[Foundation] motions

Evan Prodromou evan at prodromou.san-francisco.ca.us
Fri Jun 13 14:23:25 CDT 2003


>>>>> "DW" == David Waite <mass at akuma.org> writes:

    PSA> OK, given that it seems some server developers want this
    PSA> functionality and at least one or two client developers say
    PSA> they plan to implement it, I'll go ahead with the Last Call.

    DW> Curious, this is partially what the original 'JIG' concept was
    DW> meant for - if there weren't enough people interested enough
    DW> to form an 'interest' group, it probably shouldn't be tackled
    DW> as a standard.

    DW> I of course think standards should be based on some
    DW> measurement of need, and that we need to measure this need
    DW> before we hit the 'call for experience' to make a standard
    DW> final.

So, umm: I'm wondering if it's at all reasonable to expect to see a
reference implementation for a protocol before approving it. Many
standards organizations use a reference implementation to prove by
existence the implementability of the protocol, or conversely to
modify the protocol if it turns out to be ridiculous to implement.

Of course, there's often compliance testing suites, too. Like, if you
want to say you support JEP-XXXX, your implementation should pass this
suite of tests. Not necessarily automated tests, mind you, but some
kind of checklist. This concept may be a little wordy for the
quick-and-easy JEP format.

Anyways, maybe these things could be suggested but not required for
JEPs?

~ESP


-- 
Evan Prodromou
evan at prodromou.san-francisco.ca.us






More information about the Members mailing list