March 2000 - Zope - Zope lists

Referencing a Zope Instance through a Hyperlink
by M.J. Stahl 08 Mar '00

08 Mar '00

I have a group of ZClass (DCTasks) stored in a folder '/tasks/generator/taskstorage' The index_html of '/tasks' is the interface to view the instances and certains properties that are saved on those instances. One of the properties is 'task_title' and is displayed as a hyperlink to view the rest of the properties of the instance. This is my problem, when I check the source of the page (/tasks/index_html) and I take a look at the hyperlinks for viewing the properties of the instance, it gives a URL like so: http://diis/tasks/generator/taskstorage/<string> <string> should actually be replaced with the 'id' of the instance. This is so by the DTML code for index_html: <a href="<dtml-var URL1>/generator/taskstorage/<dtml-var id>"> <dtml-var task_title> </a> I have no idea why a string is taking the place of <dtml-var id> in referance to the instance. Any ideas would be helpful. Best regards, MJ Stahl

2 1

POLL: Preferred browser?
by Curtis Maloney 08 Mar '00

08 Mar '00

Greetings, Look, I don't want to start any holy wars, just to get a head count, more or less. But what are peoples preferred browsers for ZOPE development? Until now, I've been using Netscape, whilst desperately searching for some replacement that didn't haemorhage memory. I think I've even managed to narrow down which part of NS is losing the memory (it's the edit box, which is why it seems to happen quicker when working with ZOPE). However, in my searches I've yet to find ANYTHING under Linux that comes close to the functionality of NS. As I said... I don't want advocacy here, simply name and version, and perhaps a quick rundown of the reason(s) you chose it over others. -- Have a better one, Curtis. <dtml-var standard_work_disclaimer>

8 8

RE: [Zope] Zope needs this (and Dynamo has it)
by Alexander Staubo 08 Mar '00

08 Mar '00

> From: John Goerzen [mailto:jgoerzen@complete.org] > Sent: Tuesday, March 07, 2000 9:47 PM > To: Alexander Staubo > Cc: Zope Mailing List (E-mail) > Subject: Re: [Zope] Zope needs this (and Dynamo has it) > > > Alexander Staubo <alex(a)mop.no> writes: > > [snipped] > > > Zope doesn't have the flashy packaging and sexiness of a > product line > > Dynamo -- which, actually, I strongly urge you (and > everybody else in > > this forum) to experience yourself, before you snap back at > my again -- > > and this affects management decisions. Dynamo is > exceedingly sexy, and > > it boasts the kind of bullshit magic that have management and tech > > people alike drooling. > > OK, but see you're trying to make Linux something it is not. Linux is > not about marketing. It is not about world domination. We > specifically do not care about this :-) Now, digicool is a company, > and they probably DO care about this. Making Zope Free Software is a > great way for them -- and us -- to benefit. World domination has nothing to do with it. If Zope were the only tool out there (or if it were enjoying the same kind of monopoly iron grip on the market), that'd not be a Good Thing. At the moment, Zope is still very much a "guerilla" product. It follows that you might have to use guerilla tactics to lure it into your company. Like Linux was just 1-2 years ago. >So your comments should > be addressed at them if you want them to make the sale. Not me, a > Linux hacker. Well, you entered this thread on your own -- I wasn't addressing my post to you. If the elements of the discussion do not concern you, why are you arguing against them? > > As one of my colleagues pointed out, the problem starts > with the name. > > (Somebody out there is hyping what they call Linux' > next-generating GUI. > > And it's called... GNOME. What's wrong with this picture?) > I don't care > > myself, but I recognize the problem. > > But see again, Free Software is not about marketing. The name is > irrelevant to us. Free Software is about what's *right* and what > *works*. We leave the marketing to those that want to do it (such as > digital creations, or whomever else). So again, write to digital > creations' marketing dept. :-) Arguably, free software (or Free Software) is about freedom, not necessarily about what's right or what works. RMS, arguably the grand old man, top banana, and icon of the Free Software movement -- which is not the same as Open Source -- openly admits he'd rather have bad, free software than good, non-free software. The Open Source movement takes the pragmatic view that bad, free software isn't an improvement at all, and that the software industry has little use for RMS' moralisms. The Open Source guys just wants "software that doesn't suck". Perhaps Free Software people don't care about marketing. But then making good impressions or having good hygiene has never been top priorities for geeks. ;-) For some people, outward perceptions do count. I flinch every time somebody sets up Microsoft Exchange Server, or buys some pricey piece of non-compliant Internet software, or seriously argue the significance of ASP. The people who make such decisions are often, sadly, uneducated about what's good and about "software that doesn't suck". *Other* people suffer because of such misfires. Here's a different perspective you might not have taken into account in your salutations to the spirit of Free Software. If Linux *hadn't* been picked up by huge corporations and stock markets, if Linux *hadn't* been rampantly commercialized, would we have seen the surge of wonderful developments and innovations for the OS that we're seeing now? Methinks not. As much as I abhor commercialism and the new Linux hype, Linux' implicit "world domination" plan is doing good things for the OS. Linux is not the ultimate OS and hopefully won't be the last stop on the way to utopia, but I rather look forward to the point where it's usable as a desktop OS and I can forever throw out the software I know for a fact sucks. Marketing, in the case of Open Source/Free Software, is mostly a tool to educate ignorant brass. *We*, as you point out, don't need the stuff. > -- John > -- Alexander Staubo http://alex.mop.no/ "Do not go gentle into that good night/old age should rave and burn against the close of day/Rage, rage against the dying of the light." --Dylan Thomas

