- Zope-Dev - Zope lists

RE: [Zope-dev] ZODBCA questions...
by Schmidt, Allen J. 10 Apr '01

10 Apr '01

We too are having problems connecting to Access2000 with ZODBCDA. Seemed to be working fine (and I guess it still does) but it is slowing our whole site down. all our parts were playing together very well until (we think) about the 2.3.x introduction. Our site grinds to a halt and is very slow in every aspect; usually in the morning but on and off throughout the day. We tried some caching in various places which seemed to help for a while but it is slowing back down again. Access2000 on Win2000 with ODBC connection to Zope with ZODBCDA. Could it be the single threadedness of either the DA or Win2K ODBC driver or Access2000 itself?? We think we have narrowed it down to the database causing the slowdown. Considered a move to MySQL as it is installed. Would the ZODBCDA to MySQL cause similar problems? Better to use the native My SQL DA? Anyone else running into same problems? Have any options to try? Thanks Allen -----Original Message----- From: Tim McLaughlin [mailto:tim@BCSWebservices.net] Sent: Tuesday, April 10, 2001 7:51 AM To: 'andym(a)activestate.com'; 'zope-dev(a)zope.org' Subject: Re: [Zope-dev] ZODBCA questions... Andy- We've had no probs with ZODBCDA connected to SQL Server (7 and 2000). What leads you to think it's a prob with the DA? Tim Message: 2 From: "Andy McKay" <andym(a)ActiveState.com> To: <zope-dev(a)zope.org> Date: Mon, 9 Apr 2001 10:45:55 -0700 Subject: [Zope-dev] ZODBCA questions... Im getting some performance issues with my sites, im connecting Zope to a SQL Server (not using ZEO). I've got suspicion, but no proof, that its ZODBCA that is the problem. Does anyone have experiences with this on large sites and / or a suggestion to replace it? Cheers. -- Andy McKay. _______________________________________________ Zope-Dev maillist - Zope-Dev(a)zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )

2 1

Re: [Zope-dev] Proxy Roles Problem
by Dieter Maurer 10 Apr '01

10 Apr '01

Chris Withers writes: > .... > Anyway, it appears that proxy roles are [not] 'carried on'.... > > What gives? Is this how its intended? It is intended. Was changed with the big Zope 2.2 security policy shakeup. You may find some explanation for the change on zope.org. Dieter

2 1

RE: your mail
by Randall F. Kern 10 Apr '01

10 Apr '01

Read committed is faster, but I have not done any tests to know how much faster. Read committed mode doesn't support concurrent updates either, but fails silently. In other words, given to concurrent queries: update foo set a = a + 1 update foo set a = a + 2 Assuming these overlap, and a starting value of 0 for "a", read committed mode would result in "a" either being 1 or 2; serializable mode (with my patch) would result in "a" getting set to 3. -Randy > -----Original Message----- > From: Federico Di Gregorio [mailto:fog@mixadlive.com] > Sent: Monday, April 09, 2001 1:23 AM > To: Randall F. Kern > Cc: zope-dev(a)zope.org > Subject: Re: your mail > > > Scavenging the mail folder uncovered Randall F. Kern's letter: > > [snip] > > Concurrent updates are not supported in the serializable transaction > > mode; if one transaction tries to update (or delete) a row that has > > already been updated by a concurrent uncommitted > transaction, the second > > transaction will fail with a "concurrent update" error. > > mmm... are they supported in the read commited mode? if yes it will be > easy to add this mode as on option to psycopg cursors and use > it in zope. > do you know how much faster would it be compared to the > serialized mode? > > > Given that the default transaction mode for these DAs is > Serializable, > > and Zope has this nifty support for ConflictErrors and retrying > > transactions already, I propose Postgres DAs should convert the > > concurrent update serialization errors into ConflictErrors > > automatically. > > > > I've been running this way for a few days, and everything > seems great. > > Any reasons why this might be good/bad/ugly very welcome. > (Like: is it > > safe to throw a ConflictError at any random time during processing? > > >From a quick glance at ZODB it seems they are only thrown > at commit time > > now) > > > > Here is a small patch to ZPsyccopgDA 0.5.1 that implements > this change: > > thank you very much for the patch. i am waiting to know what > others think > about it (i am not a ZODB expert) before we incorporate it. > > ciao, > federico > > -- > Federico Di Gregorio > MIXAD LIVE Chief of Research & Technology > fog(a)mixadlive.com > Debian GNU/Linux Developer & Italian Press Contact > fog(a)debian.org > Don't dream it. Be it. -- Dr. > Frank'n'further >

