Previous Chapter | Next Chapter | Up | Next Section | Contents

How Does Zope Use It?


Netscape publishing is designed to get files from the content manager's hard drive to the hard drive at their website, using the site's Web server as the facilitator. Since Zope doesn't store documents as files, and since Zope sits behind the Web server, how can it participate in Netscape publishing?

First, Zope was taught to use the PUT protocol. Thus, when a PUT request comes in, Zope knows how to deal with it. Next, Zope's remarkably perfect integration between its object database and the Web makes a Folder look exactly like a directory on the site's hard drive. The content manager's browser software can't tell the difference.

However, since Zope DTML Methods are objects and not files, some good things and some gotchas are introduced. First, the normal Zope infrastructure for access control is still obeyed. Someone must have privilege to use Netscape publishing for their objects.

Yet, a Zope DTML Method is more than just a bunch of HTML. DTML Methods and Images have other Properties, such as id, title, and access control. Since Netscape publishing doesn't know this, Zope had to recognize this and ensure that updates kept the other Properties.

A final issue concerns DTML instructions that are mixed into the HTML. Netscape authoring tools (Netscape Gold and Netscape Composer) see these as HTML comments and do not display them. This can be both good and bad. While it makes Zope DTML Methods look more like real HTML by hiding the DTML instructions, it makes editing the DTML instructions more difficult.

Previous Chapter | Next Chapter | Up | Next Section | Contents