Although some people have suggested that running zope standalone without apache is the recommended way to go, I have heard several people suggest that the Apache solution is preferred. There are a lot of "it depends" qualifiers here. What I've heard is that running Zope behind Apache may provide more better system security. I think the rationale is that apache is more heavily debugged. (This argument may be questionable, as you have two families of bug to deal with though.) It also allows you to run other kinds of apache-based service. It also provides maximum flexibility. A key issue is to configure your httpd.conf file that specifies apache's behaviour file with suitable rewrite rules so that some/most/all requests are handed off to zope instead of being dealt with by apache. At any rate, it's what I do. I believe there are (at least) 2 common ways of doing it. The one I selected is based on "PCGI". Essentially the zope process stays active and apache talks to it via a lightweight intermediate process. It's actually quite easy to set up. Try looking here http://zdp.zope.org/projects/zbook/book/VII/IntegraWebServers/ IntegraUnix/ZopeApacheIntegraPCGI/Drafts/954770544 for details or just search for "zope pcgi apache". Greg

Running Zope standalone is the recommended way to run Zope. If you need virtual hosting support, you can re-use your Apache or Squid as reverse proxy.

-aj

--On Samstag, 4. Mai 2002 9:28 Uhr -0600 Mosquito <mosquitow@shaw.ca> wrote:

...

Hey there, I'm using linux 8.0 with KDE Desktop. I've got zope running on a medusa server at the moment, the one it came with. I've also got an apache server running and I was wondering how I can integrate the zope interface to be used with my linux apache server. I'm a total linux nub so go easy on me .

~Mosquito