- Zope - Zope lists

RE: [Zope] Zope Performance Methodology?
by Michel Pelletier 08 Dec '99

08 Dec '99

> -----Original Message----- > From: Michael A. Alderete [mailto:alderete@be.com] > Sent: Tuesday, December 07, 1999 9:00 PM > To: zope(a)zope.org > Subject: [Zope] Zope Performance Methodology? > > > - Do multiple processors help? A little. Python's interpreter has a global lock that all threads share, so all python code per Zope process is serialized. However, a lot of the underlying C code will release the python lock, for example, the oracle api calls in DCOracle free the lock so that Zope (or some other thread) can continue to work while the oracle call completes, so in these special cases another processor helps (and, of course, everything on your system at the processor level will benefit, but Zope won't benefit as much). > - How much RAM? That depends directly on what kind of performance you want. The bigger you can let your object cache get, the faster Zope will serve content. This doesn't matter much if your site only has a few objects, but for huge sites with lots and lots and lots of objects (like the Zope site 162,096) a big cache is a big win. -Michel

1 0

Zope 2.1.0 review (linuxplanet)
by Gerard Mulot 08 Dec '99

08 Dec '99

hello, you can find a review of Zope 2.1.0 from kevin Reichard http://www.linuxplanet.com/linuxplanet/reviews/1319/1/ regards Gerard Mulot

1 0

DTML Methods and parameters
by Alwyn Schoeman 08 Dec '99

08 Dec '99

Hi, Can I call a user-defined DTML-Method with user-definded parameters? Can someone please illustrate or point me to a reference? Thanx -- ~~~~~~~~~~~~~~ Alwyn Schoeman Systems Engineer Prism Secure Solutions

2 1

Re: [Zope] String to object ID
by Goodrichs 08 Dec '99

08 Dec '99

Well I still don't know how to reference a Zope object from a string, but I have a interim solution for now. I created a method in my root directory called "buildHeader"; <dtml-let newname="( _['root']+'/'+ _['prefix']+'/'+ _['path']+ _['fold'])"> <IMG SRC="<dtml-var newname>"> </dtml-let> Then in _any_ document within the site I can add the following code; <dtml-var "buildHeader( root=REQUEST.BASE0, <-- urlbase of my website(works with site access!) fold=PARENTS[0].id, <-- id of parent directory path='images', <-- path to my images folder prefix='hd')"> <-- prefix identifier of image The correct image is displayed based on the directory location of the document. It's not much but it's a start on what we are looking for. I still..... don't understand why I cannot create a Zope reference on the fly. It seems strange to me that object references must be hardwired into Zope. That I can see, if you do not already have the object within your namespace, it is inaccessible. Possibly I do not see this correctly, must every object link within a Zope website be hardcoded? Thanks, DAve.

5 5

URl1 and running zope behind proxy server ...
by Ramarao Kanneganti 08 Dec '99

08 Dec '99

Question about running Zope behind proxy server: I have a machine (internal.company.com) running zope. The company website www.company.com has proxy server at zope.company.com that gets internal.company.com's zope site. The problem I am facing is just about every link is absolute, means that I can't go beyond the first page by clicking from the front page. I noticed the reference to URL1 which is the current PATH. Now, barring getting rid of URL1 in all the standard zope default docs, is there a way that I can modify the situation? The zope docs are not very helpful in this matter. I read the zope docs and could not find a way out. Zope is really slick but the zen of zope takes a bit of getting used to. Thanks, Rama

2 1

RE: [Zope] WebDav wishful thinking WEBCR ideas.question=now Zope++Rebol
by Jason Cunliffe 08 Dec '99

