March 1999 - Zope - Zope lists

DTML Method truth value
by David Ascher 05 Mar '99

05 Mar '99

The following method tests false (correctly):   But if I change it to:     then it tests true (note that nothing happens in the #in block). Is this correct behavior or one for the Collector? --david

1 0

RE: [Zope] Can Zope*src/lib/python be copied to .../python1.5/sit e-packages?
by Paul Everitt 05 Mar '99

05 Mar '99

Skip writes: > Can the contents of Zope's lib/python directory safely be > copied to the > site-packages directory of the Python install directory? I asked Jim, who replied: "As far as I know, they can. Note, however, that it will be important for INSTANCE_HOME to be defined." --Paul

1 0

RE: [Zope] Simple DTML Method/Acquisition problem + bug report
by Michel Pelletier 05 Mar '99

05 Mar '99

> -----Original Message----- > From: David Ascher [mailto:da@ski.org] > Sent: Friday, March 05, 1999 1:52 PM > To: Michael Bernstein > Cc: The Zope Group > Subject: Re: [Zope] Simple DTML Method/Acquisition problem + > bug report > > > On Fri, 5 Mar 1999, Michael Bernstein wrote: > > > Try:  > > > > This will call the method using the root folders' namespace, thereby > > showing all qualifiying subfolders regardless of where the method is > > called from. You might have to modify your method to > construct absolute > > URL's. > > Thanks! What I actually wanted was sidebar(PARENTS[0], > REQUEST), but you > pointed me in the direction I needed. Now I have to > understand it, but > that's much easier with working code. > > This also led me to find a bug. I was wondering why the > REQUEST parameter > was needed, so I tried to remove it. The next time I loaded a page > calling this code, the ZServer process died with the Win32 > equivalent of a > segfault. Is it just me? I don't dare try it on another > Zope server =). > [Zope 1.10.2, latest ZServer, win32, NT4SP3] > Could you submit this bug report to the Collector? http://www.zope.org/Collector Thanks, -Michel > --david > > > _______________________________________________ > Zope maillist - Zope(a)zope.org > http://www.zope.org/mailman/listinfo/zope >

1 0

XML discussions
by Robert OConnor 05 Mar '99

05 Mar '99

Hi Paul and Zopesters, Paul writes in the thread: [Zope] Some tough questions about zope... >I've been awfully interested in XML in >the last few days. Can you describe what >user-visible benefits PHP delivers due to its integrated >XML parser? I found a great description about XML in: "20 Questions on XML" http://builder.com/Authoring/Xml20/index.html On Paul's Zope roadmap: (Jan 18th) http://www.zope.org/Information/Roadmap "(2) XML. Zope is a great fit for XML, even better with the machinery from Tabula going in. We also see XML becoming the basis for other initiatives, such as WebDAV and RDF. However, we just can't squeeze out the time to dive in deep. Rather, we'd love to see more community progress like that which popped up this weekend." I did a search for postings that had "XML" in the subject and found 55 messages in ~6 threads between Jan 8th and Feb 2nd Other messages have "XML" in the body. Hopefully we can get some OUTLINE of how ZOPE and XML can work together. --- A framework for further discussions. -bobo connor bob(a)rocnet.com

1 0

Simple DTML Method/Acquisition problem
by David Ascher 05 Mar '99

05 Mar '99

I have a standard_html_header/footer pair which works much like that on zope.org. The header inserts a  in the left column. The sidebar DTML method is:    <a href=""></a><p>   (show links to sub-folders which don't have the 'hidden' property set). The above works fine in the toplevel folder. In the subfolder, however, nothing shows up, even though there are two folders which do qualify. I'd tried originally to have it be a DTML Document and change the expr to _.objectValues(['Folder']), but that doesn't work at all. What am I missing? --david

5 6

evaluating attribute values
by Robin Dunn 05 Mar '99

05 Mar '99

How can I evaluate a #var equivallent for setting a tag attribute value? For example, Zope doesn't like this:  --> or this:  (it tries to use smtpHost as the host name, instead of its value) Help! -- Robin Dunn robin(a)AllDunn.com http://AllDunn.com/robin/ http://AllDunn.com/wxPython/ Check it out! Try http://AllDunn.com/laughworks/ for a good laugh.

