ned.freed@innosoft.com
Sat, 11 Mar 2000 16:43:41 -0800 (PST)
> Now I remember :-)
> It was one of the opitions described in <draft-ema-vpim-imap-01.txt> for
> message length indicator. I have exerted the section below. The original
> intent of this was for clients to get this information via IMAP and present
> it on the client.
> Despite the ugliness of subject overloading, I am afraid that is what a lot
> of products implement. I guess the question is do we want to endorse this,
Well, an additional question is whether you'll be _able_ to endorse it in a
standard. Let me say right now I'm strongly opposed to such mechanisms. For one
thing, endorsing one may open the door to endorsing lots of others. And unlike
other recent discussions like the "-xml" tag business, this is NOT an academic
concern: There are so many subject line label thingies floating around already
it isn't funny.
Another issue is the potential for conflicting use. I would not care to
bet that there isn't a conflicting use of parenthesized colon-separated
numeric values other there -- indeed, I'm pretty sure I've encountered a couple
over the past few years. Stuff that looks like it has a fair potential
to fail to interoperate doesn't make it into IETF standards.
And finally, past experience with embedding information as tags in unrelated
header fields, even in contexts far less overloaded than the subject field, has
been dismal at best. Two examples come immediately to mind: Storage of original
and recipient information in comments in Received: fields and storage of
certain delivery/return options in comments in To:/Cc:/Bcc:. Both of these
uses have been the source of numerious operational problems.
> or create a better mechanism (like the Conneg suggestion or one of the
> others below). If we choose the latter, the client vendors have to agree to
> implement it, or the industry will likely tend to what they do now.
Conneg gives us a framework for this stuff. I think we should use it.
Ned
This archive was generated by hypermail 2.0b3 on Sun Mar 12 2000 - 03:10:39 IST