Hacker News new | past | comments | ask | show | jobs | submit login
Email Is Still the Best Thing on the Internet (theatlantic.com)
435 points by plg on Aug 14, 2014 | hide | past | web | favorite | 197 comments

Wholeheartedly agree with the article. I have absolute control of my inbox with filters, labels, and signing up for newsletters and/or updates on various subjects. There is no way a centralized end-to-end service is going to eclipse email for me, unless they radically change their business models.

The value proposition is just really bad in all the services I've seen so far.

"Oh, you want me to sign up for your service so that I can look at the content you think I should see alongside the ads you're making money off of? And what exactly is in it for me?"

Something I would pay for: a rolodex social network. No centralized feed. No useless info. Your profile is 2-3 sentences and your current city (with some sort of maps integration for when you travel, to see who's near you). Two buttons, one to request to view resume, and another to request to view email. That's it. With the idea being, you use the site to enable you to keep up with people. You add people you know or have worked with to your network, and you can easily get their current email and catch up when you're in the same city. Simple, no obnoxious ads, no slimy tactics to increase time on the site.

Probably will never come to pass, but I can dream...

"Something I would pay for: a rolodex social network. No centralized feed. No useless info. Your profile is 2-3 sentences and your current city (with some sort of maps integration for when you travel, to see who's near you). Two buttons, one to request to view resume, and another to request to view email. That's it. With the idea being, you use the site to enable you to keep up with people. You add people you know or have worked with to your network, and you can easily get their current email and catch up when you're in the same city. Simple, no obnoxious ads, no slimy tactics to increase time on the site."

I really want this too. I need some sort of simple contact rolodex that simply shows who I know, why I know them, and why they're important (where do they work, how do I know them, who are they connected to, etc.). Just a simple UI, something that I would use most of the time simply to search.

Probably overkill for your needs, but I use CiviCRM (https://civicrm.org) to do that for my work. It's basically a centralized web-based contact list, for which you can then link them between each other (different types of relations), add custom fields, etc. -- the program does a lot more than that (online fundraising, mass mail, event management, and other tools mostly aimed for non-profits/volunteer groups), but you can also just use it for that.

It's then possible to expose the contact data to LDAP using "ldapcivi" (https://github.com/TechToThePeople/ldapcivi), so that the contact data can be queried from any mail client (including phones).

We also integrated CiviCRM with our IRC bot, so we can query emails and other info from there. Ex: "who is John Doe" on irc will respond with the basic data (name, email, phone), as well as the relationships of that contact, ex: "John is Manager of foo at ACME".

"why I know them" is a crucial feature missing from many services (along with "what name or alias I know them under").

I have hundreds of contacts from church, school, work, video games, etc.

Sounds like google's circles. The problem is that most folks treat their Gmail account as their "real" account and don't give it out willy-nilly. Also, that the circles thing is one-way - I can use circles to control how I see others, but can't use them to control how they see me.

> Also, that the circles thing is one-way - I can use circles to control how I see others, but can't use them to control how they see me.

No, if you share something to a circle someone isn't in they wont see it, so they do let you control how others see you. Not a huge G+ fan, but I was a fan of circles.

I mean having a fully different profile and name, not just different posts.

I was not that big fan of circles, made sharing with the wrong people easy.

I like that my personal life (facebook) and professional (linkedin) is two different networks. Its hard to accidentally share the the bad joke with my professional network.

I think something like circles could be done right -- if each circle had a custom background, etc. so that it looked and felt like a different website. I'm not worried about posting political rants to LinkedIn because it's obvious I'm on LinkedIn rather than FaceBook or Google+; I wouldn't be worried about posting to my coworkers group if there were very obvious signifiers that I was in that group.

which is good. i don't want some guy i met at a work thing once adding himself to my best friends circle.

IIRC, Facebook used to have this feature (ages ago), but it was removed. It used to ask where you knew someone from when sending a friend request, and it didn't let you send the request if you didn't know the person.

In the early days it allowed "I don't even know this person" as a reason.

yeah. Linkedin needs to implement this. Just a way to add notes that are only view-able to me would solve the problem.

On the profile page of your contact there is a "★ Relationship" button. Click that and it will let you add general notes or more detailed information about how you met.

who is that information viewable by?

I don't want to put in "obnoxious guy from conference" and then have that person be able to read it.

According to the LinkedIn help center [1]:

"Your note will be visible in the Relationship section of the profile and will only be visible to you."

Of course you may wish to store such information in a different system to avoid the possibility of LinkedIn inadvertently leaking the data.

[1] https://help.linkedin.com/app/answers/detail/a_id/43370/kw/W...

Would you trust them enough to be certain enough that it would always stay this way so you could safely put slanderous information about a person here? I certainly wouldn't.

Yeah, something that stores the data in a system I control and then pastes it on to the page with a greasemonkey script would be optimal.

But, then that would be work.

Too much work for the user or the developer?

For me... but I was assuming that I would be the developer. (and I don't know greasemonkey or javascript, so... it would be work.)

huh, cool. that's what I was asking for.

> I need some sort of simple contact rolodex that simply shows who I know, why I know them, and why they're important…

Would you be one of them? Would you actually keep and update yet another online profile, so that other people can find you on this particular network?

> something that I would use most of the time simply to search.

The answer to the above question from almost everyone you might want to search for seems almost certain to be “no”.

