On Thu, 24 May 2001, Dunigan, Craig wrote:
<dtml-tree expr="REQUEST.resolve_url(BASE3)"
You shouldn't use REQUEST.resolve_url. Use restrictedTraverse instead.
branches_expr="objectValues('Yihaw Folder')" nowrap="1" sort="title"> <a href="&dtml-absolute_url;"> <img src="/Images/FolderIcon.gif" border="0"> <b><dtml-var title_or_id></b> </a> </dtml-tree>
A user hits /folder1, generates a tree, modifies the tree state and gets a cookie. Then he hits one of the links above, goes to /folder1/folder2, generates a new tree, but with the state saved in the cookie. If he modifies the tree state yet again, does he get a new cookie, obliterating the previous tree state, or does the same cookie get modified, saying "for page /folder1/index_html the tree state is 'x', for page /folder1/folder2/index_html the tree state is 'y'?"
The former: the old cookie gets obliterated. There is only one 'tree-s' cookie, and it can only hold the state for one tree, the last tree viewed by the user.
My users seem to expect that they can return to a page and have whatever tree state they "saved" on that page reappear, and reading what you just wrote, I don't think that's true, is it?
Nope.
Obviously, it isn't happening, but it _appears_ (key concept!) that it is happening at some times and not others, and I'm trying to find out if it _can't_ happen, or if I did something wrong.
When it apears to happen, it may just be a caching issue. I'm sure you could try and extend the tree cookie to be more all-encompasing as you described above. This would have the upshot of allowing two trees on the same page. -Michel