Almost 20 years ago Nathaniel Borenstein, then working for Bellcore, proposed to the Internet Engineering Task Force (IETF) the notion that a new protocol be used to swap multimedia data, rather than just ASCII text, over email. Today, that protocol, MIME, is a bog-standard aspect of the way we communicate and is used across email messages, web pages and operating systems. Now chief scientist at (appropriately enough) cloud-based email management and security firm Mimecast, Borenstein took half an hour on the phone to tell us why (SPOILER WARNING: my ancient, clunking Mark Twain quote approaching) rumours of the death of email are exagerrated.

Q. What do you think of Facebook COO Sheryl Sandberg's famous suggestion that email is going to go away because young people are rejecting it?

A. The trivial response is 'remind me what users do if they forget their Facebook passwords?' They mail it to them. Facebook, like other social networks, is a closed environment where users can only send messages to people on the network. Teenagers travel in packs and you tend to have social groups following each other. When those people go into the workplace and tell all their business contacts they have to be on Facebook, they'll lose all those contacts.

Q. But you see a lot of people turning to things like instant messaging as an alternative...

A. It depends on what you're trying to use it for but if I were Facebook I would an add email capability. Before joining Mimecast I worked at IBM for seven years and IM was used as heavily as anywhere but it's for very different purposes. If someone is not online it's useful; if not its useless. There are times you don't want [a rapid exchange of messages]. It's like when you call someone just to get their voicemail.

Q. So do you see a future for email that outlives us both?

A. Yes. With each of these messaging media it builds on the previous generation. Email didn't replace the telephone. Email is for 'store and forward' delivery to any format. It's more capable [than alternatives].

Q. But 20 years ago fax was dominant in electronic business written communications...

A. The fax example is a very rare one and sometimes there are things that are just plain better.

Q. With the rise of unified communications do you see the various types of messaging being masked by the front-end so users don't know or care which medium they're using - they're just sending messages as dictated by recipient's profile, location etc?

A. I do expect that merging to happen but I suspect email will still be called email [and remain distinct].

Q. Do you think there's some generational bifurcation between younger people who prefer the instantaneous nature of IM versus older users who require email?

A. I think there's some bifurcation or trifurcation but that might be a stage in life.

Q. I used Lotus Notes for many years and was wholly unimpressed with its messaging component. Do you look around and feel impressed by how things have improved in mail?

A. I worked with Lotus at IBM for seven years and if you Googled 'Lotus Notes sucks' I'm afraid you'd find me near the top. I was a fierce critic even then and I think it's essentially un-improvable. The reason is there's a lot established code in place and a lot of people are used to it. [In environments of] lower-level workers the need to be retrained on a new user interface [is a constraint]. To Lotus's credit, after a few attempts to do the impossible, they have done something much more interesting in [online collaboration service] Lotus Live.


Q. Are you impressed by some of the newer approaches to email?

A. There are several interesting systems with really good email interfaces and capabilities. My biggest concern is splits in the community though: when one mail system starts sending messages another can't read, that's a bad thing. In terms of standards, it's not moving as fast as it should be but the IETF's support for non-ASCII character email adresses where either side of the @ [symbol] can be Chinese languages will be huge for interoperability because [of the size of the potential new user base] and because a lot of older email systems will just roll over. I also like Mac Mail's integration with RSS and GMail's message threading.

Q. Is it a surprise to you that in so many business scandals, email is still the smoking gun?

A. Something like 84 per cent of business data is held in email but viewing it as a problem is like saying it's a problem to identify wrongdoing.

Q. You mentioned GMail threading and it seems to me that free webmail has changed the user's view of email. Will web-based email services become the dominant form over in-house?

A. I think cloud email will do for business what free webmail did for consumers.

Q. How do you win the trust of CIOs and those outside IT?

A. 'Trust' is ambiguous. There are two kinds of trust in choosing cloud service providers trust: competence and integrity. I trust my wife but I don't want her running my email systems. I trust my sysadmin but I wouldn't trust him enough so I'd want him to date my daughter. With a cloud service provider you need the closest thing to someone who you trust both ways: competence and integrity. Datacentres are full of understaffed and overworked admins trying to keep your email running. You're going to be better off in most instances with a specialist even if their reputations aren't yet as established as we might like.

Q. Apart from SLAs (which CIOs thend to demand then dismiss as not worth the paper) what can cloud service providers do to assuage the fears of executives that don't like data residing in the cloud?

A. They can tell you what they will do if they ever want to leave you. We're sad if you want to leave but were going to help you get your data out. That's a pledge. The reason SLAs are not as useful as you'd like is they're hard to define because if we offer you 99.999 uptime and your Ethernet hub falls over, have we failed? We can't make promises beyond our control.

Q. What do you think Microsoft will do? Exchange/Outlook has been the top double-act in business email for a while now but hye seem to be having it both ways with client/server and cloud offerings...

A. Microsoft traditionally tries to cover a lot of bases and what typically happens is that the paradigm that works is the one they pursue in the long run. They're hedging their bets.


N.b This entry was updated 22 July after a cut finger -- caused by grabbing a basket that contained a razor -- meant that I had accidentally cut and pasted a wodge of copy into another file. Finger and owner now doing just fine, thanks for asking.