October 2002 - Zope-Dev

Re: [Zope] A tip about 2GB Data.fs issues
by Johan Carlsson [EasyPublisher] 16 Oct '02

16 Oct '02

At 22:16 2002-10-15 +0100, Phil Harris said: >On Tue, 15 Oct 2002 14:54:41 -0400 >"Matthew T. Kromer" <matt(a)zope.com> wrote: > > > Just FYI, the Binary Zope 2.6 betas include a Python 2.1.3 with LFS > > built in, > >Does the large file support extend to Win32 systems? > >yours hoping-rather-than-expecting-ly Not until Python 2.2 :-( But with Python 2.2 it works fine. The downside is that Zope doesn't support Python 2.2 officially. Regards, Johan Carlsson -- Torped Strategi och Kommunikation AB Johan Carlsson johanc(a)easypublisher.com Mail: Birkagatan 9 SE-113 36 Stockholm Sweden Visit: Västmannagatan 67, Stockholm, Sweden Phone +46-(0)8-32 31 23 Fax +46-(0)8-32 31 83 Mobil +46-(0)70-558 25 24 http://www.easypublisher.com http://www.torped.se

1 1

Idea: Awaking Zope to Life
by Ulrich Eck 15 Oct '02

15 Oct '02

Hi there, I just had an Idea and want to hear your comments on the following: Zope is a Request-Based System. It has no internal scheduling functions except one installs Xron. I have written a MicroThread Scheduler using python2.2 generators that supply cooperative Multithreading. Now the point: I want to integrate this Scheduler as one Thread (like Xron) to Zope as "Service" where clients could register a callback with parameters which are called regularly with a certain priority. Zope's Persistent Object Database would morph to an completely persistent Program that handles requests too. One could use it to implement better Workflow capabilities, or an Event Sytem (a client would register a Thread that checks for incoming messages and handles them). Many more things could be done .. What are the pitfalls i can run into, when trying to implement this ?? I could think of concurrency issues, problem with threads, context, security .. any concrete hints ?? What do you think about that ?? Ulrich Eck ------------------------------------------------------------------------ net-labs Systemhaus GmbH Ebersberger Str. 46 85570 Markt Schwaben fon: +49-8121-4747-11 fax: +49-8121-4747-77 email: ueck(a)net-labs.de http://www.net-labs.de

5 6

Re: [Zope-dev] This was fun!!!
by Joachim Werner 15 Oct '02

15 Oct '02

I won't look into the DTML Parser code right now. Don't need no headache ;-) But this is my first guess: > > <dtml-if getEasyLanguageService> DTML looks at "getEasyLanguageService" and does an implicit "getEasyLanguageService()". So now "getEasyLanguageService" (as an attribute in the DTML namespace) represents the actual service object the method should return. > > <dtml-let ELS="getEasyLanguageService"> ELS represents the service object now > > <dtml-in getAllLanguages> > > <dtml-var "ELS.getLanguageByRFC3066(_['sequence-item'])" the getLanguageByRFC3066() method of the service object is called And here it fails because: > > <dtml-if "is_language_aware() and getEasyLanguageService()"> getEasyLanguageService() is called explicitly, and in the "", so DTML does no magic. It does NOT put the returned service object into the namespace as "getEasyLanguageService" (as it would do in the first example) > > <dtml-let ELS="getEasyLanguageService"> the method "getEasyLanguageService", not its return value (i.e. the language service object) is passed to ELS > > <dtml-in getAllLanguages> > > <dtml-var "ELS.getLanguageByRFC3066(_['sequence-item'])" This fails because Zope tries to get the getLanguageByRFC3066() method of the method (or function) getEasyLanguageService (without ()), and that is not allowed here. Yep Joachim

2 1

This was fun!!!
by Lennart Regebro 15 Oct '02

15 Oct '02

Here was a fun error I had. Try and figure out where the bug is! I had this dtml-code: <dtml-if getEasyLanguageService> <dtml-let ELS="getEasyLanguageService"> <dtml-in getAllLanguages> <dtml-var "ELS.getLanguageByRFC3066(_['sequence-item'])" html_quote> </dtml-in> </dtml-let> </dtml-if> This worked fine. I then changed it to this: <dtml-if "is_language_aware() and getEasyLanguageService()"> <dtml-let ELS="getEasyLanguageService"> <dtml-in getAllLanguages> <dtml-var "ELS.getLanguageByRFC3066(_['sequence-item'])" html_quote> </dtml-in> </dtml-let> </dtml-if> Now this fails, and it fails in: <dtml-var "ELS.getLanguageByRFC3066(_['sequence-item'])" html_quote> with an "function attributes not accessible in restricted mode" error! Now, where is the bug? :-) First answers get 5 ZopeZen coolness credit points. An explanation to why the first version works gets 100 points! :-) Best Regards Lennart Regebro, Torped http://www.easypublisher.com/

