[Zope] Netscape/Zope interaction bug?

[Jens Vagelpohl]

Hi everyone,

as i mentioned here before, i am working on a class project implementing a
web-based customer database. yesterday i exchanged a few emails with the
professor and he looked at the pages from home. i followed his progress
through the site from my internet gateway's proxy logs and noticed something
funny. at one point zope just died. of course this is slightly embarrassing.
i had worked on the site all day and it was as good as gold.

after poking at it a little further and even downloading netscape (in this
case the latest milestone of mozilla, M8) i can replicate the following
behavior: when i try to click on a page link exposed as a leaf in a
<!--#tree--> tag i am getting the following error dump in the telnet session
i use to watch what is happening:

uncaptured python exception, closing channel
<zhttp_channel connected 192.168.168.3:4137 at 844fe00 channel#: 1
requests:0> (exceptions.TypeError:unpack non-sequence
[/usr/src/Zope-2.0.0a3-src/ZServer/medusa/asyncore.py|poll|59]
[/usr/src/Zope-2.0.0a3-src/ZServer/medusa/asyncore.py|handle_read_event|285]
[/usr/src/Zope-2.0.0a3-src/ZServer/medusa/asynchat.py|handle_read|126]
[/usr/src/Zope-2.0.0a3-src/ZServer/medusa/http_server.py|found_terminator|44
4])

this error, strangely enough, does not happen using MS IE4 as well as IE5,
which are the browsers i have at work and at home, respectively, and which i
use every single day to work on the site.

my setup is Zope 2.0.0a3 and python 1.5.2 on RedHat6. I'm not sure what
browsers the DC guys are working with, maybe someone has seen this behavior.

By the way, the code for this "tree" implementation is mostly cut and pasted
from the ZBook tutorial.

any help would be appreciated :)

Jens Vagelpohl