[Standards] Simple Jingle example(s) and spec(s) wanted
remko at el-tramo.be
Wed Feb 7 08:53:11 UTC 2007
> I don't see a problem with extensions or alternative mechanisms, as long as we
> don't forget our standards.
Once you have an alternative mechanism, you have a standard, so there
is a choice anyway. And when there is a choice, the best standard
(where 'best' is a combination of most powerful, most implemented,
easiest to write, easiest to extend) will be picked by client
developers, and will survive.
As Peter said, we shouldn't consider the standards we write to be holy
and untouchable if they don't work, and try to progress towards better
standards with the knowledge we have about the past. Of course, we
should avoid making standards that will be obsoleted by all means, but
we shouldn't restrict ourselves by saying that all our standards are
perfect and will hold forever.
The process of replacing approved standards happened with avatars (and
still isn't finished), which everyone considered a good thing; this
happened with composing notification/chat states (which was less
necessity, but cleaner anyway); this happened with VoIP (although
almost no one implemented that standard). We seem to have managed
pretty well with those cases in the past, most of the time
implementing both for a while (practically at no extra cost).
Personally, I'm always excited to implement new standards if i'm
convinced they improve the situation. I especially like to do this if
I know they will make my life easier and my code cleaner in the
future, which is what frameworks like Jingle and PEP do for what used
to be ad-hoc standards.
Then again, all I do is code for fun, I don't know how corporate
people feel about progress in standards and interoperability.
More information about the Standards