On Aug 18, 2006, at 11:03 AM, Spencer Dawkins wrote:
Hi, Arnt,
I apparently forgot to reply back in March. Sorry.
Spencer Dawkins <[EMAIL PROTECTED]>
I was thinking about this from the client's perspective (if the
collation
doesn't report "error", how does the client know to report
anything to
someone who might be able to fix the input?).
There's no point in doing that. Take my home, IMAP. If an IMAP client
tells the server to sort the messages by subject, and one message's
subject contains an illegal ISO-2022-JP sequence, the collation will
return "error" a few times. That's not a bug in the client or server,
it's just another broken mail message. Usually it's spam. Routine.
If you guys are OK here, I'm OK.
I'm OK Enough here. It's hard to make things truly general. We
really need to solve the IMAP need, and the authors have done a good
job of trying to make it useful elsewhere. Most of the work of
making the registry useful elsewhere would belong in other documents
anyway. For example, if IRIS wanted to use collations to sort
registry entries and report errors, the reporting of errors during
sorting would have to be part of the IRIS proposal anyway.
Then if that fails there's always new RFC numbers for later versions
of this work.
Lisa
_______________________________________________
Gen-art mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/gen-art
