- Zope - Zope lists

Security Logon
by Kulpinski, Dejan 11 Feb '00

11 Feb '00

Hello all, Does somebody know how to create a logon screen without using the default Zope's Security Logon Screen? Essentially, I want html screen for logon from witch I am going to pass user name and password to Zope as parameters... Thanx, Dejan Kulpinski Sierra Systems Consultants Inc.

1 0

Getting paramters in DTML XMLRPC server?
by Robb Shecter 11 Feb '00

11 Feb '00

Hi, How does a DTML method get the XML-RPC paramters? These parameters aren't named, so I'm not sure where to get them from. The howto doesn't show this. Thanks, Robb

2 1

[Fwd: Security problem]
by Chris McDonough 11 Feb '00

11 Feb '00

Keith, I'm forwarding your question to the Zope mailing list, as it's more likely to be answered there... you can join by visiting the Resources - Mailing Lists link on www.zope.org. Thanks! KSherman2(a)excite.com wrote: > > Hi, I'm new to zope, and I'm trying to figure out a problem. I'm using zope > as a pcgi process running on a remotely hosted sever. It seems to be set up > correctly, and I've had no problem accessing the website I'm working on (a > squishdot site) with my username and password. Unfortunately I cannot get > it to be viewed anonymously. I have changed the security permissions > everywhere it seems to need to be changed, but the only html that its > sending to anonymous users is the main zope index_html. I've tried moving > the Squishdot html to that spot and renaming it to index_html and I can't > get it without a password request, (all security tab things were set to > anon) I've also tried setting up other DTML documents but they won't serve > either without an authorization request (neither in the same folder as > 'manage' or an offshoot directory) I'm running the solaris binary that > you've got on your site, version 2.1.3. As you can imagine that anonymous > viewings would be a mission critical situation. Nothing other than the > manage/index_html can be served, and if I try to tick it by putting another > DTML document in there, even that won't be served without first requesting a > userid/password. I know that my web hosting service is running Zeus, but I > followed the instructions on their FAQ. At one point was able to log in > with IE5 without a password, but it still needed a userid. I'm hoping > there's some sort of easy fix for this. I was wondering if maybe there's > some sort of DTML tag that can force the serving of the page without going > thru the authorization? Thanks for any help. I should also mention that I > think you have one of the most exciting products out there right now for web > development. > > Sincerely, > Keith Sherman > > _______________________________________________________ > Get 100% FREE Internet Access powered by Excite > Visit http://freeworld.excite.com -- Chris McDonough - Digital Creations, Inc. Publishers of Zope - http://www.zope.org

1 0

Logging users in Zope 2.x
by Hannu Krosing 11 Feb '00

11 Feb '00

How should I configure logging to get usernames into Z2.log I first hoped to get it from Apache's access_log, but it saves username only when doing user authentication itself. Then I thought that Zope at least would log it and migrated the site to Zope 2.1.3 to get logging support. Unfortunately it does still not log usernames and I have not found the place where I could turn it on. Also the format is buggy: Apache access_log: 193.40.215.135 - - [11/Feb/2000:08:29:55 +0200] "GET / HTTP/1.1" 200 2673 Zope Z2.log 127.0.0.1- - [11/Feb/2000:19:04:46 +0300] "GET /misc_/OFSP/dtmlmethod.gif" 200 1000 the bug being the missing space after the IP address. what I was hoping for would have looked like: 127.0.0.1 - superuser [11/Feb/2000:19:04:46 +0300] "GET /manage" 200 1000 ------- Hannu

