What kind of content is it that you are uploading to Zope, and what is your workflow? As someone working in the news industry with several online newspapers and also fairly experienced with Zope and content automation, I might be able to make some suggestions that might help you out. Zope has some weird issues with namespaces and syntax that sometimes clash with Python's. That said, I think that things are constantly improving. Perhaps _['identifier'] solves your problem, at the slight cost of being just a bit kludgy. Do you use Zope for your staging server, or just for your live site? There might be ways to use it at both ends, integrate with dreamweaver, and bypass using wget...

From the sounds of it, you are using zope as the back end to an "editorial-system" workflow using Dreamweaver, but you are not taking advantage of most of Zope's features that facilitate content management. If I was familiar with your workflow, I'd be happy to provide advice if I could be of any help.

Good luck, Sean -----Original Message----- From: Ausum [mailto:ausum@mail.cosapidata.com.pe] Sent: Tuesday, December 19, 2000 11:20 AM To: jpenny@universal-fasteners.com; zope@zope.org Subject: [Zope] Zope's "." vs "_" unsolved issue (Re:To retrieve properties from a dtml document) document) jpenny@universal-fasteners.com wrote:

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.

Hello Jim, I'm finding myself greatly dissapointed that Zope doesn't have a built-in routine for this "." vs "_" problem. Although it may not be a big issue if you remain in Zserver and Unix, it is if we want the Zope's users base to keep growing, because we all benefit from it. And this point is a sort of unpleasant barrier for not-programmers Windows users like me. Dreamweaver is great for layout and design, and for nothing else. In despite of its templates and libraries and everything else, it would be unclever to let the journalists and editors in charge of content creation at my job, to open Dreamweaver and insert their stories in each page, asking them to not forget to write the summary in the description tag and the story title as the page title tag. :) For the time being we can't afford a 3 Mb/s connection (our current bandwith needs) so we must upload to a remote server almost two hundred pages daily. And even when relying page updating on trainned personnel exclusively, there's been this funny experience that the printed edition releases before the digital one, meaning that we have to deal with this gap between releases all the time. So that's why we need Zope, and why we need to get over its behaviours. Maybe it's time to throw the ball to Macromedia's ground. :) Thanks again for your comments, Jim Ausum jpenny@universal-fasteners.com wrote:

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

_______________________________________________ Zope maillist - Zope@zope.org http://lists.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope-dev )