Xref: utzoo comp.protocols.tcp-ip:16777 comp.protocols.tcp-ip.ibmpc:6296 Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!van-bc!ubc-cs!uw-beaver!milton!mrc From: mrc@milton.u.washington.edu (Mark Crispin) Newsgroups: comp.protocols.tcp-ip,comp.protocols.tcp-ip.ibmpc Subject: Re: Where Can I Get The IMAP Specification? Message-ID: <1991Jun29.030833.9957@milton.u.washington.edu> Date: 29 Jun 91 03:08:33 GMT References: <2869DA1E.614E@intercon.com> <1991Jun28.024011.10286@milton.u.washington.edu> <1991Jun28.224703.9915@socrates.umd.edu> Organization: University of Washington, Seattle Lines: 36 In article <1991Jun28.224703.9915@socrates.umd.edu> coates@UC780.UMD.EDU writes: >Wis IMAP? Please post. IMAP is an Interactive Mail Access Protocol. It differs from the POP style of protocols in that IMAP focuses on maintaining a mailbox on a remote repository and not on a client. POP on the other hand is designed to fetch mail from a repository onto the client and remote it from the repository. In other words, they do fundamentally different things even though the idea (email on a personal machine but using a central machine as a maildrop) is the same in both. IMAP also has some fancy facilities to make it possible to do various things (e.g. remote searching) that would be prohibitively expensive with POP. IMAP was designed to scale well with very large mailboxes and with high latency/high RTT networks. I regularly read my mail from home on a NeXT using 2400 baud SLIP with IMAP. Another neat thing about IMAP is that it has the capability to parse the message remotely, so a client doesn't need to know about all the details of RFC-822 or of the new multi-part/typed mail body stuff. It's all converted into a easily machine-readable structure unaffected by the various perverse ideas of visual esthetics which inflict RFC-822 headers. IMAP was first conceived by me in 1985. That first version is now extinct. The modern version, called IMAP2, was originally documented in my RFC-1064. That RFC had a few typographic errors introduced in the RFC editing process and a few extensions to the protocol had become widely accepted. I issued a subsequent RFC-1176 to replace RFC-1064; it corrected those errors, documented the (minor) extensions, and clarified some aspects of the architecture that people had found confusing. IMAP2 has been implemented on a variety of client and server platforms; unfortunately for some reason a finished PC client has eluded us. A number of people are reportedly working on a PC client.