1 0

Re: [Zope] Products breaking other Products?
by Tres Seaver 11 Feb '00

11 Feb '00

"James W. Howe" <jwh(a)allencreek.com> wrote: > > I just encountered a problem which scares me to some extent. I have > developed my own product with my own set of ZClasses. In my Products > folder I had my product, some other products and the PTKBase product. The > version of the PTK was from a couple of weeks ago (or so). Yesterday I > retrieved the latest PTK stuff from the CVS repository and updated the > contents of my products directory with this version of PTK. Suddenly > existing instances of ZClasses that I had created previously, were > appearing as broken. They apparently couldn't find their product. I was > confused by this, because I was able to add new instances of my > ZClasses. After posting a message to the zope mailing list, I received a > reply which suggested that perhaps a product unrelated to my product was > causing a problem. Having nothing else to lose, I replaced my PTKBase with > the old version of PTKBase. After restarting Zope my broken ZClasses were > once again functioning. > > My question is this, why should a product unrelated to my product cause > problems with my ZClasses? I feel a bit uncomfortable with the notion that > one product can seemingly affect another. I don't believe any of my code > relied on anything relating to the PTK. Can someone explain what might > have happened? When Zope starts up, it runs through each Product, importing it as a Python module. My guess is that the broken PTKBase was earlier in the list than your custom product, and that an exception propagated uncaught out of it, interrupting the normal product loading. If this is so, then we need to look at hardening Zope's product loading mechanism against such behavior. Just a guess, Tres. -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

2 2

Dreamweaver and _notes
by Chris Withers 11 Feb '00

11 Feb '00

Hi, I think this problem is best solved by modifying Dreamweaver, but since their news server is down I thought I'd ask here. I use Dreamweaver to edit the content of my Zope site. It works very well, including Dreamweaver's check in/check out stuff. It also has a system for keeping Design Notes associated with your files. Unfortunately, it chooses to store them in a folder called _notes on the server. Zope doesn't like folders that start with and _ so it doesn't work. In Dreamweaver, there is a file called FileClass.js which specified a constant that is '_notes.' However, changing this doesn't seem to affect Dreamweaver as it still tries to create the _notes folder. So, is there any way to ACTUALLY change this in Dreamweaver (like compiling the javascript of something) or is there any way to sort it out at the Zope end? cheers for any help, Chris

2 3

Alpha Linux
by kmail＠dolphinsearch.com 11 Feb '00

11 Feb '00

I am having trouble installing Zope 2.1.4, on and Linux server running on a Alpha. I have compiled python 1.5.2 with thread support and when I run python wo_pcgi.py, this is what I get. Any help would get great. Kelly(a)dolphinsearch.com [root@admin Zope-2.1.4-src]# python wo_pcgi.py --------------------------------------------------------------------- --------- Compiling python modules --------------------------------------------------------------------- --------- Building extension modules ------------------------------------------------ Compiling extensions in lib/python cp /usr/lib/python1.5/config/Makefile.pre.in . cp: /usr/lib/python1.5/config/Makefile.pre.in: No such file or directory Traceback (innermost last): File "wo_pcgi.py", line 116, in ? File "wo_pcgi.py", line 104, in main File "/usr/src/redhat/Zope-2.1.4-src/inst/build_extensions.py", line 96, in ? make('lib','python') File "/usr/src/redhat/Zope-2.1.4-src/inst/do.py", line 133, in make do("cp %s ." % wheres_Makefile_pre_in()) File "/usr/src/redhat/Zope-2.1.4-src/inst/do.py", line 104, in do if i and picky: raise SystemError, i SystemError: 256 [root@admin Zope-2.1.4-src]#

1 0

RE: [Zope] ZServer breakage wrt HTTP - 2.1.3
by Martijn Pieters 11 Feb '00

11 Feb '00

From: Brett Morgan [mailto:brett@pulse.itd.uts.edu.au] > > I am playing with ZopeStudio on a stoke Zope 2.1.3 build (wo_pcgi) > running ZServer on port 8080. It don't work. I uncomented the > debugging dump at the top of the zopestudio.js file in the chrome and > noticed that the server was just sending out 25 pings for the two rdf > files it wants. I develop ZopeStudio on straight ZServer, so I don't think that that is the problem. The 25 pings you see is ZopeStudio waiting for Mozilla to have retrieved and parsed the RDF, not actual data over the network. Look in your Z2.log file, in the var directory. You should see entries for "GET /zsTreeRoot HTTP/1.0" and "GET /zsContainerRDF HTTP/1.0". If they are not there, Mozilla cannot find your server. Thanks to some fixes from the Mozilla team, better error reporting is under way. > I retreived them without a hitch from within netscape > 4, but mozilla can't even get the main page from zope. Again, my copy of Mozilla has no problems. What build are you using? > > So here's a theory: > ZServer can't grok http 1.0 compliant request thus: > > [brett@localhost brett]$ telnet localhost 8080 > Trying 127.0.0.1... > Connected to localhost. > Escape character is '^]'. > GET http://localhost:8080/ HTTP/1.0 > host: localhost:8080 > > HTTP/1.0 404 Not Found > Server: Medusa/1.13 > Content-Type: text/html > Content-Length: 134 > Date: Fri, 11 Feb 2000 15:04:23 GMT > This should be GET / HTTP/1.0 The host part and the protocol part of the URL are never sent to the server. After all, they define on what port and on what server you place the call. -- Martijn Pieters, Software Engineer | Digital Creations http://www.digicool.com | Creators of Zope http://www.zope.org | mailto:mj@digicool.com ICQ: 4532236 | PGP: http://wwwkeys.nl.pgp.net:11371/pks/lookup?op=get&search=0xA8A32149 -------------------------------------------