2 1

I am looking for a DragDrop target.
by robert at rocad 10 Apr '01

10 Apr '01

Hi there, I would like to be able to allow people to upload files to a zope site by dropping them on a drop target someplace on a webbrowser showing the zope site. Can anyone point me towards a solution. Is there an example someplace? Thank you Robert

3 2

standard_error_message manipulation
by Dyon Balding 10 Apr '01

10 Apr '01

Hi, As part of a large python product I wanted to modify the normal "Zope Error" standard_error_message page. Most exceptions will be caught in the code, and therefore those that do get through and cause an error message to be displayed to the user probably indicate a bug in our code. I would like to give the user the ability to submit the details of the exception along with their own information as a bug report. I created my own standard_error_message, and wanted it to contain a form with the values of error_value and error_tb among others. It however appears that the values of these variables get rendered into HTML much earlier than desired and therefore aren't very readable in a bug report. Is there a better approach to this problem? or a work around? thanks -d -- Dyon Balding <dyon(a)devcoder.com.au>

2 3

Objects, Roles and Permissions
by Danny William Adair 10 Apr '01

10 Apr '01

Hello warriors, what's the fastest way to find out whether a specific role has a specific permission on a specific object? What's the fastest way to find out if it's acquired? Even more interesting: How would I find out whether a specific role has a specific permission on a specific object, _taking_acquired_rights_into_account_??? In other words (management screen): "View" permission is not checked for the Anonymous role, but acquired. Therefore the "Anonymous" role maybe _does_ have the permission to "View" this object, through acquisition. Would I have to climb up the ladder manually and check every parent until acquisition has been "turned off", or is there some function - even if it's only available for an external method - that already does this work? I looked through the ZMI pages (and the .py's under "lib/python/AccessControl") but couldn't find anything useful for this purpose. Iterating through permission_settings takes too long and I don't know how to access the specific roles/permissions directly (by name), mainly because mapping constructions like "p199r0" are not very handy... Thank you very much in advance for your help (dtml, pyhton, or docs I didn't see), Danny

2 2

ZODBCA questions...
by Andy McKay 10 Apr '01

10 Apr '01

Im getting some performance issues with my sites, im connecting Zope to a SQL Server (not using ZEO). I've got suspicion, but no proof, that its ZODBCA that is the problem. Does anyone have experiences with this on large sites and / or a suggestion to replace it? Cheers. -- Andy McKay.

2 1

ZPatterns sub-Specialist traversal
by Michael R. Bernstein 10 Apr '01

10 Apr '01

Hello all, I've been trying to follow the instructions for getting subSpecialist instances to appear as directly traversable from Specialist instances, but have not yet had any success. Here is what I've done: ArchiveImages (Specialist) defaultRack traversal_method (Python Script) traversal (SkinScript) Renderings (sub-Specialist) defaultRack traversal_method has 'REQUEST, name' as parameters, and the following body: return getattr(container.Renderings, name) The 'traversal' SkinScript has the following body: WITH SELF COMPUTE __bobo_traverse__=traversal_method Nevertheless, when I try to traverse an instance in the Specialist to an instance into the subspecialist, I get an error: http://localhost:8080/ArchiveImages/001/002 Resource not Found Traceback (innermost last): File /usr/local/zope/2-3-1/lib/python/ZPublisher/Publish.py, line 223, in publish_module File /usr/local/zope/2-3-1/lib/python/ZPublisher/Publish.py, line 187, in publish File /usr/local/zope/2-3-0/lib/python/Zope/__init__.py, line 221, in zpublisher_exception_hook (Object: 001) File /usr/local/zope/2-3-1/lib/python/ZPublisher/Publish.py, line 162, in publish File /usr/local/zope/2-3-1/lib/python/ZPublisher/BaseRequest.py, line 372, in traverse File /usr/local/zope/2-3-1/lib/python/ZPublisher/HTTPResponse.py, line 547, in notFoundError NotFound: (see above) The 002 sub-Specialist instance exists and is accessable normally (http://localhost:8080/ArchiveImages/Renderings/002) so I must be doing something wrong in the PythonScript or the SkinScript, but I don't know what. Any help would be appreciated, Michael Bernstein.

