18 Mar
2003
18 Mar
'03
3:23 p.m.
Hi Janko,
first of all: Thank you very much for your help!
Ok, I see! This is something like the virtual host configuration of
Apache. The reason why we start and configure swazoo from code and do
not use the configuration file is the following:
1. SmallWiki should be running out of the box, just by loading
the source from StORE. We do not want that people have to
modify external files before they are able to run it.
2. We want to stay independent of any external library. So
web-servers, storage, etc. should be pluggable in our wiki.
server :=
Swazoo.HTTPServer new.
server addSite: self defaultWikiSite.
server port: 8080.
server ip: '*'.
server start.
Swazoo philosophy is a bit different as you start above. You first
make a new Site, configure host: ip: port: combination, and then start
it. This will start appropriate HTTPServer automatically. This also
makes possible to run many wikis on the same Swazoo server, and not
only wikis, you can run other resources (like AIDASite) in parallel.
Very usefull for hosting a lot of sites on the same server. Admins
love that :) So, in my new AIDA/Web with integrated Swazoo I start
a thing that way:
site := AIDASite new name:'eranova';
host: 'www.eranova.si' ip: 'mici-web.eranova.si' port: 80.
SwazooServer singleton addSite: site.
SwazooServer startSite: 'eranova'
So AIDASite is a subclass of Site, I think. Until now I did use
Resource as superclass for my SwazooWikiSite. Do you think I should
change this?
The main problem is that I parse and analyze the url all by myself.
Actually I do not care if there are additional path elements before.
But when overriding the message #answerTo: of the Site-class it does
not work! Do I have to change something else?
Thanks again for your help
Lukas
--
Lukas Renggli
http://renggli.freezope.org