Hello guys! I`m really new to zope now i need help! I have made a new product called 'minimal' and now i want to inizialize it and i get this error: 2001-08-24T12:25:16 ERROR(200) Zope Couldn't install minimal Traceback (innermost last): File C:\PROGRA~1\zope240\lib\python\OFS\Application.py, line 657, in install_product (Object: ApplicationDefaultPermissions) (Info: minimal) File C:\Program Files\Zope240\lib\python\Products\minimal\__init__.py, line 9, in initialize File C:\PROGRA~1\zope240\lib\python\App\ProductContext.py, line 166, in registerClass TypeError: unsubscriptable object what i have done wrong? Thank you for your help! Oli

2 1

Proxy role + zclass + python script = Unauthorized
by Ayhan Ergul 24 Aug '01

24 Aug '01

Folks, I have a python script (called "clone") defined as a method of a zclass that, when invoked on the instance of the zclass (KM Chart Applet), makes a copy of the instance in a different folder. This works fine when the user has role manager, but fails with "Unauthorized" complaining about the permission "Add KM Chart Applets" although the script has proxy role manager. In the python script manage_copyObjects succeeds but manage_pasteObjects fails (traceback below). Running zope 2.3.3 Any clues why I'm losing the permission when running the script proxied to manager as opposed to a user with manager role? I don't think I'm running into the post 2.2 policy of not propagating proxy roles, since the python script is called directly from browser and it calls the stock manage_....Objects() directly?? Ayhan Traceback (innermost last): File E:\PROGRA~1\TPMC_Z~1\lib\python\ZPublisher\Publish.py, line 223, in publish_module File E:\PROGRA~1\TPMC_Z~1\lib\python\ZPublisher\Publish.py, line 187, in publish File E:\PROGRA~1\TPMC_Z~1\lib\python\ZPublisher\Publish.py, line 171, in publish File E:\PROGRA~1\TPMC_Z~1\lib\python\ZPublisher\mapply.py, line 160, in mapply (Object: clone) File E:\PROGRA~1\TPMC_Z~1\lib\python\ZPublisher\Publish.py, line 112, in call_object (Object: clone) File E:\PROGRA~1\TPMC_Z~1\lib\python\Shared\DC\Scripts\Bindings.py, line 324, in __call__ (Object: clone) File E:\PROGRA~1\TPMC_Z~1\lib\python\Shared\DC\Scripts\Bindings.py, line 354, in _bindAndExec (Object: clone) File E:\PROGRA~1\TPMC_Z~1\lib\python\Products\PythonScripts\PythonScript.py, line 336, in _exec (Object: clone) (Info: ({'script': <PythonScript instance at 035AF170>, 'context': <KMChartApplet instance at 03510C10>, 'container': <KMChartApplet instance at 03510C10>, 'REQUEST': <TemplateDict object at 36f2c70>, 'traverse_subpath': []}, (), {}, None)) File Script (Python), line 19, in clone (Object: Traversable) File E:\PROGRA~1\TPMC_Z~1\lib\python\OFS\CopySupport.py, line 216, in manage_pasteObjects (Object: Traversable) File E:\PROGRA~1\TPMC_Z~1\lib\python\OFS\CopySupport.py, line 414, in _verifyObjectPaste (Object: Traversable) Unauthorized: Add KM Chart Applets __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/

2 1

Accessing Dictionary Values with URL Traversal
by Thomas Guettler 24 Aug '01

24 Aug '01

I just had the idea that it could be possible to access the values of a dictionary with URL traversal: A object of the following class: class MyClass: myDict={ "foo": "<html> bar </html>", "dtml": DTMLFile("foo.dtml") } could be accessed like this: http://localhost:8080/myClassObject/myDict/foo --> <html> bar </html> http://localhost:8080/myClassObject/myDict/dtml --> foo.dtml Is something like this already possible? The above thing would be sometimes better than an ObjectManager, because you can have serveral dictionaries in one class. OK, you could use an ObjectManager which contains ObjectManagers, too. Nevertheless I would like to know what you think! -- Open Source Software Solutions Thomas Guettler <thomas(a)thomas-guettler.de> http://www.thomas-guettler.de

3 3