- Zope - Zope lists

Is there an APE mailing list?
by Johan Carlsson 20 Oct '03

20 Oct '03

Hi Shane, Is there an APE mailing list? If not, shouldn't it be one :-) Best Reagrds, Johan Carlsson PS. APE - Great work (as always :-) -- Johan Carlsson Tel: + 46 8 31 24 94 Colliberty Mob: + 46 70 558 25 24 Torsgatan 72 Email: johanc(a)easypublisher.com SE-113 37 STOCKHOLM

2 1

how to fix this code problem?
by Jianping Zhu 20 Oct '03

20 Oct '03

''''code 1 index_html <html> <body> <table > <tr tal:repeat="item container/objectValues"> <td tal:content="item/getId">Sample.tgz</td> </tr> </table> </body> </html> with above code when i hit test index_html i get following result which are all files and sub-directoy on current folder index_html ''' a file myfiles ''' a file Files ''' a folder I am now tring to use python script to show all file I put two files on current diectroy '''' index_html <html> <body> <table > <tr tal:repeat="item container/myfiles"> <td tal:content="item/getId">Sample.tgz</td> </tr> </table> </body> </html> and a python file '''myfiles return container.objectValues('File') when i test index_html i expect to get a list of two files index.html and myfiles. but disponitedly it return nothing!! how can i fix this problem? Thanks

3 2

RE: [Zope] Corrupt database after packing with Zope 2.6.2
by Bjorn Stabell 20 Oct '03

20 Oct '03

Douwe wrote: >> I am sure, the ZODB developers are very intersted in this >> "Data.fs.old". >> >> I am interested as well, provided it is small (less than >> about 500 kb when bzip2 compressed). > > Well, it is +- 200Mb uncompressed, so 500 seems to be out of > reach. How do I get it to the zopdb developers? Ours is 170MB uncompressed. But I'll reconfirm that we still have the problem first; we haven't tried packing for a few days. We store all text as unicode... not sure if that'll help.

1 0

long response times?
by Andrew Altepeter 20 Oct '03

20 Oct '03

Hey All! I've developed an alumni directory application. This directory is accessed through a search interface, and pulls the data from ldap. LDAP can't do an offset+limit to return a subset of the results, so I'm returning all results and storing them in request.SESSION. The results are display 10 / page, with links to move forward and back. The responsiveness is fine if there are few results (< 8 seconds for < 100 results). However if the set is large ( > 100), response time grows enormously. One search I do that returns 402 results takes about 30 seconds! I find this strange, because performing the search and rendering the entire page (template and all) takes < 5 seconds. Any ideas why it is taking so long? It seems there is some sort of house-cleaning being done after the page is rendered and before it is sent to the browser. Thanks, Andy

2 2

Re: Re: can't pack Data.fs due to time-stamp error
by martin.van.nijnatten＠informa.com 20 Oct '03

20 Oct '03

> >I'm trying to move a site to a different computer. > >Copied (with -p) the Data.fs, started the new Zope (2.6.2). > >Everything seems fine, but when trying to pack I get: > > > >Error Type: CorruptedError > >Error Value: /usr/local/Zope-2.6.2-src/var/Data.fs:5733365: > >time-stamp reduction: 03431be874d772d5 <= 034be54b5ea3dedd > > > > > It's possible that your new computer's time is screwed up relative to > the previous machine. > > Set the time on the new machine just ahead of that on the old. Might > work. Later you'll want to set your time to the "correct" value. Take a > look at ntpdate for that. Solved it by running fsrecover.py on the orgininal Data.fs

1 0

(no subject)
by Manish 20 Oct '03

20 Oct '03

I am a web designer and working for last 5 yrs in the same field. i would like to learn content management solutions. do u think it will be helpful to enhance my career. may i know where exactly i can go for a training at bangalore. thanking you, Manish Mishra Executive - Website Development Corporate Communications The Himalaya Drug Company Makali Bangalore - 562123 Karnataka India. Ph: 91 080 3714444-65 (551 Extn.) Mo: 94482 25925 website: <http://www.himalayahealthcare.com/> http://www.himalayahealthcare.com