08 Dec '99

At 16:50 07/12/99 -0500, Brain Lloyd wrote: >I hadn't looked at Rebol before today. My impression (FWIW): while >I'll agree that your example above is exceptionally clear, that >clarity also degrades as the logic becomes more complex. This is >not to make a judgement on Rebol necessarily - I think that this >is an issue in _any_ language or environment. Before you know it, >people will have moved beyond: > > "send luke(a)rebol.com read <http://www.rebol.com/>http://www.rebol.com" > >to need: > > "send luke(a)rebol.com read <http://www.rebol.com/>http://www.rebol.com but only if he's > allowed to see it and only if its sunday and oh yeah convert it > to multi-part while you're at it" > >>From what I saw of Rebol, the end result would be significantly >uglier than my contrived example above :) Again, that's not to >bash Rebol in any way - I'm certain that the equivalent DTML is >just as or even more ugly. Complexity just naturally grows in step >with flexibility. Yes I agree with you. in fact I find as one works one's way down the examples list at: <http://www.rebol.com/examples.html>http://www.rebol.com/examples.html It starts to look more and more like any.py file To me there are two clear ideas in Rebol at a glance: - it uses spaces instead of other punctuation. This makes it a little more like 'natural human' prose langauges such as written english. - Rebol uses context lookup so that a command such as 'send' or 'read' apllies some common snese get-it-out-my face tricks to apply embdeed code to parse anything whihc has www. in it and go: "Ok - its http" or 'zope(a)zope.org', thats pop or smpt" etc.. The context 'dialect' ideas in Rebol perhaps do relate well to the acquistion features of Zope and what could perhaps be a 'cleaner' way forwards for future version of Zope syntax. As I understand _good_ use of Zope Acqusition allows powerful context-based structures to evolve. The designer of Rebol seems to have given a lot of thought to context-oriented sytnax and how to maintain it. There is an interesting interview at: <http://www.rediff.com/computer/1999/sep/29carl.htm>http://www.rediff.com/c omputer/1999/sep/29carl.htm It could be an interesting test to see how far one could push pseudo code, python or similar to write a short story/novel say. At what point would that become illegible to non-programmers or vice versa? >A lot of this really goes back to the root question of "what is >DTML really for?", which I won't go into at length (there have >been several threads about it lately). I would, however, suggest >that in many ways the use of External Methods or Python Methods >is the natural and Pythonic way to deal with reducing complexity >(at least apparent complexity from the point of view of the DTML >user). While there is no equivalent to your send-page example >implemented as a simple embedded DTML command, it is pretty >easy to hide that complexity behind External Methods or Python >Methods that implement those "commands": > ><dtml-call "send_page_to_someone('luke(a)rebol.com', '<http://rebol.com/>http://rebol.com')"> > >I'll be the first to admit that this is not quite as pretty, >but it is not unreasonably ugly. Precisely. Parse out some pun5tuation adn wired characters and one is back to a REBOl like regualkr prose style use of english. I am left wondering how to improve my Python skills so that my code reads more simply. How to construct a 'send' or 'read' c;ass metthod which can parse ahead a little or at least sniff out protocols under the hood and run defaults unless over-ridden. Rebol looks little like 21st century [forward polish] FORTH or good LINGO. >> How could Zope use Rebol or a rebol-like python external >> method to best >> effect? >> any ideas? > >I don't quite understand this - by this do you mean some kind >of object like an external method but that could be written >in Rebol? If so, that wouldn't really fix the DTML ugliness >problem would it? You'd still have to say something like: > ><dtml-call "send_page_to_someone('luke(a)rebol.com', '<http://rebol.com/>http://rebol.com')"> > >...even if the actual work was done in a different language >wouldn't you? Perhaps I'm missing your point though. hmm..well that's close and quote acceptable to zope99ers. But how about: <dtml-rebol send Brian(a)digicool.com "Hello Brian, check out <http://www.reol.com/>www.rebol.com and www.zope.org" or <dtml-rebol zope(a)zope.org read <http://www.rebol.com/>www.zope.org/members/brian/messages/today> ?? ie leverage on the Rebol clarity and scalability and excellent potential for mobile/PDA apps market, but using Zope as richer sophisticated environment for hosting and integrating = ie what it does well Or alternatively some friedly construct like <dtml-in rebol> or <dtml-with rebol> where 'rebol' dialectfolder can fhold rebol library I hope this does not seem like I am trying to reinvent Zope. I'm not. But it is clear that Zope can get unneccesarily messy (especially in some hands:-) and despite its excellent concept and power. thoughts? - Jason ------------------------------------------------- Jason Cunliffe <jasonic(a)nomadicsltd.com> NOMADICS.STUDIO(Design Director) Geo-Digital Arts and Technology Le Vieux Moulin, Route de Mons 83440 SEILLANS, FRANCE Tel: +33 (0)4 94.76.98.72 Fax: +33 (0)4 94.76.97.77

1 0

RE: [Zope] no data matching query
by Chris McDonough 08 Dec '99

08 Dec '99

> is it possible to test if a database-query returned data or not? > i'm not interested in the result - data, only if there where > one or more > rows matching the query... It's easier than you think. If you have a ZSQL Method "sqlMyQuery": <dtml-if sqlMyQuery> ...do stuff... </dtml-if> That's it. > > another simple question: > is there a more simple possibility to get only the first > query-match than: > > <dtml-in dataquery size=1 orphan=1> > > ... > </dtml-in> > errr... this isn't simple enough? you could always do: <dtml-in sqlMyQuery> <dtml-if sequence-start> ...do stuff... </dtml-if> </dtml-in> But that's more complicated, I think.

1 0

no data matching query
by Martin Hofer 08 Dec '99

08 Dec '99

hi out there, is it possible to test if a database-query returned data or not? i'm not interested in the result - data, only if there where one or more rows matching the query... another simple question: is there a more simple possibility to get only the first query-match than: <dtml-in dataquery size=1 orphan=1> ... </dtml-in> thanx in advance! martin hofer ------------------------------------------ o EDV-Referent an der HTU-GRAZ o ------------------------------------------ Rechbauerstrasse 12 4010 Graz - Austria mail: hofer(a)oeh.tu-graz.ac.at tel: +43/0316 873-5109 | _|_ fax: +43/0316 873-5115 |__ | mobil: +43/0699-1016 1847 | | | | | internet: oeh.tu-graz.ac.at | | | |__|

1 0

RE: [Zope] Zope Performance Methodology?
by Chris McDonough 08 Dec '99

08 Dec '99

> I haven't run any standardized benchmarks against them yet, but I plan > on doing so for the Dual CPU machines, and the two HP babies. On the > latter, I am checking to see if my boss will let me publish my results > on them externally (wouldn't know why they wouldn't ... :). > > Anybody have a preferred webserver benchmark that is appropriate for > Zope? > > If desired, I could turn this into a howto ... > > Bill Wow, this is all great information! Can I supply my vote for a HOWTO? As far as a benchmark tool... errr... its such a touchy subject. But let me stick my neck out and put in a vote for "ab" (the Apache benchmark tool). I for one would love to see how those different platforms compare to one another, even if the benchmarks are unreplicable. Thanks again! - Chris

1 0

external method as an href
by Robin Becker 08 Dec '99

08 Dec '99

Hi, I want to pass the current REQUEST into an external method which is used as an href. eg <a href=X?T=3>link_X</a> where X is defined as def X(REQUEST,RESPONSE,T=0): ..... How can I make Zope pass the current REQUEST into this external method? -- Robin Becker

1 0