July 2003 - Zope-Dev

ZopeCTL from debian sid could it be included?
by kosh 28 Jul '03

28 Jul '03

I was chatting on irc with some people on #zope and some where interested in it because it does make managing many zope instances easy. You have config files for your instances and you can stop and start all instances from a single script. You can also create new instances, remove them, reset the access password etc. It makes it much easier to manage overall then a regular zope install however it still needs a little more work for some of its rough spots. I would like to see something like this as part of zope since currently zope is a bit of a pain to manage many instances on one machine and so lots of things have been developed to deal with that and having a system that is just part of zope to address that would help a lot from a maintenance standpoint. This one has the advantage of being written in python, having a simple config file format and it already works however any such system would be an improvement and I am sure debian would switch if zope gained a built in way to do this. http://packages.debian.org/unstable/admin/zopectl.html

2 1

Restart button and Products dir in Zope 2.7 beta 1
by PieterB 28 Jul '03

28 Jul '03

Hi, I tested todays Zope CVS checkout on Python 2.2.3 on FreeBSD 5.0. I noticed the following issues: (1) I noticed that the "restart" button is missing on the Control Panel. Is this intended behaviour for Zope 2.7? Or did I encounter a bug? There is still a Shutdown button, BTW. (2) I also tried to install some products in the INSTANCE_HOME/Products directory. There was a README.txt there telling: "Additional products for your Zope instance should be installed in this directory. A freshly created instance should only contain this README.txt file in this directory." I tried adding a product to this directory (Formulator from CVS), but it didn't loaded when I restarted the Zope instance (from commandline because of (1)). It loads fine if I move it to SOFTWARE_HOME/Products. Can anybody comment on these issues? Should I create bugreports in the collector? PieterB -- http://zwiki.org/PieterB

3 4

looking for zope consultant, DC area
by Craeg K Strong 28 Jul '03

28 Jul '03

Greetings, all: Regular zope-dev poster here with an unusual item... We are looking for a Zope developer in the Washington D.C. area; US citizenship required. Is this the right forum to post such a request? If not, please advise. Thanks, --Craeg

1 0

Re: CVS: Zope/lib/python/TAL - TALGenerator.py:1.65
by Philipp von Weitershausen 26 Jul '03

26 Jul '03

Chris McDonough wrote: > Update of /cvs-repository/Zope/lib/python/TAL > In directory cvs.zope.org:/tmp/cvs-serv15492/lib/python/TAL > > Modified Files: > TALGenerator.py > Log Message: > Merge TAL i18n fixes which break CMF to HEAD from 2.7 branch. The agreement Jim was basically that this backward compatability will only go in 2.7, so that 2.8 will be fully compatible with the spec, however break backward compatability. What now? Phil

3 4

breaking apart INSTANCE_HOME.
by Chris McDonough 26 Jul '03

26 Jul '03

Hi folks, Since we started configuration and installation geddon on Zope 2 (which has resulted in Zope 2.7), we've made a number of improvements in the flexibility of file placement behaviors: - You can specify in the config file where you want your lock and pid files to be placed. - You can specify in the config file where you want your filestorage files to be placed. - You can specify in the config file where you want each log file to be written. - You can specify in the config file any number of Products directory paths and these will be knitted together into the Products pseudopackage. This is good, because it makes Zope much friendlier to some filesystem hierarchy standards which require that you don't put library and volatile data in the same place (e.g. the Linux FHS). But it is still a bit monolithic. You need an instance home directory, which must contain (at least) these two directories: import Extensions I propose that we add two more options to the config file: import-directory extensions-directory These would both be multikeys which specify some number of directories that contained importable zexp files and external methods, respectively. This would allow us to not require any fixed instance home directory. Instead, each path required by each subsystem is specifiable by itself in the config file. I'm sure that utilizing these options in the config file will break things that rely on having a monolithic INSTANCE_HOME such as products that attempt to do something like "import_dir = os.path.join(INSTANCE_HOME, 'import'). So I propose that the stock Zope instance home install continue to follow the old pattern (where everything is installed into a single instance home directory), but we provide the advanced config file options for roll-your-own packagers and advanced users. I would like to do the same thing for the software home, but I haven't thought much about it yet. - C

