4. make zope content easy to co-operative with current technonlgy especially search engines. zope oodb is greate, but it prevent other apps to access the data, so it is difficult to contruct a local search facility.
This is kind of vague, nothing in ZODB is specific to the way Zope
I have a kindof related point about search engines... Zope doesn't like objects which have extensions. For example: index.html, so instead index_html is used. However, search engines don't like pages which DON'T have extensions. They use extensions to work out the content type of the file which saves them having to do a HEADER get to find it. A few of the major search engines simply ignore files without an extension assuming them to be scripts or downloads. It would be nice if Zope had better support for objects ending in .html, for example, instead of having to use the convoluted "_['index.html']" and others... perhaps even change the default from index_html to index.html and maybe even add a default extension of .html to DTML documents when they're served up...
Hm, it seems to me this could be turned into an advantage using Zope. Using objects with externsions is no problem, unless you have to reference them from dtml (or python for that matter). However, if you design your site such that only the objects that should be visible to the outside world have extensions (and make these DTML Documents) and all objects that are invisible to the outside world have not, this actually makes it easier to make search engines only index what _you_ want to be indexed.
That is to say: if your dtml/sql/python methods, including the headers and footers etc have no extensions, they won't be visible to the people who search your site by way of search engines. And if your site is well designed (and ZOpe design encourages this), the final Documents rarely have to be called from other objects.
Rik