Gmail: POP vs. IMAP

Okay, while it’s nice that Google has now rolled POP support to my Gmail account, I kind of wish it hadn’t. Why? Because in the absence of POP I would have cheerfully convinced myself that Google’s Gmail crew was working on IMAP support, that they were going straight past POP without stopping.

IMAP makes so much more sense than POP in Gmail:

  • It is a better fit with the gigabyte of storage
  • It is consistent with the Gmail idea of a delete-nothing persistent message store
  • It is lighter-weight than Gmail, so less server load from clients
  • It would make it hard for me to not make Gmail the hub of my personal emails, given that I could have a universal message store

So why is Google wasting time with POP on Gmail? My bet — and this is unabashedly optimistic on my part — is that it is holding out to offer IMAP support in a paid version of Gmail.

Related posts:

  1. GMail Drive as Lightweight App
  2. The Gmail Economy
  3. Gmail as Atom Feed?
  4. Fun with Gmail
  5. The Tech Support Sinkhole

Comments

  1. David Ascher says:

    I’m sure that Google could come up with an interesting IMAP interface that used virtual folders associated with its tagging system. I’m less sure that many of the IMAP _clients_ could deal with a folder hierarchy that they have to represent, but that they can’t control. I find that Thunderbird for example doesn’t really deal very well with the server changing the folder set “behind its back”; Outlook 2000 had related problems to do with when to apply filtering rules.
    My big unfulfilled dream for gmail is much smarter processing of email based on categorization of email — I find it just doesn’t scale to heterogeneous email traffic (personal, blog comments, mailing lists, etc.), let alone large throughput (as opposed to archive size).

  2. Paul K says:

    Right, that’s close to my wish as well. One of the main reasons I use Caelo’s NEO email product as an Outlook overlay is because of its multi-pointer classification system for email. It lets me deal with heterogeneous email traffic, as well as having messages appear in multiple places/indexes.
    My dream Gmail would combine the multi-pointer aspect of Caelo’s NEO with a nice IMAP back-end that has, in effect, virtualized boxes. The biggest problem, as you say, is that few (read: none) email clients could handle it all neatly.

  3. b7j0c says:

    IMAP client configuration is still an issue…although admittedly anyone smart enough to ask for IMAP likely knows how to configure it

  4. Alexander says:

    IMAP puts considerably more stress on a server. Perhaps that is one reason.