[Seaside] Content Management System in Seaside
renggli at gmail.com
Fri Feb 10 08:07:31 UTC 2006
> I saw the suggestion to use Pier as a CMS. I downloaded Pier from
> SqueakMap. I'm sure I didn't quite get it right, since some Report
> sometimes results in an error (if I'm in a leaf node, I think), and
> since clicking on Wiki prompts me to download the file (at least in
> Firefox). But what I saw looks like a good basis for a CMS.
- Reports: This was fixed some time ago. You should update to the
latest version using Monticello from
- Wiki-download: This is a feature, not a bug. It might be confusing,
but it should indeed offer to download the raw wiki-text of the page.
> thing I'd be concerned about, though, are the URLs. I realize that
> Seaside by default has session info in the query string, but for a CMS,
> this could be problematic. What do you think about issues like
> bookmarking and search engines?
Bookmarking of the current page, and of all links within a page works
perfectly. If the session string is missing, simply a new session is
created. By going to the Seaside configuration application, you could
of the URL.
> work without the query string, and that they're consistent, but I'm
> wondering in particular how a search engine would read the page (I would
> guess this differs from search engine to search engine).
I don't know about the behaviour of search engines together with Pier,
but there are certainly solutions. I think the plugin called Sitemap
of Philippe Marshall does submit all a sitemap to google using a
special XML protocol. I've never tried it but it might be worth
> Also, is there any sort of role-based security planned for Pier?
There are two security frameworks available that can be loaded as plugin:
- Philippe Marshal build an ACL based security, that is around for
quite a while now and that provides a lot of nice features. It
provides sophisticated tools to manage users, roles, command sets and
access control lists. I don't know if it works with the latest version
of Pier, but you might get some information by browsing to
- I build a tiny security model being very similar to the one of Unix.
Every page gets a collection of executable commands for the owner, the
group and the others. There is a set of commands to (recursively)
change owner, group and modes (similar to chown, chgrp, and chmod).
Users and groups cannot be managed trough the web right now, but that
will be probably improved. There is no inheritance. It is very simple
from implementation point of view, only a couple of classes and it
also works perfectly with OmniBrowser. Load Pier-Security from
http://mc.lukas-renggli.ch/pier or from SqueakMap.
More information about the Seaside