4 5

Re: segfaults in cPersistence under 2.6
by Tim Peters 26 Jul '03

26 Jul '03

[Anthony Baxter] > *sigh* > > So I tried a fresh checkout on a different box, no crash. > Completely nuked old checkout (which was a new checkout > last night), checked out again, rebuilt, no crash. > > Beats me. I'm going to blame the electrician who took out > power to our floor this morning, because it's that or > cosmic rays. I hate these. Causes for stuff "like this" in the past included: + "make clean" not cleaning everything it should have cleaned. + Compiling not recompiling everything it should have recompiled. + Wild load or store in Zope or Python's C code. These *can* be extremely touchy, appearing to go away for weeks at a time, and never occurring on some peoples' machines. I've only seen one instance of that last one triggered by Zope, which turned out to be a bug in Python (fixed long ago) specific to the intersection of Python's new-style classes and Zope's ExtensionClass. Jeremy and I saw cases of ring corruption while putting together the experimental ZODB 3.3, but they were 100% reproducible on all platforms, and turned out to have clear causes (some of the C objects we backported from ZODB4 were getting ghostified by the ZODB4 code without removing themselves from the ZODB3 ring). So "ring corruption" rang a bell for me, but not one that makes any sense in your context. Keep your eyes open, Anthony! If it happens again, I want to know.

2 1

Server consolidation in the domain grandaschoolnet.it and Groupware suite for Zope/CMF/Plone
by Piero Giuseppe Goletto 26 Jul '03

