Hacker News Reader

Comment by ultimoo | Hacker News Reader

It's amazing as to how ubiquitous email has gotten in the recent years in spite of being a relatively older technology. It was only a few years ago that you needed to afford the luxury of a blackberry to email on the go.

And now we have very affordable phones zapping out emails over reliable cellular data connections, hundreds of small apps and companies curating and shooting out content-rich emails that can be viewed in your palm, and browser based email clients on desktops evolving to have features comparable to installable thick clients. I wonder what else is in store for the trusty email.

View on HN

It's not just relatively older, e-mail is one of the oldest technologies on the net (we're talking 1960s technology here!)

Frankly, I wish it would die already as it's pretty much held together with ducttape:

* it's all transmitted via ASCII; so any binary attachments have to be base64 encoded that increases their file sizes by ~30%;

* transport encryption still isn't the norm (this is the 21st century, why the hell are we still sending stuff in clear text!?),

* there's no standard error responses meaning it's often guess work whether mail has been received or not

* there's no standard way of rendering HTML mail, so anyone sending mailshots has to revert back to a 1990's web developer mindset (ie assuming that everything doesn't render so resorting to HTML tables and hours of frustration).

* it's not real time / real time daemons are not standardised. (I'm not saying we need something like instant messengers, but it seems silly that we have a small handful of proprietary push systems (eg Exchange) and the rest of us normal users have to instead rely upon re-syncs every 5 minutes (which means a fresh set of handshakes as well). A persistent connection would make much more sense for a 21st century e-mail network.

There's been so many attempts to "fix" e-mail, but most have either missed the point (eg Google Wave) or just added yet more ducttape to the existing problem (eg pretty GUIs that change the way we interact with e-mail clients but without actually touching the core technology itself).

I might actually write my own mail server and client one day because it's becoming obvious that most people are happy with "good enough"

I would add postage to the protocol, as well. Spam consumes so much resources because it's basically free -- adding required computational complexity to send a mail (or, even better, an escrow based real money postage system) would go a long way towards cutting the legs out from under the spammers.

Network effects mean that all this is idle speculation, however.

I don't think it would make that much difference to spam as most of it is sent from hijacked machines anyway and the legitimate stuff is heavily regulated so must have features like "unsubscribe".

There definitely does need to be a better way of managing it though. But I'm yet to find a better way to manage spam (every method I do consider would either be ineffective or open to abuse to blacklist legitimate senders). But I guess that's the $million question.

Reply on news.ycombinator.com