[jdev] Jabber Certification Program
temas at box5.net
Thu Jun 17 19:44:32 CDT 2004
On Jun 17, 2004, at 6:04 PM, Justin Karneges wrote:
> On Thursday 17 June 2004 3:25 pm, Julian Missig wrote:
>> Protocols do not move forward by sheer will of the council. The
>> /reason/ those JEPs are /still/ experimental is because of lack of
>> implementation attempts.
> Do we really want to relive the DTCP disaster again? Ever since that
> backlash, I'm afraid to implement anything that doesn't say Draft.
>> JEPs do not move forward by sheer will of Council or JEP authors.
>> Client-based JEPs need client implementations to test them and work
>> bugs... and then the Council can start moving things forward based on
>> real experience.
> Has the Council opinion changed? If so, I wasn't aware of it. Real
> experience is apparently overrated. Time for everyone to go read some
> about File Transfer in late 2002.
> And Tkabber _still_ does DTCP based File Transfer instead of
> Bytestreams. Yay
> for standards.
>> How many test implementations of jep-secure are there?
> jep-secure doesn't even have a JEP number yet, hence the name. It's
> stuck in the council/jep-editor void (for 3 months now). Surely an
> implementation is not expected.
The DTCP argument is old and dead. It was a matter of multiple
standards doing the same job coming out at the same time and then
people pushing and shoving to make something move to the head of the
class. That issue is over and dead, let's move past it, we're big kids
As to the council opinion, implementations are fine, but if you're
using a "real" JEP and it's experimental, than it's expected that
you'll use the current version and then whatever becomes draft. We
can't have ideas that are just written down, they need to be tested.
This is also, related to the recent council discussion about not as
quickly accepting or pushing JEPs through the process, and using the
JEP-Secure has political issues and you know that. Don't make
statements that are loosely blaming other people for that hold up.
There's no reason not to be implementing it and figuring out what the
problems are. For the JEP-84 mods there's no way I would be ready to
update it if I had not started implementing it in Gabber2. I
immediately found many practical problems with the described methods,
and how I initially thought I could fix them.
More information about the JDev