- Zope - Zope lists

problem migrating ZClasses from Zope 2.6.1 to Zope 2.9.8
by Yuri 27 Mar '08

27 Mar '08

Hi! I've got a trouble importing a ZClass app to 2.9.8 from 2.6.1. First, I've tried 2.8, and it worked but had the __new__ problem, so I've seen it is fixed in 2.9.8. After the upgrade, all seemed to work. My ZClass is a container (with ExtFile). I've been able to create a new instance, then had trouble after uploading an ExtFile. The file is there, in the repository, but when I clicked over it, I get a fast scrolling in the log. After the restart, the ZClass instances and the Product in the control panel, gived me a maximun recursion depth, here the last part: File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 597, in getState return unpickler.load() File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 471, in _persiste nt_load return self.load_oid(reference) File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 537, in load_oid return self._conn.get(oid) File "/usr/lib/zope2.9.8/lib/python/ZODB/Connection.py", line 205, in get obj._p_jar = self File "/usr/lib/zope2.9.8/lib/python/ZClasses/_pmc.py", line 65, in __setattr__ super(ZClassPersistentMetaClass, self).__setattr__(name, v) File "/usr/lib/zope2.9.8/lib/python/ZODB/persistentclass.py", line 84, in __se t__ jar.setstate(inst) File "/usr/lib/zope2.9.8/lib/python/ZODB/Connection.py", line 732, in setstate self._setstate(obj) File "/usr/lib/zope2.9.8/lib/python/ZODB/Connection.py", line 786, in _setstat e self._reader.setGhostState(obj, p) File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 604, in setGhostS tate state = self.getState(pickle) File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 597, in getState return unpickler.load() File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 471, in _persiste nt_load return self.load_oid(reference) File "/usr/lib/zope2.9.8/lib/python/ZODB/serialize.py", line 537, in load_oid return self._conn.get(oid) File "/usr/lib/zope2.9.8/lib/python/ZODB/Connection.py", line 205, in get obj._p_jar = self File "/usr/lib/zope2.9.8/lib/python/ZClasses/_pmc.py", line 65, in __setattr__ super(ZClassPersistentMetaClass, self).__setattr__(name, v) File "/usr/lib/zope2.9.8/lib/python/ZODB/persistentclass.py", line 84, in __se t__ jar.setstate(inst) File "/usr/lib/zope2.9.8/lib/python/ZODB/Connection.py", line 737, in setstate exc_info=sys.exc_info()) File "logging/__init__.py", line 999, in error File "logging/__init__.py", line 1078, in _log File "logging/__init__.py", line 1064, in makeRecord File "logging/__init__.py", line 226, in __init__ # Must happen after ZopeStarter.setupInitialLogging() RuntimeError: maximum recursion depth exceeded in cmp 2008-03-26 11:12:05 INFO Signals Caught signal SIGINT 2008-03-26 11:12:07 INFO Z2 Shutting down ======= Any idea?

7 11

running in debug mode
by David Bear 26 Mar '08

26 Mar '08

Is it possible to have one zeo client running in debug mode while another is not? It seems that it should be the way most people would make a development environment. One zeo is running in standard mode so you could see how the site performs -- while the other zeo is running in debug mode so you can do file system development. -- David Bear College of Public Programs/ASU 411 N Central, Phoenix, AZ 85004

5 7

Running multiple transactions during single request
by Charl Coetzee 26 Mar '08

26 Mar '08

Hi, I have an app under Plone (2.5.1, Zope 2.9.6, Five 1.4.3), with a browser view that gets fed csv data, and creates a bunch of objects (with some further processing) (essentially like a data import problem). For various reasons, while this view is doing its work (ie during course of the single request), it seems attractive to commit the added objects regularly - e.g. to commit after every 5 newly created objects, say. (as opposed to using savepoints). I would like to confirm if a view like so: from Products.Five.formlib import formbase import transaction class UploadForm( formbase.PageForm ): ... ... @form.action("Process") def process(self, action, data): ... while someLongLoop: ... # Create some persistent objects, # set some attributes on them, etc. ... if everyFifthPass: transaction.commit() transaction.begin() # end of while loop really does have the desired effect, where also ConflictError handling etc continue to work as expected, without nasty side effects etc. I've looked at ZPublisher/Publish.py that seems to do essentially this in the publish() function, except that this also does recordMetaData() { which method in turn carries the comment "Is this code needed?" -- but I'm guessing this may be related to the "Undo" ZMI functionality? Anyway my view is obviously not calling this repeatedly, is this cause for concern? } This is the deepest I've yet dug into the transaction machinery, so I thought it would be prudent to ask the question -- google doesn't have much to say on this one, Thanks in advance for any insights, Charl