4 7

Documents for web and paper
by Robert OConnor 05 Mar '99

05 Mar '99

I'm looking for the best way to store and retrieve documents with these requirements: SPECS *Easy storage and retrieval on the web *Pretty printing on paper *Easy simple collaboration (no checkout) *One BASE format for both web and paper *Base format *could be* edited with Word or WP *Preserve text markup: bullets, bolds, underlines, tables, graphics, links...etc. *Search by keyword *Search by full text ******* I look forward to comments and Ideas on implementing this in ZOPE Perhaps some of this has been solved already and I don't know it. EXAMPLE: Say you have a small book/article that is fully contained in a WORD or WP file. It has chapters and sections: Chap1 Plants Section1: Trees Section2: Flowers Section3: Vegetables Section4: Fruits Chap2 Animals Section1: Dogs Section2: Cats Section3: Fish Section4: Humans Ok, as a paper document in the word processor it is laid out *nice* with a new page for each chapter, section, and subheadings for groups of paragraphs in each section. It prints *nice* and can even have a table of contents and index. For the web, the whole book could be one LONNNG web page. Ugly and it doesn't work well. Ideally on the web each chapter and section would be navigable with a left hand frame like table a la http://www.zope.org. Then sections can be placed on separate pages on the right. This means that the one source document becomes multiple html "files" on the web site. *Tree Navigation Ok, for tree navigation like http://localhost:9673/QuickStart Each Chapter and Section becomes a folder and the text from each section is contained in the same named: "dtContent" as in ...9673/QuickStart/Outline/Applications/External/dtContent While the tree is nice, and separate html "files" make the website better, *REVISIONS with TWO OUTPUTS How do you do: Make revisions to the source document in word/WP and have it FLOW automatically to html format or tree format for the web? *Perhaps* XML could come to the rescue as holy grail to help solve this problem??? I understand the idea of mark up but how does it solve the problem of two outputs? Can you have a template that automatically chops up the SOURCE into the appropriate folders/files etc and feeds it to ZOPE? Can this or another template reformat for pretty printing including page numbers that are updated based on the planned printing? Maybe I need to reconsider how documents are stored as source, i.e. not Word. And if not, what tool or mechanism would be best? Maybe the NATIVE source could flow to Word/WP or via a template making for pretty printing. Maybe for simple stuff, the source could be simple text entered on a webpage or sent by e-mail with *simple* markup and parsed automatically? *KEYWORD SEARCH Ok, on the web (and Zope), the benefit is it's searchability: Do we "hide" keywords in the source somehow and in addition use Chapter and Section titles as Keywords? Search for: "dogs" gives a result of Chapter 2 Section 1 in the example as one entry. Additional entries might be lessor cross references such as a keyword "dog" in the Section 4 humans as "man's best friend". *FULL TEXT SEARCH Ok so we search for the word "Daisy" and this word is found in the text in sections: Flowers, Dogs (as a pet name) and Humans (as a proper name). The result is then three line entries that take you to the words. *Zope IDEAS for this? I assume that it would be best to store the documents and source in the ZOPE database rather than an external datasource. Thanks for any input on this -bobo connor bob(a)rocnet.com

1 0

Re: [Zope] Using Zope with CORBA (new user question)
by Jeffrey P Shell 05 Mar '99

