Dennis Allison wrote:

...

To debug this problem, you might want to use the nice tcpwatch (-> google). It acts as a proxy but displays the complete http conversation in a window.

Thanks for the pointer...

...

Another remark to your redirects, AFAIK the the rfc mandates to redirect to complete URLs, not just the relative link. I don't know offhand if zope "fixes" you redirect, but I would use something like absolute_url()+'/index_html' as a target instead.

That might be the problem although I have used relative URLS in redirects without problems in other contexts. I checked the Zope docs for this restriction and didn't find any information. Might you have a pointer?

Sorry, no special pointer, see the http rfcs, they always talk about URI, and your relative links is not an URI. Note that I doubt that your specific problem is caused by this, it was just a general note. The same thing holds when you use javascript to load another page via location.href='somewhere.html' (syntax only IIRC). This works for most browsers, but is against the spec, and netscape 4.x gives bizarre results on this: It doesn't find the domain 'somewhere.html' and returns the netscape search results page for that term. In some cases it immediately redirects to the first search result. We were absolutely stumped when this happend, and it's a hell to find out _what_ happend when you just see a strange website in a frame where you expected to see your navigation bar. Therefore I'm a little bit careful with things like that now. cheers, oliver