26 Jul '03

Hello everybody, we are Roberto Negro, an Information Technology teacher in an High School in Cuneo (near Turin, Italy), and Piero Giuseppe Goletto, a software engineer and Linux expert. We are the project leaders of "GRANDASCHOOLNET". This project has the main purpose to leverage on some existing capabilities in the open source software field by which to create a network of all Schools in the Cuneo Province, which is one of the largest in Italy and where there is a demand for some advanced services both for the students and their families. 1. THE STORY SO FAR Initially we set up our portal following a "sure" and "consolidated path": Apache as a web server, adding PHP and Postgresql in order to deliver some dynamic content. You can see our portal at http://www.grandaschoolnet.it. The server is Linux-based, has had record uptimes, and we are glad to say that all in this project is made exclusively with open source software also by students. We added phpgroupware as a groupware solution in order to allow teacher and students to work together on projects. We presented the portal in a convention on March 14th and 15th 2003, here in Cuneo and that convention created broader interest especially in the content management area. So, we devised the need to have a real content management and content delivery solution in order to allow both teachers and students to create and publish content in a secure and user-friendly way. We chose Zope/CMF/Plone, on which we are migrating, and you can see the result of our efforts so far at http://www.grandaschoolnet.it:81/granda. We beg your pardon since the portal is only in Italian at the moment. Zope/CMF/Plone complies perfectly with our requirements, and since it is, in our opinion, an excellent product it formes the reference platform for our project. We will soon integrate in our Plone portal both Silva (Infrae's document management solution which is also open source) and FLE3 (e-learning application). Both are Zope based so this integration is a non-issue but we feel that it's interesting for you to know. 2. WHAT WE NEED TO DO First of all we need to consolidate and power-up our portal. This requires a tight integration between various components: Apache web server, Squid proxy server, Zope/CMF/Plone, AwStats as a log analyzer, and, more, consider that we have devised the opportunity for a Plone/OpenGroupWare integration which we will discuss in detail later. This is the reason why we crossposted this message on various mailing lists, since we feel it involves suggestions and contributions from various parts. We beg your pardon about that and we will make (and publish) an excerpt of all what we learn in this process on our portal. Our intention is to create an infrastructure similar to the one that Matt Hardy described in this document (see http://matthardy.us/mywiki/HighPerformanceZope) which he says he is working at. Although we won't implement at this stage a ZEO cluster like Matt will do, we think it is an opportunity we need to explore; we expect to get to serve users in the thousands in 1-2 years. Our idea, and we are kindly asking you for confirmation on feasibility and correctness of these details, is to activate the Zope/CMF/Plone and Apache (with mod_php, mod_kerberos, mod_ssl, mod_rewrite), FastCGI, Squid (using this as a proxy) and Postgresql on the SAME server in a way such that Apache/Squid will work as a cache and as an authentication mechanism based on Kerberos or on OpenLDAP. We could estimate a need for at least 256 Mbytes RAM on this particular solution (the server is an assembled Pentium IV standard machine with 40 GBytes Hard disk; the kernel is Linux 2.4.20 and distribution is a standard RedHat 8.0 in which Zope has been installed from tarballs). At this very moment neither Apache nor Squid nor FastCGI are installed on this server. 3. THE GROUPWARE COMPONENT We read an interesting article from Reuven Lerner, published by Linux Journal, from we quote: "The latest example of a killer app in the Open Source world is Plone [...]. Plone has become an increasingly prominent piece of software in recent months and has brought many people into the Zope community. And [...] Plone may well be the killer app for Zope." In our opinion Plone is the killer application for Zope but we feel that Plone's success would be much easier if a complete groupware solution is available for it. We have recently (in fact, last week) read about the OpenGroupWare project and we feel it very promising, and we think that Open Groupware is an easy path is to create simply the Plone interfaces to OpenGroupWare and get, that way, the specialized solution integrated with Plone. 3.1 How would we do this? We feel that such a project, released as open source as a Plone app, would be a no-nonsense contribution both for Plone (and by the way for CMF and Zope) and for OpenGroupWare. There would be a single Plone "product" for every OpenGroupWare functionality, which would interface to OpenGroupWare as an XML-RPC client. The user interface would be integrated in the Plone Look and Feel. Of course all this will be released with the same license as Plone. As we ask your opinion on this subject, we feel that it makes sense to adopt the "next generation" releases. We read that the beta of release 2.7 is out, so is it better for us to wait for the completion of next fall's Zope 2 and Zope 3X releases, since, we read, next fall's Zope 2 will be the basis for the migration from Zope 2 to Zope 3, instead of starting with Zope 2.7? Again, should we stick to CMF 2.0 (which relies on Zope 2.7, is it correct)? What are the exact relationship between Plone 1.1 and the Zope and CMF releases? We thank you all for your kindness and patience and for the explainations of our mistakes you certainly will give us, and hope to read from you soon. Best regards Roberto Negro (roberto.negro(a)multiwire.net) Piero Giuseppe Goletto (piero.goletto(a)tin.it)

