[standards-jig] RE: JIDs (JEP-0029)

Matthew A. Miller linuxwolf at outer-planes.no-ip.COM
Tue Apr 30 16:42:29 UTC 2002


(This isn't really a reply since I'm getting the digests; too much
e-mail as is...)


Overall, I like that someone has taken the time to write up an
informational JEP on JID's.  I know I've asked countless questions about
the structure and limitations... (-:

Here's my $0.02, although I'm sure others will express something of this
opinion themselves.  I just want to be on record.

Having a limit on the resources is not that big of a deal.  I can see
specialized applications possibly wanting to have a "large" size for the
resource, but 256 is pretty darn large.  Others have noted that many
languages only recognize the first n characters as unique, and this
could work, too.  Personally, I don't see this being that big of an
issue (and haven't heard it to be one so far).

Using bytes instead of characters, however, is not a good idea IMO.  It
restricts Jabber(tm) to only those that can express their language in
ASCII.  Also, the the "node" and "resource" are considered UTF-8, which
makes them multi-byte character strings.  This (apparently) imposes an
"arbitrary" limit on the string, which does not result in a fixed
length.

In my most humble opinion, a fixed length of 256 *characters* is
acceptable, but not 256 *bytes*.


Thanks for allowing me to rant...

-- 

Matt "Linuxwolf" Miller

- Have you "Jabbered" recently? (http://www.jabbercentral.org/)




More information about the Standards mailing list