[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: RFC 3280 error WRT rfc822Name


At 12:10 PM 9/25/2002 +0200, Marc Jadoul wrote:

Sometimes it is better to break things!

I agree that can be true, though you generally need a transition plan.

Having 2 users with the same email
address except cases is really a stupid thing to do,

That's probably true in general, but someone some where may have a non-stupid use. For example they have layered some application on top of email that involved user identification that is case sensitive and that application is one that can't be modified. Or maybe some one has a gateway to another mail system that is case sensitive.

 and you won't break
that much by changing that!

RFC-822 email addresses are used by more than 100 million users and thousands of products. These addresses are even used outside of email in things like SIP. Today the rules you follow for this particular part of the standard are clear.

If we were to change the rules to say that case doesn't matter then we should say that all software using these addresses should do comparisons independent of case. It would take years for all the currently deployed servers and clients to be upgraded to have that behavior. We'd probably never really get it straight. It would probably also take a few years to update RFC 2822.

Seems more like fixing things to me.
Does it exist 1 real world example for having as requirement 'cases
sensitive email addresses'?

Don't have a really good concrete example (which doesn't mean one doesn't exist). What I would offer is the fact that UNIX, Linux, and Solaris login names are case sensitive even though sendmail seems to ignore the case.

LL