Re: [Evolution-hackers] calendar implementation (Partly OT)

[Adam Williams <awilliam whitemice org>]

>But that is the whole point, I would not mind starting this project, but not
>based on BEEP. CAP is theoretically sound ( I read the draft, this
>afternoon) but indeed vaporware.
>Smartest thing at this moment would be to use xml-rpc or http as protocol
>engine.

Agree.

>Point here is that it will be to big for one person, design decisions will
>have to be made and it would be nice if we could get some idea's support
>whatever from ximian. I think it would be really good if we could make the
>server in such a way that writing a backend for evo would be working fast

My point is that several working, if not ideal, servers already exist. 
Both Horde and PHP Groupware have gone out of there way to support
"external" clients with an XML-RPC API.  Both these products work,  and
web based calendar access is an important feature.  I don't think having
a Apache/PHP application as your core calendering solution is perfect,
but these WORK, now.  If someone creates a client module; when your done
writing the module - you have something that you can use, right then.  A
calendering server that doesn't support shared calenders, etc... might
as well not be a calendering server, and both the above mentioned
solutions do so.  Writing a calendering server with sufficient features
is very NON-trivial, evidenced by the number of derelict and abandoned
projects littering the net.