1 1

Re: [Zope-dev] 27 million objects.
by Sin Hang Kin 09 Apr '01

09 Apr '01

Concerning Zcatalog: 1. Search of vocabulary is fully matched, partial search, trancate search or fuzzy search will be useful for searcher. 2. If we catalog all things in one catalog, then we need to make filter when we want to search on one particular portation. 3. I would like to suggest : putting cataloger in branches and make them catalog only those they were interested. for example, if we put catalog in every branch of tree, then they catalog only that branch. catalog can be set to propagate their collection into higher level. This can be done in background. Then a search on a higher level will result in : 1. result in that catalog which meet the search specification and result from the lower level of those the higher level does not known as which match the search specification. In such settings, for example, we can arrange the catalog into multi-layer or have catalog specialize in one section. And have the propagate hierarchical which they update their upper level catalog in batch. The user is allow to search in a level and get all results up to that level and below. for example, in zope.org, we can put catalog in every document level, which catalog one document and one mother catalog which combine all catalogs from every document and one level up to handle all zope site. The we can search specifically in whatever level we want. It seems very difficult for all the catalog hierarchical to share their storage, so there will be some storage inefficent. But this can provide the necessary scalability. Rgs, Kent Sin --------------------------------- kentsin.weblogs.com kentsin.imeme.net

1 0

ZUserFolder usage and development
by Tim McLaughlin 09 Apr '01

09 Apr '01

Dave, 1. I say sync because UserFolder keeps its objects in an attribute called "data" (list of Users) whereas ObjectManager does not. The 'name' attribute is no longer necessary because I usee the id of the object and allowing it to be specified presents potential discord. 2. Correct, you're supposed to check for that in your constructor, but this is a new Product and could obviously be changed. In fact, I'm thinking of eliminating the constructor params altogether and allowing them to be set later. manage_editUser does check that password and confirm match since in fact it calls the Super. 3. I'm gonna pitch this out. say it all to the benefit of all. Incidentally, great working with you on this. Excellent feedback, and your catchin' on quickly (hooray! another infected by the Zope bug). Tim -----Original Message----- From: David Hassalevris To: Tim McLaughlin Sent: 4/9/01 4:20 PM Subject: Re: More progress - ignore last message Tim, (1) I notice that in zUserBase you omit "name" from def __init__ Your comment says # name parameter allows for id and name to unsync # so it is no longer Can you clarify this? Note: I noticed that the standard acl-user folder throws an error if duplicate name is added. Wouldnt that "default" behavior protect the linkage (or sync) between your class and its related user object? (2) I also notice that while I can sucessfully add a new user via z User Folder ... it omits password and confirmation checking. Is this me or is this in your code? (3) Trying to solved problems like this raises alot of interesting Zope questions. Do you mind if I throw some of the issues that occur to me out to the zope list? Best, David Tim McLaughlin wrote: > > cool. I was just about to form some vague conjectures as to what your > problem was... but good luck and lemme know if issues come up. > > incidentally, don't use 'name', 'roles', 'domains', or 'authenticate' as > property names or methods (clash with base class). > > Thanks, > Tim > > -----Original Message----- > From: David Hassalevris [mailto:bluepaul@earthlink.net] > Sent: Sunday, April 08, 2001 1:21 PM > To: Tim McLaughlin > Subject: More progress - ignore last message > > Tim, > It finally occured to me that in order to obtain User attributes not > stored in a "property sheet" I simply had to use things like > "getUserName()" > Ring! methods from the base classes are available! > > Ill update you as things progress. > David > > Tim McLaughlin wrote: > > > > David, > > > > Try it again if you have previously (0.8.0). Includes more instructions > > (thanks to some testing of my collegues) and fixes. > >

1 0