[Zope] To retrieve properties from a dtml document

[jpenny@universal-fasteners.com]

On Mon, Dec 18, 2000 at 09:10:54PM -0500, Ausum wrote:
> Hello Andy, thanks for the advice.
> It didn't work. Maybe I'm wrong at any point, so please tell me whether
> it worked for you.
> 
> I'm aware that this problem has been discussed here before, that it
> happens because of Python's object naming, and also that there's a
> workaround at least (and just)for expressions, using _['whatever.html']
> 
> Due to the kind of workflow we have, we need all of the documents to be
> editable with Dreamweaver at any moment, directly, after the daily WGET
> process. Batch-replacing every "_html" with ".html" in file names and
> content seems to be very complicated within this scenario.

Going from dreamweaver to Zope, if you are working on 
unix, batch replacing is EASY, a single find command to 
change the file names and a simple
sed script (or python or perl)  to change hrefs.  
You will need to write such a script anyway, if your servers
are Unix or Linux and your Dreamweavers are Windows.  You will
find that your windows people mix filename case randomly and
expect it to work; you will have to canonify all hrefs to a
single case convention.

If going from zope to dreamweaver, either write the corresponding
_ to . scripts and batch it, or write a checkout external script
that does it for you.  I have not thought about this latter option
very much, but the export facility should give you an example of
how to do this.  I doubt if it is much more than a day's reading
and two days coding (and this is being damn generous, as I suspect 
that this is a ten line script).

You still have given no cogent reason for using Zope.  If you have only,
or predominatly static content, you are paying a huge overhead penalty.
If you have mixed static content and dynamic content, use apache and
ProxyPass to front-end the dynamic content, and apache to serve the
static content.  Then you have no conversion worries at all on the
static content.

Jim