On Fri, 31 May 2002, Larry Osterman wrote: >Actually I'm pretty certain that (I think) pine will generate message >sets in the form of a:b where b>a. I remember that it surprised me when >I first saw it, so I quickly changed the server to handle that case. >If it wasn't pine, it was one of the other common MUA's.
At least Outlook Express uses a fetch to MAX_UID+1:* to check for new messages. As much as I hate to say so, Outlook Express isn't breaking any point in the RFC AFAICS. In my first implementation, I assumed nothing about the bounds in a message set. I don't understand why anyone should do so. Andy > >Larry Osterman > > >-----Original Message----- >From: Arnt Gulbrandsen [mailto:[EMAIL PROTECTED]] >Sent: Friday, May 31, 2002 8:36 AM >To: Paul Smith >Cc: [EMAIL PROTECTED] >Subject: Re: max+1:* fetches > >Paul Smith <[EMAIL PROTECTED]> >> >Precisely. As far as the RFC says, "1:2" and "2:1" are equivalent. >> >> It doesn't say this.. (as far as I can see). It's open to >interpretation >> from reading the RFC. > >Exactly. ;) > >It says a:b means all the messages with MSNs/UIDs between a and b, >inclusive. That's clear. > >It does not require that a<b. The absence of a requirement is also >clear... more or less. Sometimes the non-requirement can be usefully >mentioned. > >--Arnt > -- Andreas Aardal Hanssen