2 3

Re: [Zope] problem migrating ZClasses from Zope 2.6.1 to Zope 2.9.8
by Yuri 26 Mar '08

26 Mar '08

> > On Mar 26, 2008, at 11:19 , Yuri wrote: >> > Hi! >> > >> > I've got a trouble importing a ZClass app to 2.9.8 from 2.6.1. >> > > IMHO you're not investing your time wisely trying to make a ZClas > application work on a later Zope version. They're unsupported and > dead. I would keep the old setup running and spend time on recreating > the ZClass product functionality as a real Zope product, and then > create a script to migrate the old ZClass instances over. > Sure, but then I would switch totally to zope3 techs, at least. The point is, why a working envinroment in zope, don't works anymore just restarting zope? It seems a database corruption or something not written in the disk, so it is worth an investigation. All works, you restart, boom. Can't recover in any way. At least there's should not mentioned any fix in 2.9.8, if it does not works. It is time comsuming for nothing, and kills any love with Zope, imho. Say: in this version this and this works, in that version not. Say "it works", then it don't, who knows.... > jens

2 2

ZClass problem again when zope restarts
by Yuri 26 Mar '08

26 Mar '08

Hi! redoing the process of upgrading, it has worked again. When I restart it with zopectl fg, it gives back the same error again. So there's something in the startup process which creates problem. It seems like something don't get written in the zodb, could be?

1 0

I would like to join your group; my membership is pending approval
by Jan Ulrich Hasecke 26 Mar '08

26 Mar '08

Hi there! -- DZUG e.V. (Deutschsprachige Zope User Group) http://www.zope.de

3 2

Google Summer of Code 2008 - Student looking for Python Zope ZODB Grok Project / Mentor
by Mali Ozbay 25 Mar '08

25 Mar '08

Hi there: I would love to see if there is anyone looking for a student to mentor to do some hacking in fields such as Python Zope ZODB Grok etc., during this summer, in the Google Summer of Code 2008. If you are interested in learning more about my background and experiences please see my resume on : http://denali.cs.uh.edu:9080/mali/re/Mehmet_Ozbay_Resume-3-25-08.htm You can also view my profile on Google Summer of Code 2008 web space ( http://code.google.com/soc/2008/) in the applicant students list. Sincerely Mali Ozbay University Of Houston - Computer Science

1 0

Question about dot in method name
by bard stéphane 25 Mar '08

25 Mar '08

hi every body i'm changing a very old zclass product. In this zclass i declare a python script with this name "mydoc.xls" this script return an xls file with an argument now i'm writing a zope2 product. How can I explain to zope that he should publish a certain method with a different name so when people type http://myIntranet/folder/mydoc.xls?fileName=hello.xls it point to a real python method (dot caracter are forbiden in python class method) mydoc.xls => mydoc_xls(REQUEST=None) I know that Zclasses are bad :), but I cannot change this URL, because lot's of people record it. thank's a lot for an advice.

1 0

ExtFile 1.5.6 bug
by Dragos Chirila 25 Mar '08

25 Mar '08

Hi, I am experiencing a very strange problem; any thoughts or ideas will be highly appreciated. I am using ExtFile 1.5.6 product to store images and files in my application. During the publishing process, at some point, a rename of a container with images and files it is involved. From time to time this results in broken objects with 0 byes in size. My application runs on a Zope 2.7.6. On the other hand, I am running the same kind of publishing process on a Zope 2.10.4 with ExtFile 2.0 and this problem never occours. I noticed a line in the CHANGES.TXT file: - Bugfix: Exceptions at a late stage of a cut/paste operation could lead to data loss. Can this be the bug that affects my application when I use the ExtFile 1.5.6 product? Because a migration to a higher Zope version is out of the question right now, is there a fix for this problem? Or is there another way to overcome this? Thank you very much, Dragos -- Dragos Chirila objectValues(a)gmail.com (+4) 0722 395375

1 0

Zope utf-8 problem using MySQL
by Hugo Ramos 23 Mar '08

23 Mar '08

Yellow ppl, I'm using Zope 2.10.5 + MySQL 5.0.51a + MySQL-python-1.2.2. Everything is set to utf-8. In MySQL server and client is set to utf-8 and utf8_general_ci In Zope zpublisher is set to utf-8 I even created the (in)famous sitecustomize.py in Python 2.4.4 site-packages The problem is that when retrieving data from a table the international chars become alien symbols... The only way I get the right symbols is using SET NAMES 'utf8' <dtml-var sql_delimiter> on top of each sql method. What am I missing? Thanks -- Hugo Ramos http://www.hugoramos.eu/ ramosh(a)gmail.com

3 8