IMAP's real problem was scale and timing

IMAP's real problem was scale and timing

The problem IMAP had initially was scale.

POP3 is really simple as a protocol, and scales well. But more importantly, POP3 tends to assume (in most client defaults) that mail will not remain on the server. IMAP doesn't, and therefore increases storage use and associated costs.

Back then storage space on large mail servers was a serious issue. The machines weren't all that capable, and solutions like SANs were exotic and expensive.

By the time the machines became mature enough to handle IMAP accounts, the market wanted webmail rather than IMAP, as it doesn't tie you to a machine. (Although it does tie you to a browser.)

Ironically, from what I've seen many of the webmail implementations out there are actually based on having data served from an IMAP server, albeit one that's only accessible on the loopback address or behind a firewall you can't access...

I agree with your client config point 100% - IMAP is not thought of as a viable option for most enterprises though, so it's only to be expected. To be honest, I think it was only added in the first place for the education sector, as some large universities used it.

I couldn't figure out why someone didn't just pick this market up buy the scruff of its neck and provide a simple {XML|CSV} file format that was open and freely licensed, and could be read by your mail client to configure itself for you. Then configuring a mail client could literally be as simple as "click on this link, then type your username and password"...

But I still don't think that would have changed things much... Webmail, once used, has a certain accessibility that's far more attractive to light users than a more restrictive rich client will ever be...

Note: you can avoid this step in the future by logging into your LWN account.