1 0

Re: [Zope] can I eliminate the data.fs file in favor of an relational database?
by John Goerzen 08 Mar '00

08 Mar '00

Ron_Dagostino(a)SSGA.COM writes: > I'm thinking it would be better to eliminate this file in favor of an > Oracle or Sybase database (I have access to both). Can I do this? If it Well, basically what you are doing here is moving the point of potential corruption from zope to the relational database. This may or may not give you any comfort. For me, it makes me more nervous since I don't have the source code to those particular databases, and thus couldn't fix things if the worst does happen.

1 0

Zope authentication problem
by Ken Weiss 08 Mar '00

08 Mar '00

I apologize in advance for this note. I know this topic has come up several times in the archives, and I swear I've read every post, every how-to, and every other bit of documentation I can find. I can't get past the management screen authentication for Zope with Apache. Here's my environment: OS: RedHat Linux 6.1 Zope: version 2.1.4, installed from Jeff Rush's RPMs (thanks, Jeff) Zope PCGI: version 2.1.4, also from Jeff Rush's RPMs (thanks again, Jeff) Apache: version 1.3.9, installed from the Red Hat RPMs (both base and devel RPMs are installed) Python: version 1.5.2, from the Red Hat RPMs I'm configured as follows: Zope install directory: /var/zope Zope.cgi: copied to /home/httpd/cgi-bin/Zope (not a symlink, and renamed 'Zope') access file: /var/zope/access, contains: admin:cleartextpassword httpd.conf contains: RewriteEngine on RewriteCond %{HTTP:Authorization} ^(.*) RewriteRule ^/static/(.*) /home/httpd/html/$1 [l] RewriteRule ^/Zope(.*) /home/httpd/cgi-bin/Zope$1 [e=HTTP_CGI_AUTHORIZATION:%1,t=application/x-httpd-cgi,l] (that last line is all one long line, and I'm pretty sure I've sorted out the ls and the 1s properly) Zope is launched with the following command: cd $INSTANCE_HOME (/var/zope) /usr/bin/env python z2.py \ -z /usr/share/zope \ -u zope \ -Z /var/run/zwatchdog.pid \ -p /home/httpd/cgi-bin/Zope \ -w '' \ -f 8021 \ -m '' \ -l /var/log/zope \ 2>&1 >> /var/log/zope \ & I can access all my other (non-Zope) CGI programs normally. I can access my static html with URLs of the form: http://server.ucop.edu/static/some_silly_url.html So, I'm pretty sure that both CGI and the rewrite modules are installed, configured, and working properly. I can run Zope and see the QuickStart materials, so I'm pretty sure that the basic Zope installation is correct. But I can't get authenticated to Zope. There's nothing in the logs for httpd or Zope that looks interesting to me: httpd - access_log: 207.212.67.31 - - [08/Mar/2000:10:14:18 -0800] "GET /cgi-bin/Zope/manage HTTP/1.0" 401 1388 Zope log (syslog messages file): Mar 8 10:14:18 akpi ZServer[24689]: 127.0.0.1- - [08/Mar/2000:18:14:18 -0700] "GET /manage" 401 1782 /var/log/zope is empty. I've tried about a zillion other things, including installing the Zope binaries instead of the RPMs, and installing and building the Zope source. In every case I end up with the same result. Zope doesn't accept the authentication. One oddity I noticed is that if I access Zope through: http://server.ucop.edu/Zope it immediately requests authentication. It doesn't even show me the opening screen or allow me to browse the QuickStart. If I dodge around the RewriteRule by going directly to: http://server.ucop.edu/cgi-bin/Zope then it doesn't ask for authentication until I try to hit the management screen. I should probably add that I tried using the built-in Zserver Python package and it worked fine. I haven't tried Zap. I would prefer to use Apache because I want to serve a mixture of legacy HTML and Zope content. I would prefer to use my own Apache instead of Zap because in general I'm not comfortable with using a package that I can't maintain at the current release level. Zap Apache trails plain Apache releases, at least by a little bit. I've tried a whole raft of variations on the rewrite rules in Apache. I'm stuck. Anybody out there been through this already? --Ken ------------------------------------------------------------------------- Ken Weiss ken.weiss(a)ucop.edu California Digital Library Technologies (510) 710-3356 (voice) UC Office of the President (510) 763-2471 (fax) 1111 Franklin Street #7313B ken.weiss.pager(a)ucop.edu (text page) Oakland, CA 94607-5200 http://dcas.ucdavis.edu/kenhome.html PGP public key stored at: ldap://certserver.pgp.com http://pgpkeys.mit.edu:11371/pks/lookup?op=get&search=ken+weiss