3 2

segfaults in cPersistence under 2.6
by Anthony Baxter 25 Jul '03

25 Jul '03

Wow. I'm having _so_ much fun with 2.6. I'm now seeing a reproducible segfault on startup. Program received signal SIGSEGV, Segmentation fault. 0xfe4517c8 in accessed (self=0xedbe58) at ZODB/cPersistence.c:160 160 self->ring.next->prev = self->ring.prev; (gdb) p self $1 = (cPersistentObject *) 0xedbe58 (gdb) p self->ring $2 = {prev = 0xe5d8fc, next = 0xeb7b210c} (gdb) p self->ring.prev $3 = (struct CPersistentRing_struct *) 0xe5d8fc (gdb) p self->ring.next $4 = (struct CPersistentRing_struct *) 0xeb7b210c (gdb) p self->ring.next->prev Cannot access memory at address 0xeb7b210c. #0 0xfe4517c8 in accessed (self=0xedbe58) at ZODB/cPersistence.c:160 #1 0xfe4526b4 in Per_getattr (self=0xedbe58, oname=0xbf280, name=0xbf295 "as_key", getattrf=0xfe6d7cec <EC_findiattro>) at ZODB/cPersistence.c:524 #2 0xfe452774 in Per_getattro (self=0xedbe58, name=0xbf280) at ZODB/cPersistence.c:539 #3 0x64bd8 in PyObject_GetAttr (v=0xedbe58, name=0xbf280) at Objects/object.c:1052 #4 0x21c34 in eval_code2 (co=0x8090f8, globals=0x49daac, locals=0x8, args=0xbf280, argcount=2, kws=0x0, kwcount=0, defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:1773 #5 0x245b8 in call_eval_code2 (func=0x80ac24, arg=0xd9230c, kw=0x0) at Python/ceval.c:2979 #6 0x24108 in call_object (func=0x80ac24, arg=0xd9230c, kw=0x0) at Python/ceval.c:2818 #7 0x23f64 in PyEval_CallObjectWithKeywords (func=0x80ac24, arg=0xd9230c, kw=0x0) at Python/ceval.c:2753 #8 0xfe6d4614 in callMethodWithPossibleHook (inst=0xea55d8, meth=0x80ac24, args=0xd9230c, kw=0x0) The code in question is here: /* Do nothing unless the object is in a cache and not a ghost. */ if (self->cache && self->state >= 0) { CPersistentRing *home = &self->cache->ring_home; self->ring.prev->next = self->ring.next; >>> self->ring.next->prev = self->ring.prev; self->ring.next = home; self->ring.prev = home->prev; home->prev->next = &self->ring; home->prev = &self->ring; } So it seems that the ring's getting corrupted in some way. Any tips on tracking this down? I'm thinking of a simple function that just zips around the ring checking that it's sane in both directions... (yes, I've confirmed that all the C extensions have been rebuilt cleanly). -- Anthony Baxter <anthony(a)interlink.com.au> It's never too late to have a happy childhood.

3 4

Re: i18n:attributes in 2.7
by Godefroid Chapelle 25 Jul '03

25 Jul '03

Godefroid Chapelle wrote: > To summarize, as message ids translation for attributes was added, 2.7 > and 3 are working according to the doc/spec, > but 2.7 breaks backward compatibility to 2.6.1. > > This leads to the following question : do we (1) accept this rupture in > i18n:attributes management or do we (2) want 2.7 to cover space > separated attributes ? > > IOW, could people that need backward compatibility stand up ? Plone 1.1 needs backwards compatibility. Plone guys would like not to have to maintain two versions for 2.6 and 2.7 This implies that we need to implement the second possibility : 2.7 must accept space separated attributes. > > BTW, second choice would lead us to a bad situation where we cannot > discriminate between > > <span value="value" title="title" > i18n:attributes="value title" /> > > and > > <span value="value" title="title" > i18n:attributes="title msgtitle" /> Quoting Jim on #zope3-dev : > ok, then I suggest: > In the case where there are other than 2, there is no ambiguity > If there is a semicolon, then use the new rule > "foo title;" -> title is message id > so if there are two things: "foo bar" > if *bar* is an attr, then: > Translate it and issue a depreication warning using the warnings module > if *bar* if not an attr: > Treat it as a message id. IOW if space separated and there are more than two items, all are treated as attributes plus deprecation warning if space separated and two items : - i18n:attributes="value msg_id;" (with semicolon) will always be treated as <attr> <msg_id> - i18n:attributes="value title" will be treated as <attr> <attr> as long as title exist as attribute plus deprecation warning -- Godefroid Chapelle BubbleNet sprl rue Victor Horta, 18 / 202 1348 Louvain-la-Neuve Belgium Tel + 32 (10) 459901 TVA 467 093 008 RC Niv 49849