1 0

Re: A few more dumb questions...
by Dylan Reinhardt 19 Oct '03

19 Oct '03

[forwarded to list] On Fri, 2003-10-17 at 06:43, Ted holden wrote: > Dylan, First off, please post to the list if you have questions. You may have a question that someone else is better-qualified to answer and I sometimes take weekends off. :-) > The method you suggest of using a statement like: > > thank_you = DTMLFile('thanks', globals()) > > and then returning self.thank_you from the function which loads the three > instance variables would work if I were to create a corresponding dtml file. So far so good. > > The idea of returning a handle to an existing dtml file rather than trying to > code a dtml file in python certainly makes sense in cases in which the > desired response is either static or can be expressed in terms of dtml code > and available variables. Or even in cases where you would like to use new dtml variables. Ex: some_dtml = DTMLFile('path/to/file', globals()) def some_method(self, request): """docstring""" request['new_var'] = 'some_value' return some_dtml > Is the thinking that virtually all desired responses can be expressed that > way, and there is never really a reason for having python code create > strings with html coding in them and return them? Basically. Templating languages are the proper tool for assembling and returning markup. Any non-trivial logic you're going to do in this scenario is done in Python. The degree to which you observe this distinction will determine, to large extent, the maintainability of your system. But there are gray areas... I usually find <select> controls easier to render in Python, as they tend to be logic-intensive. YMMV. > Another question: Having a user browse to a web page which involves a python > object clearly creates the object. Not typically. Some URLs call object-creating methods, most don't. > What happens when the user either > switches his browser to some other web site or turns off his modem? If the user isn't making a request, there is nothing for Zope to do. > Does > that somehow delete the object and, if not, does the developer need to > concern himself with garbage collection? Since objects are not typically generated by usage, there is usually nothing to collect. If you *are* generating new objects with your product, the answer is "it depends." In Python, objects are generally collected when they go out of scope or have zero references. But Zope's persistence machinery makes this less certain. If created objects are persisted, they should never get collected without some explicit deletion. Some non-persisted objects might loiter around until next time you refresh their container or restart the server. It might be helpful to hear more about why you believe you *are* creating objects & maybe we could guess what's happening with them. > What exactly is going on with the slash rather than a dot or pointer symbol > '->' in the form statement which activates the method directly from the form? > I noticed that using a dot '.' didn't work, and I've never seen that sort of > a usage in other languages. In fact, executing class methods using dtml-call > statements requires a dot. When you're calling a method on the server side, use a dot. When you're generating a URL for the client to hit, use a slash. > > The developers' guide gave an example of executing class methods which looked > like this: <snip> > > The guide doesn't appear to give an example of the usage you described, which > is more useful. What I described was for product code. If you wanted to return one dtml method from within another, just call it: <dtml-var other_method> > Last dumb question here: Is Zope powerful enough to handle a large online > forum such as FreeRepublic or LibertyPost? Probably... but you'll want a pretty serious setup to do it well. Do some searching on ZEO and "high availability" for past discussion. HTH, Dylan

1 0

Zope and garbage collection
by Ted holden 19 Oct '03

19 Oct '03

Dumb question here: Having a user browse to a web page which involves a python object clearly creates the object. What happens when the user either switches his browser to some other web site or turns off his modem? Does that somehow delete the object and, if not, does the developer need to concern himself with garbage collection? In other words, is garbage collection automatic with Zope?

3 2

tal:define limitations
by Michael Bernstein 19 Oct '03

19 Oct '03

I just stumbled across the fact that variables that are defined via tal:define apparently must conform to Python identifier conventions (must start with a letter or underscore, etc.). This doesn't seem to be mentioned anywhere in the documentation. However, URL arguments and form field names are not subject to the same limitations at all, which leaves me with no simple way to define default values in TAL for optional URL arguments or form fields that use a number as the identifier for a name/value pair. Example: http://example.com/sample?1=off I would like the following code to work: <html tal:define="1 request/1 | string:on"> This should have the effect of defaulting the argument to 'on' within the context of the page when the URL is http://example.com/sample but allowing it to be overridden via the URL parameter 1=off. However, when adding similar code to a Page Template, I get an error: Compilation failed TAL.TALDefs.TALError: invalid define syntax: '\n 1 request/1 | string:on', at line 7, column 3 My use-case is simple presentation logic, but this mismatch between allowable URL and form names on the one hand and allowable tal:define variables on the other is inconvenient, to say the least. Questions: - Is there another way of doing this simply (more simply that populating the REQUEST directly with a python expression or script)? - Should the restrictions on tal:define variable names be relaxed? Thanks in advance, - Michael Bernstein

2 1

authenticated_user in Z SQL method not defined
by Christoph Landwehr 19 Oct '03

19 Oct '03

Hi, I'm using the exUserFolder together with a PostgreSQL-db to log in users. Works so far. With <dtml-var "AUTHENTICATED_USER.getUserName()"> I get the username which is the same as the field "username" in the db users. But I can not use AUTHENTICATED_USER in an sql-querry like SELECT foo FROM users WHERE username=<dtml-var "AUTHENTICATED_USER.getUserName()" sql_quote>; Why is AUTHENTICATED_USER not availabel in Z SQL methods? Traceback (innermost last): * Module ZPublisher.Publish, line 98, in publish * Module ZPublisher.mapply, line 88, in mapply * Module ZPublisher.Publish, line 39, in call_object * Module Shared.DC.ZRDB.DA, line 328, in manage_test * Module Shared.DC.ZRDB.DA, line 306, in manage_test * Module Shared.DC.ZRDB.DA, line 415, in __call__ * Module DocumentTemplate.DT_String, line 474, in __call__ * Module DocumentTemplate.DT_Var, line 221, in render * Module DocumentTemplate.DT_Util, line 201, in eval __traceback_info__: AUTHENTICATED_USER * Module <string>, line 2, in f Error, exceptions.NameError: global name 'AUTHENTICATED_USER' is not defined Regards Christoph

3 2