05 Mar '99

>>>>>> "Paul" == Paul Everitt <Paul(a)digicool.com> writes: > > Paul> Jody wrote: > >> My basic questions are: > >> o Can I have an ILU mainloop inside of Zope? (I really don't > >> care if I have to hack Zope or if I can use an external > >> method.) > > Paul> This is a poor answer (as most here know I don't have the > Paul> mental horsepower to provide good responses :^), but...our > Paul> _intention_ over time is to have "ZServer" allow objects to > Paul> be published via PCGI, FastCGI, HTTP, FTP, and CORBA > Paul> simultaneously. > > Paul> Ideally we'd get some kind of Python implementation into the > Paul> Medusa handler layer. Obviously this approach has some > Paul> serious drawbacks... > > Paul> Anyway, I realize that I'm not providing much of an answer, > Paul> so sorry :^( > > Would it be worthwhile to start looking at and hacking ZServer? Yes. ZServer really restores Zope's original intent from the Bobo days: publishing objects through a variety of protocols. I think ILU support was once attempted back in those days. However, a Bobo app at that time could only have one publishing mechanism (ie, CGI, PCGI, FastCGI, ILU, etc...). ZServer allows for multiple publishing mechanisms in the same application, allowing access to the objects through the web, through FTP, or through any other protocol developed for ZServer/Medusa. .jPS | jeffrey(a)digicool.com zope, through the web. www.zope.org

1 0

RE: [Zope] Using Zope with CORBA (new user question)
by Michel Pelletier 05 Mar '99

05 Mar '99

> -----Original Message----- > From: Jody Winston [mailto:jody@ldeo.columbia.edu] > Sent: Friday, March 05, 1999 10:19 AM > To: Paul(a)digicool.com > Cc: zope(a)zope.org > Subject: RE: [Zope] Using Zope with CORBA (new user question) > > > >>>>> "Paul" == Paul Everitt <Paul(a)digicool.com> writes: > > Paul> Jody wrote: > >> My basic questions are: > >> o Can I have an ILU mainloop inside of Zope? (I really don't > >> care if I have to hack Zope or if I can use an external > >> method.) > > Paul> This is a poor answer (as most here know I don't have the > Paul> mental horsepower to provide good responses :^), but...our > Paul> _intention_ over time is to have "ZServer" allow objects to > Paul> be published via PCGI, FastCGI, HTTP, FTP, and CORBA > Paul> simultaneously. > > Paul> Ideally we'd get some kind of Python implementation into the > Paul> Medusa handler layer. Obviously this approach has some > Paul> serious drawbacks... > > Paul> Anyway, I realize that I'm not providing much of an answer, > Paul> so sorry :^( > > Would it be worthwhile to start looking at and hacking ZServer? > It's allways worth while! :) I think the point paul was trying to push is that we are interesting in publishing our objects over various transport protocols. If you want to look in this direction, probably the best idea would be to write a Medusa handler. Check out Sam's site at: http://www.nightmare.com/medusa There currently exists an ftp and http medusa handler for Zope, which you can look at in the latest ZServer alpha release availabe at zope.org. I'm not sure is there is a Medusa CORBA handler or not, I suspect not, but maybe there's some type of CORBA out there written in Python? -Michel > Jody > > > _______________________________________________ > Zope maillist - Zope(a)zope.org > http://www.zope.org/mailman/listinfo/zope >

1 0

RE: [Zope] Using Zope with CORBA (new user question)
by Paul Everitt 05 Mar '99

05 Mar '99

Jody wrote: > My basic questions are: > o Can I have an ILU mainloop inside of Zope? (I really don't > care if I have to hack Zope or if I can use an external > method.) This is a poor answer (as most here know I don't have the mental horsepower to provide good responses :^), but...our _intention_ over time is to have "ZServer" allow objects to be published via PCGI, FastCGI, HTTP, FTP, and CORBA simultaneously. Ideally we'd get some kind of Python implementation into the Medusa handler layer. Obviously this approach has some serious drawbacks... Anyway, I realize that I'm not providing much of an answer, so sorry :^( --Paul

2 1