1 0

Ninth Zope 3 Newsletter
by Gary Poster 25 Jul '03

25 Jul '03

WELCOME TO THE ZOPE 3 NEWSLETTER: ISSUE 9 (25 JULY 2003) Welcome to the ninth Zope 3 newsletter. Information about Zope 3 and newsletter contributions and suggestions can be found at the bottom of this newsletter. GLOSSARY FOR THE UNINITIATED Encounter a term in this newsletter you don't know? Try this glossary: http://dev.zope.org/Zope3/NewsletterGlossary NEWS SNIPPETS: - Jim Fulton released the third Zope X3 milestone release on June 30. A few "geddon"s (big changes to Zope 3) are still planned before the beta release. - The Australian OzZope Zope 3 sprint, organized by Jan Smith, was a big success. See http://www.ozzope.org/OzSprintWiki for a fun summary, and http://www.ozzope.org/news/LinMagAu for a link to a nice explanatory interview about Zope, Zope 3, and sprints with Jan. FRED DRAKE: Vocabulary Fields and Widgets A vocabulary is an object which represents a collection of distinct values (where distinct means that any two of them compare as not equal using !=, and no two compare equal using ==). These are most interesting when the set of values is highly dynamic. A vocabulary may be provided by (for example) some object in the ZODB, a query to an external database, the contents of a file maintained by another process, or static data provided by a separately maintained piece of code. Vocabulary fields provide a field type for situations in which the set of options is defined by a vocabulary. This is most useful when the set of possible values may be highly dynamic or otherwise be decoupled from the schema definition itself. A vocabulary field in a schema contains either a vocabulary object (``IBaseVocabulary``) or the name of a vocabulary. If a name is used, the specific vocabulary to use will be supplied by a registry when the schema field is bound to an instance of a content object. Entries in the registry may be made through an API or using ZCML. More information is available in the Zope 3 source tree in the file doc/schema/vocabularies.txt. This is an especially important document to read before creating or configuring widgets to use with vocabulary fields. ZConfig for Zope 3 (Submitted June 30) Zope 3 will be using a configuration file defined by a ZConfig schema for site administration. This change is intended to make the operation of a site easier for system administrators to work with. Additional configuration parameters will be added to the configuration file during the continuing development of Zope 3. PHILIPP VON WEITERSHAUSEN (submitted July 1): XML: After Martijn Faassen and I had implemented a first bit of XML in Zope3 in Louvain-la-Neuve, I thought it was time again to advance things in that department. After discussing things with Martijn on IRC, I moved all zopeproducts concerning XML to one zopeproducts.xml package. The ParsedXML DOM implementation, which Martijn had ported to Zope3 looked like the most advanced part of the whole xml package. It was not yet ready to be used as components though, so I wrote interfaces for the DOM, following the W3C specs tightly, and added ZCML configuration directives for the classes. I also attempted to implement something like ParsedXML in the form of DOM Document, which was simply a persistent document node. While providing views for it and simple adapters to and fro IXMLText (XML contained in a string) was - thanks to the CA - quite easy, trying to implement editing capabilities turned out to be a trip with much ContextWrapper pain, so I finally had to give up realizing that ContextWrappers were not the right choice for a DOM implementation. Martijn had used them because ParsedXML's DOM implementation used acquisition and ContextWrappers are its successors. It is therefore questionable whether we will hang onto this DOM implementation. It is also questionable whether we will have to implement our own DOM at all. In order to store a DOM persistently, we will most certainly have to, but it would be wise to research use cases first before starting to re-implement the whole DOM API. XHTML compliance ("xhtml-gheddon"?): When browsing through some page templates I realized that a lot of the times, XHTML compliancy was neglected. Thus I validated all page templates against XHTML 1.0 Transitional with the xmllint command line tool from http://xmlsoft.org/ and corrected all obvious errors that it reported. Unfortunately, the tool was very disturbed by the TAL and METAL namespaces and the flood of error messages was enormous, so I might have missed a few errors. I am yet looking for a solution to this problem, so that we can maintain XHTML compliance by validating page templates once in a while. For all emacs users: I used emacs's grep-find mode entering the following command:: find . -name "*pt" -exec xmllint --noout --dtdvalid \ http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd {} \; The grep-find mode can parse this output. Instead of specifying the URL, one can also download the DTD to speed up the process. Zope products: I kept a few zopeproducts up to date with the changes in Zope3 head (package refactorings, new-style implements, configuration -> registration). Most time I spent on the NewsSite product which was grossly out of date. I got it to work so that I could add a news site and news entries, but it's not pretty. This question goes out to both the initial developers (participants of the DZUG sprint) and other people interested in this product: is anyone willing to take further care of this product? If it is only kept up to date with changes in Zope3, it will remain in its miserable state for eternity. Then we might just as well remove it. CHRISTIAN THEUNE: Zope Security evaluation gains speed The Zope Security Evaluation aims at getting Zope 3 security mechanisms certified with a "common criteria certificate" which is an approved certificate within 15 countries. (http://www.commoncriteria.org) During the 9. and 10. july there was a workshop at the TUV-IT (http://www.tuev-it.de) which included Steve Alexander, Aroldo Souza-Leite, 3 people from the TUV and me. We have begun explaining them the Zope 3 security model which produced a fair explicit layout of components that are involved and started writing the documentation of the security target which means determining the components that belong to security, possible threats to them as well as selecting functional and environmental security requirements and objectives to defend against those threats. We gained a pretty good picture of what the certificate will give us (better understanding of the whole complexity of the security machinery) and that even the entry level certification means a lot of explanation work and in-front work I now continue writing the security target and am aiming to finish it before end of june (which is some kind of unlikely but possible). Everybody who likes to see the progress of the certification work can find a page including the schedule in the "Subprojects" area of the Zope3 Wiki named "SecurityEvaluation". Also there are first checkins within the Zope 3 file tree to find under Zope3/doc/security. The documentation format is reStructuredText and is additionally checked in as a rendered .html file. Stay Zope3ed. STEPHAN RICHTER Documentation I have made significant progress with the Zope 3 Development Cookbook. The "Content Components" section with 12 recipes (roughly 124 pages) is now done, some in outline form and others fully written out. The source code can be found at ZopeProducts/demo/messageboard. The recipes build a message board product from scratch and start out relatively easy and become then more involved. There are a couple other recipes dispersed throughout the book outline, which you can find at http://dev.zope.org/Zope3/DevelCookbook. For everyone who does not know yet, the Zope 3 Python Developers Cookbook will be published by Sams (same people as New Riders who did the Zope Book). I will be the primary author, and Jim and Steve signed up as technical editors. The book will be published under an open content license (http:// creativecommons.org/licenses/by-nd-nc/1.0/) Please send me or post comments about the recipes. The more early feedback I have, the better the book will be! Online Help While writing the recipes, I also fixed the Online Help to look for views correctly. One of the above mentioned recipes (http://dev.zope.org/Zope3/ OnlineHelpForZopeApps) has detailed information on how to use it. The Zope 3 Development team would like to encourage people to start developing help screens. SQLExpr Alan Runyan had the great idea to support a TALES expression that would evaluate SQL statements. Short story: He told me, 2 hours later it was done. The syntax looks as follows:: <html tal:define="rdb string:PsycopgDA; dsn string:dbi://test"> <body> <ul tal:define="name string:Stephan; table string:contact"> <li tal:repeat=" contact sql: SELECT * FROM ${table} WHERE name = '${name}'"> <b tal:content="contact/name" /> </li> </ul> </body> </html> The add-on is mainly aimed at scripters, who do not want to worry about design so much. You can find the product at http://cvs.zope.org/ZopeProducts/ sqlexpr. BTW, I think it would be easy to port this code to Zope 2, since very little is Zope 3 specific. Making zope3.org our Home It has been in our heads since last year, but we did not have the time to do it (other than registering the domain). But now it is time! We have a ZWiki product and soon I will also have a Bug Collector ready for Zope 3, so that we can start moving the content over to www.zope3.org. I hope to be able to move there in the next weeks. CONTRIBUTING Please send Zope 3 news, and newsletter suggestions and requests, to gary(a)zope.com, with "Zope 3 newsletter" somewhere in the subject line. As you can see above, casual and quick news items are acceptable and even desirable. MORE INFORMATION ON ZOPE 3 The central place to find Zope 3 information is currently http://dev.zope.org/Zope3 The Zope 3 mailing list is archived and managed at http://lists.zope.org/mailman/listinfo/zope3-dev The Zope 3 development IRC channel, #zope3-dev at irc.openprojects.net is (strictly) for discussion of Zope 3 development issues. Newsletters are (supposed to be, blame the editor) archived at http://dev.zope.org/Zope3/Zope3Newsletter

1 0