Re: [Evolution-hackers] Simple Calendar Protocol Connector for easy implementation of calendar servers

ext Michael Wyraz wrote:
i've spent some time with research on different calendaring client/server implementations. I believe that most use proprietary or open but very complex protocols (e.g. CalDav) or protocols that are restricted to read-only access (e.g. WebCal). On the other hand there are some really simple protocols which are not dedicated to calendar access (e.g. xmlrpc or json over http). So the barrier for a developer of a calendar (i.e. a web calendar) to implement a bridge to evoluition is very high.
hmm. <shameless rant> I've submitted an open source Apache module implementation of CalDAV <>. I won't argue that it (implementation) is perfect but I'm inclined to say that CalDAV _is_ pretty simple compared to many other new specs (i.e. when you don't have to implement the whole http/dav stuff). The only exception to this may be the iCalendar spec which is probably even too flexible. Fortunately, there's this, many times forked libical libary, which we can/should fix. And Evolution is a pretty decent calendar client as well (although there are some missing features). So with these constraints, I'll doubt that you could significantly simplify or improve the situation by creating yet another new spec ;-) br, jari
Ok, the CalDAV spec is not extremely complex. But it depends on iCal, WebDAV and the CalDAV Spec (which is in fact not a spec but a draft that still might change). The current CalDAV draft (draft 15) was expired on March 17, 2007. The evolution caldav connector is from December 2, 2005 ( It implements the protocol draft version 5 or 6...
CalDAV is already an rfc: <>. Evolution's CalDAV backend mostly lacks freebusy reports (sure acl handling is totally missing). That said you can still start fixing it fairly easily (I do have some fixes at
CalDAV draft has 115 pages. WebDAV consists of at least 7 RFCs, the most important are together about 300 pages. ICal rfc has about 150 pages.
right, but in OSS world you have many of these implementations, why reinvent the wheel ?
I know about 2 (free) servers (ok, since your mail, i know 3 :-) ) and maybe 5 clients which supports caldav. Most (all?) of them supports a subset and all implements different draft versions.

So in my opinion CalDAV is not a simple protocol. It's a big all-in-one-approach for calendaring with the cost of high implementation expense. It will not be implemented in many calendars until stable caldav-libs (client and server) are available.
I assume quite many of us are trying to boost it ...
I'd prefer a proprietary (in sence of non-standardized, but well documented) protocol which is easy to implement in a few hours or days. This gives progammers the opportunity to add client-support to their calendar servers whitout too much effort.


Perhaps I have to stop producing running code, I'm way too slow ;-)
br, jari

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]