2 1

ZServer breakage wrt HTTP - 2.1.3
by Brett Morgan 11 Feb '00

11 Feb '00

I am playing with ZopeStudio on a stoke Zope 2.1.3 build (wo_pcgi) running ZServer on port 8080. It don't work. I uncomented the debugging dump at the top of the zopestudio.js file in the chrome and noticed that the server was just sending out 25 pings for the two rdf files it wants. I retreived them without a hitch from within netscape 4, but mozilla can't even get the main page from zope. So here's a theory: ZServer can't grok http 1.0 compliant request thus: [brett@localhost brett]$ telnet localhost 8080 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. GET http://localhost:8080/ HTTP/1.0 host: localhost:8080 HTTP/1.0 404 Not Found Server: Medusa/1.13 Content-Type: text/html Content-Length: 134 Date: Fri, 11 Feb 2000 15:04:23 GMT <head> <title>Error response</title> </head> <body> <h1>Error response</h1> <p>Error code 404. <p>Message: Not Found. </body> Connection closed by foreign host. But lives fine with this slacker http: [brett@localhost brett]$ telnet localhost 8080 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. GET / HTTP/None 200 OK Server: Zope/Zope 2.1.3 (source release, python 1.5.2, linux2) ZServer/1.1b1 Date: Fri, 11 Feb 2000 15:05:25 GMT Content-Length: 1290 Content-Type: text/html <HTML> <HEAD> <TITLE>Zope</TITLE> <base href="http://localhost:8080/"> </HEAD> <BODY BGCOLOR="#FFFFFF"> ... which is not actually correct since I pinged it in http 0.9. No matter. Everyone of note communicates in HTTP/1.0ish. But Mozilla the standards compliant client that it is, breaks ZServer. I hope this helps the guys trying to figure out why ZopeStudio is breaking :) brett -- email: Brett.Morgan(a)uts.edu.au phone: +61-2-9514-2091 The rhino is a homely beast, For human eyes he's not a feast. Farewell, farewell, you old rhinoceros, I'll stare at something less prepoceros. -- Odgen Nash (1902-1971)

2 2

XML-RPC woes
by Tony McDonald 11 Feb '00

11 Feb '00

Hi all, I'm trying to get this to work **from a XXX PythonMethod** (ie the ones that allow you to import things and are coloured red and orange!) PythonMethod do_par import xmlrpclib server_url="http://xmlrpc.server:9999" server=xmlrpclib.Server(server_url) server.getTranscript(number) I call it with a dtml doc <form action="do_par" method=post> <input type=text name="number:int"> </form> The PythonMethod returns a blank page (whether from the form or the 'try' tab) If I make the PythonMethod be; import xmlrpclib def doit(number): server_url="http://xmlrpc.server:9999" server=xmlrpclib.Server(server_url) return server.getTranscript(number) doit(number) Then I get this error, when called from either the form or the 'try' tab.   I'd like to work from PythonMethods as they offer the promise of speedy editing, so what am I doing wrong this time? :( getTranscript does work, because all this code works from the Python interpreter. ta tone. ------ Dr Tony McDonald, FMCC, Networked Learning Environments Project http://nle.ncl.ac.uk/ The Medical School, Newcastle University Tel: +44 191 222 5888 Fingerprint: 3450 876D FA41 B926 D3DD F8C3 F2D0 C3B9 8B38 18A2

3 4