The market is people who want to run their own mail server, but haven’t already either spent the week figuring out existing software or just purchased an account where the hosting company manages the email servers. That feels small and hard to reach, to me. Where will you find a thousand people (per year) who’ll buy this at $250? Writing an email server is easy. Figuring out how to market an email server, now that’s hard. ;)

Also, if you want to run your own mailserver, having redundant systems (ideally on a UPS and redundant network connections) is also important. Which, actually, raises an interesting question: would that be the additional feature that makes this a great product? How well is failure handled by pop3d and imap4d and their friends?

Btw, authentication should be modular, even inside the app, so you can start with ’simple’ built-in authentication, but expand to include links to system accounts, LDAP, etc. It ought to be a “30 seconds from install to serving email for every account on the system”, because for a lot of little easy-install servers, figuring out where the passwords go is one of the trickier bits that always annoys me.

Still, even if this is a weak idea, the analysis of the marketing and business side is already more detailed and useful than 99.999% of all software development blogs out there, and I *really* hope you’ll keep that up. That’s the really itneresting bit of this series, to me, at least.

After fighting with sendmail (and a hosting company with flaky uptime) for two years, it was a huge relief to move all email-related functionality off my server entirely.

Thanks, everyone, for the feedback. I’ve got to admit that this is one of the weakest of my 30 ideas. It’s hard (though not impossible) to make a compelling case, from either a marketing perspective or a technical perspective.

But, nevertheless, it is something I’ve thought a lot about over the last few years, so I decided to put it out here anyhow.

Another point is why will a non-techie small business owner be bothered with linux based systems or hire a “Guru” on a retainer.

Price point wise selling your s/w over $250 might be difficult i guess.

Ryan

* Webmin is not a “good” interface. It’s ugly, and written for Unix system administrators. I’d think a product with simple terminology and a straightforward object model would be most likely to succeed in the market. If you have to ask a very technical question, see if you can eliminate the need for that question (infer the answer or make a one-time decision yourself) or else provide an easier-to-understand explanation about what the concepts mean. We could all do a lot better than Webmin.

Having said that, I guess making yet another mail server, especially for geeks (who can and often proud to be able to configure things themselves) must be a bad idea.

For example, many small shops might keep their users in Window’s active directory, and use pam_ldap on the linux side. If you aren’t using PAM, then either you need to independantly support LDAP or you alienate a whole chunk of your audience.

Ditto for about a whole bunch of other auth options….