2 2

can I eliminate the data.fs file in favor of an relational database?
by Ron_Dagostino＠SSGA.COM 08 Mar '00

08 Mar '00

I just installed Zope (on NT) and would like to have reasonable backup/recovery options. One way to do this is to just make a copy of the data.fs file in the var directory, but I'm concerned that this file may not be in a consistent state when it gets copied somewhere. Should I worry about this? I'm thinking it would be better to eliminate this file in favor of an Oracle or Sybase database (I have access to both). Can I do this? If it is possible, would I want to do it? If I should want to do it, how do I do it? I've spent a few hours searching the documentation, but this has only convinced me that it may be possible -- I've been unable to find a definitive answer. Ron_Dagostino(a)SSGA.Com

3 2

Object Manager and Folder
by Mario Lang 08 Mar '00

08 Mar '00

Hi. I am quite new to Zope and currently try to create a BookClass ZClass Product. I came across several issues where you probably can give me some useful tips: 1. What is the difference between inheriting rom Object Manager and/or Folder. 2. My product will include Cover Images of the books. (a small and a large one). Whats the best way to implement this? I thought about creating emtpy images (smallcover and largecover), but I dont know how to ask the Objects if they contain something. Or should I just add the Images in my BookClass Instance programmatically? 3. Wasn't their any web-client product for Zope so that I could retrieve content from the web directory and import it in Objects (like a Image) without having to upload it manually? And the last one: Is their any special reason why FTP access doesnt work in the /Control_Panel/Products folder? All methods are empty their. I work with Emacs and ange-ftp is quite a good method to access content of Zope Objects. Besides, if I change the content of a Object through the web interface, Zope seems to convert < > and " chars into < > " I understand why, but is this really necessary. It seems as if nearly all textmode browsers (as lynx and emacs/w3) dont care about converting the entities back tto their corresponding characters. Thanks in advance, Mario <mlang(a)delysid.org>

2 1

Re: [Zope] Creation of an ZClass Instance.
by M.J. Stahl 08 Mar '00

08 Mar '00

><input type=hidden name=id value="<dtml-var "REQUEST['id']">"> > >Variables within REQUEST only exist for that particular web request. To pass >values on to your _add method you need to put them in form fields or >cookies. > >Kevin Ok that seems to work because I am getting other errors and I think this task is a little advance for a zope infant like myself. Now, this line is causing me problems (this happened after I called entered the hidden field): <dtml-call "REQUEST.set('id',_.str(_.int(id)))"> The error it gave me was: Error Type: ValueError Error Value: invalid literal for int(): generator It seems to be refering to the "(.int(id)))" of the above code. -M.

2 1

calling background/child processes
by Philipp Dunkel 08 Mar '00

08 Mar '00

Hi everyone. When calling my ZClass_add method I want to call an external python method. Is there any way to let the method return immediately so that the script is running in the background? thanks a lot philipp dunkel ** ** ** ** ** ** ** ** pgp public key available gpg public key available ICQ# 60149094

3 2

Re: [Zope] FYI: large files & linux.
by John Goerzen 08 Mar '00

08 Mar '00

tsarna(a)endicor.com (Ty Sarna) writes: > Even simpler than buying new hardware, switch to a system like NetBSD or > FreeBSD that has supported large files for years and years :-) It supports it to no greater or less extent than Linux does. The standard library defaults to using an int or a long for *seek calls. The difference is semantic only, as they're both 32 bits on both systems anyway. Even if you assume unsigned (fseek is signed), that's still 4gb, max. However, on my Linux/Alpha system, a long is 64 bits. This effectively removes the problem. I submit, therefore, that you are not seeing a Linux problem, but an x86 problem :-)

4 6