[Members] MOTION: JEP-0106 (JID Escaping)

Peter Millard me at pgmillard.com
Thu Mar 25 10:50:28 CST 2004


Richard Dobson wrote:
>stpeter said:
>> The problem is that, even though JEP-0100 is experimental, it merely
>> attempts to document best practices among gateways, some of which have
>> been around for almost 5 years. The question is whether we want to break
>> all the existing MSN gateways.
>
> I would say yes, since they will have to be reprogrammed to properly support
> jabberd2 anyway, people are not going to want to be running both jabberd1.4
> and 2 along side for very long, also surely its not that big a change in
> reality anyway. Yes this causes some short term problems but IMO it is
> definatelly worth it in the long run to have a much more standards based
> approach to escaping rather than using a less than desirable solution to the
> problem, IMO we shouldnt set a precident and restrict ourselves because of a
> bad implementation decision, otherwise we will encourage this kind of thing
> in the future.

What about ALL of the legacy clients that folks still use that do the old-style
JID escaping?? Are you going to somehow force folks to upgrade (which in a
corporate environment is not always easy)? If I'm understanding you correctly,
your saying we should use URI escaping because "everyone already does that", and
we don't care about backwards compatibility??? That just doesn't make sense. As
a community, we've ALWAYS strived to maintain backwards compat, and to sacrifice
it here just for the sake of being able to re-use code that is down-right
trivial to create is crazy. Do you have other arguments for not seeing this JEP
move forward?

IMO, I also disagree with your assesment that folks are going to "quickly"
migrate to jabberd2. It's a non-trivial task to move a large installation (like
jabber.org) over to the new sever. Folks are going to continue to use 1.4.2/3
until they need something that j2 offers, or security problems are found in
1.4.2/3.

pgm.




More information about the Members mailing list