2 1

Re: Catalog Bugs?
by Casey Duncan 15 Oct '02

15 Oct '02

(continuing discussion on zope-dev) I am to blame for those changes which were motivated by the addition of a bunch of new indexes. The most significant of which is ZCTextIndex which has its own brand new kind of "Vocabulary" object (called a Lexicon, which is what the old thing actually was). We tried to use the old one or develop a compatible one, but it was beyond saving, even its API blew. Its one the worst pieces of cruft in Zope IMHO. That is why a vocabulary is no longer created. Also, the Catalog used to have a hand in managing the vocabulary (basically locating it for textindexes), this behavior harkens back to before plug-in indexes and was actually horribly broken, so I ripped it out. The other problem was that creating default indexes in the Catalog, although perhaps convenient for newbies, gets in the way much of the time. I found myself creating catalogs programmatically and then clearing out all the indexes and adding different ones all the time. Also, a few indexes namely summary and bobobase_modification_time were actually broken and/or useless. The latter is not something the application should use as an index source as many have discovered. Many of these indexes were also not useful in the CMF. Now this is not to say that adding some default indexes (as an option on catalog creation) would not be useful. Here are some suggestions: id FieldIndex title ZCTextIndex meta_type FieldIndex path PathIndex PrincipiaSearchSource ZCTextIndex Granted these are much the same as before minus a few. It would also create a ZCTextIndex Lexicon named "Lexicon" in the catalog for you and bind the text indexes to it. How does that sound? Anybody else want to suggest index possibilities? -Casey On Tue, 15 Oct 2002 07:02:36 -0400 Luke Tymowski <luke(a)seeto.com> wrote: > Hello Casey, > > > Just noticed that you said that you thought the Catalog was broken in > > 2.6.0b2. What is the problem you found? Is it in the collector? > > Previously, before 2.6.0, when you created a new search engine or > catalog, inside the catalog Zope would create for you a Vocabulary, > Indexes, and Metadata with default values. You could add to the > defaults. But the default was usable and you simply created search and > results forms. > > Now when you create a catalog you have to create the Vocabulary, > Indexes, and Metadata by hand. Then create the search and results forms > or use the "wizard" to create them for you. > > It's not a big deal for someone who is used to Zope, but for newbies > it's v difficult to get them up and running with a search engine - and > it was one of the beauties of Zope - you had a search engine built-in > that was just about ready to go, making it a much less frustrating > experience for newbies. > > Maybe I'm doing something wrong. > > Luke >

2 1

Formulator Objects and Error Messages
by Chris MacKenzie 15 Oct '02

15 Oct '02

Hi, Is it possible to output error messages generated by a formulator object in a region at the top of a page template rather than to a column on the form. Also is it possible to place the button where you want on the form. An online example would be cool if one exists regards Chris MacKenzie -----Original Message----- From: zpt-admin(a)zope.org [mailto:zpt-admin@zope.org]On Behalf Of Martijn Pieters Sent: 14 October 2002 14:50 To: Chris Withers Cc: zpt(a)zope.org Subject: [ZPT] Re: TAL and METAL in <script> and <style> On Mon, Oct 14, 2002 at 12:14:03PM +0100, Chris Withers wrote: > Well, taking this use case, the current behaviour is confusing. > > I'm a designer, I get this template I know I'm allowed to tweak but I have > no idea what this tal: stuff is. I do however, know a bit of javascript. > So, I happily tweak the <script> stuff ignoring the funny tal: bit until it > does what I want. I hand over the template to the scripter who does his > thing and hands off to the customer. I'm then confused as to why the > customer is unhappy until I find out that none of my javascript is working. > I wonder what to do next. > > So, if this use case is broken anyway, could we at least make it useful to > the coders and scripters of this world and allow tal: inside <script> and > <style> tags? ;-) I don't think it is broken; as a coder you always have to weary of changes to a template that break the templating code. Imagine a table row with a tal:repeat copied over to a new table for convenience of the designer. Any template editied by a non-coder needs inspection as an unrendered template anyway, as well as discussion with the designer as to what he/she expects will happen when it is templated. Page templates support the use case, but as any tool will not make it work by magic. -- Martijn Pieters | Software Engineer mailto:mj@zope.com | Zope Corporation http://www.zope.com/ | Creators of Zope http://www.zope.org/ --------------------------------------------- _______________________________________________ ZPT mailing list ZPT(a)zope.org http://lists.zope.org/mailman/listinfo/zpt

