ietf-822
[Top] [All Lists]

Re: Length Limit for display-name

2010-09-23 19:10:07

Alright, let's look at an alternative to what I'm trying to do.

Perhaps you should describe what you're trying to do. It's very difficult
to make suggesions without knowing the semantics you're trying to get.

How is it with using Optional Fields (section 3.6.8). Are they safewith 
respect
to "intermediaries perform[ing] all sorts of processing of header fields"?

First of all, the semantics of display-names and optional headers are very
different. A display-name is always associated with a header address. If
whatever you're doing needs to track along with header addresses, then optional
fields do not offer a solution, because agents that mess with header
addresses won't know to make the corresponding change in your private
information fields.

OTOH, if whatever you're doing isn't associated with header addresses, then you
shouldh't have tried to cram the information into a display-name to begin with.

Now, as for using optional fields, the answer is that as long as you pick one
nobody else is using for something, I'd say they're safe up to 4K octets or so.
(You can also use more than one if you need the space.) Beyond that you may
start running into implementation limits again. Putting a vendor name in the
field name is often a good plan - it makes it far less likely some else will
use it for something different.

You'll need to properly fold the field if it is very long, and you should
be tolerant of folding changes.

                                Ned

<Prev in Thread] Current Thread [Next in Thread>