Stefan Hoffmeister wrote:

How should one validate the HTML produced by Zope?

An external validator on the site? But that pinpoints only the problem at the macroscopic level, not at the Zope document or product (i.e. microscopic) level?

you need to put HTML comments at the beginning and end of your DTML methods. These will be passed unchanged to the final page, and any problems pointed out by your validator will be pinpointed.

I would like to validate individual Zope documents. The problem I mainly have is that Tidy doesn't know the DTML markup - meaning that I cannot use the validation and cleaning (!) services offered by Tidy on Zope "snippets" (documents, methods, ...)

The problem is a bit deeper than that. Even if a validator could be convinced to ignore DTML, you would still have the problem of trying to analyze an incomplete HTML document.

My interest in Zope is in modularizing a site, making it more maintainable this way - much in the same way as a source base in any OOPL is modularized into classes, I would like to accomplish the same with *content* on a web site. I do not seem to have found the key to gaining both modularity *and* robustness with Zope.

I'm unsure what you mean, can you give an example? Michael Bernstein.