1 0

DateTime.rfc822() bug?
by Lennart Regebro 15 Oct '02

15 Oct '02

RFC 2822 (which is the currently valid one, if I understand correctly) specifies the date format to have four digit zone specifications, ie "GMT+0200", while DateTime.rfc822() happily returns "GMT+2". Not that this seems to be any problem, I'm just looking for an answer if this is how it's supposed to be? I discovered it, because Outlook Express handles RFC822 dates incorrectly and will display both the "Fri, 11 Oct 2002 13:47:03 GMT+0200" and "Date: Fri, 11 Oct 2002 13:47:03 GMT+2" as "2002-10-11 15:47". Obviously some programmer at Microsoft hasn't read the rfc properly. :-) "The date and time-of-day SHOULD express local time. The zone specifies the offset from Coordinated Universal Time (UTC, formerly referred to as "Greenwich Mean Time") that the date and time-of-day represent." While MS obviously thinks it represents GMT. Easy mistake to do, if you don't read the rfc... Best Regards Lennart Regebro, Torped http://www.easypublisher.com/

3 4

Zope RPMs/debs and Linux FHS
by Chris McDonough 14 Oct '02

14 Oct '02

Hi all, I am working towards a unified Zope configuration and installation system on a branch of Zope named the 'chrism-install-branch'. I have given the buildout process on that branch the ability to create an RPM distribution of Zope. I intend later to give the buildout process the ability to create Debian .debs as well and maybe Solaris packages... I am doing this with the expectation that we might be able to provide RPM and .deb distros of Zope from zope.org instead of our current "generic Linux as tarball" distro. I haven't looked yet at the Debian packaging of Zope (by Gregor Hoffleit), but I intend to do that next to get some more ideas. I know there are already at least two flavors of Zope RPMs which Jeff Rush helped to package. There are a number of differences between the packaging of the RPMs generated by my branch and the packaging of Jeff's RPMs: - One of Jeff's distros breaks Zope up into many different packages, while another installs it as one or two. Mine only has one distribution flavor: a single package. - Jeff's puts some stuff into the current prevailing python's site-packages directory and some other stuff into /usr/share/zope. Mine puts nothing into site-packages, and installs all Zope software into /opt/zope. - Jeff's creates an INSTANCE_HOME in /var/zope. Mine creates an INSTANCE_HOME in /var/opt/zope. I don't know if this is the right thing but in reading the Linux FHS, it advises to not create subdirectories of var directly... so I don't. - Jeff's puts pid files into /var/run, while mine creates pid files directly in INSTANCE_HOME/var. - Jeff's puts log files into /var/log while mine puts them into INSTANCE_HOME/var. I am wondering if: - anybody has opinions on the packaging layout. Why is it advantageous to have many packages rather than one? - anybody has any opinions of where Zope files distributed via RPMs and debs should really go, especially wrt to the Linux FHS. I'm not sure there is a right answer, but I don't know beans about this, so I figure I'll ask. A file named 'Zope.spec.in' is attached to this email which is the input file to create a Zope RPM spec file during the make process, to give a better idea of how this works. Thanks! - C

7 14

Security Testing
by Chris Withers 14 Oct '02

14 Oct '02

Hi, I'd like to build a suite of security tests for a product I'm writing using unittest.py. Is this possible? I thought about using newSecurityManager with various known users, and restrictedTraverse to get to the appropriate methods, but then how do I test if those methods are callable? cheers, Chris PS: How is all this being tackled in Zope 3?

3 2

Debugging conflict errors? Hints please!
by Stefan H. Holek 14 Oct '02

14 Oct '02

Hi All! I am experiencing a lot of 'ZODB conflict error at ...' that I can reliably reproduce by hitting my browser's Refresh button at a high rate. The bad news is that the conflicts happen on pages that are not supposed to change anything in the ZODB but only display results of some SQL queries. My question is now how to find out what actually is causing the conflicts. The log entry refers to the page being served ('/very/long/path/detail_view'). As a matter of fact I am not even sure whether the log entry means that the 'detail_view' document has changed (which of course it hasn't), or something along the traversal path caused a ZODB write. Zope 2.5.1, ZEO 1.0 Any hints on how I could find the culprit appreciated. Thanks, Stefan -- Those who write software only for pay should go hurt some other field. /Erik Naggum/

3 4