Have you tried Rapportive? https://rapportive.com/ It shows additional information about the person in your Gmail sidebar.

Also you might want to check out Nimble, like Rapportive it pulls info from other social networks. If there's a feature you want, it seems they listen to ideas.

I've also used http://www.cerberusweb.com/tour for sorting email and importing additional data about people, but that might be overkill. With Cerberus you would have to import the location data info yourself, for example. It's a "power user" type of application, very versatile.

Something like Contatta? I used to work there a year or so ago, so I may be biased, but still feel it's an improvement over email and sterile CRM. I personally like how it brings several pieces together, even if it's still missing a calendar for example. It's still in open beta, but I know it's extensible too. I may sound like an ad, but there's nothing for me to gain from it now, just my opinion.

I picture something as simple as "Gmail Directory".

Allowing you to share your contact info with people you choose and request access to others information.

Humin sounds like what you're looking for: https://www.humin.com/#/product

Humin looks interesting but logging in with facebook is a deal-breaker for me. Logging in with google would be slightly better, since all of my contacts are there already and if Humin doesn't integrate with my current contacts I'm not sure what value it has.

Contacts+ uses my google and facebook and twitter credentials to link to those services, not to verify who I am - and if I don't want to link to FB or twitter, I don't have to; I'm definitely uncomfortable with being forced to use a particular identity provider.

I wonder if Humin has really thought through their identity and credential management requirements. A great many sites these days allow one to create an account simply by logging in "on the fly" using an existing email address and a password you enter then and there. One email validation email and you're good to go.

Why would Humin need more than that?

"Why they're important". I feel like that is a pretty tough one to generalize.

For people that you _might_ forget it should be pretty easy to summarize in a sentence or two. Two sentences should easily cover when, where, and why you met. Throw in common contacts and you're good to go. Obviously these same two sentences might not encapsulate why your wife is important to you, but, really, you aren't quite likely to forget that either.

I just meant it felt kind of creepy.

Why do you say that? It seems like an important thing to keep track of, for people who you don't keep in regular contact with. Prevents "Who the heck is Greg Bobberton?", if you can just look and see "Greg Bobberton: that guy from the Widgetpalooza conference, was interested in that one idea I had" (obviously with less vague specifics).

Why do you feel it would be a creepy feature?

the UI is a bit off, but doesn't Google+ circles provide this functionality?

Or the "Notes" section in the Contacts app on iPhone?

Or LinkedIn even?

What is missing or wrong with the current social network/address book things?

How much would you pay? I might build that.

It's called "Contacts.app".


I recently gave serious thought to building something similar to a rolodex social network, but eventually poked too many holes in it to follow through.

I'd just come back from a conference at which I'd been asked for business cards by lots of people, maybe a hundred. Of all those people, I only really cared to hear back from a few. Though it's nice to have my name out there and be cordial in the industry, I wanted to place some kind of filter on all the vendor spam that inevitably followed. If...I could easily provide "redacted" or "enriched" contact info, maybe via QR on a smartphone screen, for instance, it would be possible to meter who gets which contact info without hurting any future networking opportunities.

If any of you can figure out how to make this work, I'd definitely pay for it.

This is totally off the cuff, but here's my first thought:

To sign up for the site, you need only a username, password, and verification method (email/sms/facebook/something to authenticate you as a person, but this wouldn't necessarily be tied to the account). The usernames are unique, and your personal url is domain.com/username or something along those lines (like reddit).

Then you can give your business card to everyone and their spammy recruiter, and if they follow up and visit your domain.com/username, they can hit the "request email" button, at which point you can choose which email to give them or ignore it all together. Potential problem is that unique usernames will mean that professional ones run out quickly. Not sure at what scale that becomes an issue....

In other news, if anyone here is going to be at hackMIT and would be interested in working on something like this, email me (its in my profile).

Hmm, a kind of double opt-in? (Or possibly triple if you count handing out the card in the first place.)

Interesting idea, at the same time there are ways to replicate this today in terms of 'send me a blind message on Facebook/LinkedIn and I'll maybe reply' and it doesn't seem to work well, at least not in my experience. It's about the power dynamic, fundamentally.

Doesn't the facility to redirect email from certain addresses straight to a spam box or to higher priority accounts make regular email filters more-or-less functionally equivalent?

I don't think I give my business card out to enough spammers for the need to manually opt-in everyone that wants to email me to be a less laborious task than blocking/ignoring the occasional promotional mailing.

I couldn't reply on your last comment, so it will be here. How would be your "9 or 10 cards"?

I imagine you can just control what info a conctact would have access. Maybe: professional email, personal email, office phone, personal phone, where do you work. Maybe one profile for each info..but that is 5... how to reach 9 or 10?

I think of a relationship regulated by tags. Tags similar to Google+ Circles. You can use to set context: <church>, <college>, <company>. Or access: <prospect>, <top investor>, <important network>, <conference contact>. You can define an access profile for a particular tag.

You can change contacts though an app, but he won't have any access until you alocate a tag to him. So you won't offend anyone upfront, as even the VIP investors will have to wait until you put an all-accesstag on them. So it is a much more subtle acknowledge when after weeks the contact still don't have access to your email.

I could imagine setting up the following cards right away. And by "card", I'm not specifically referring to a piece of paper, but perhaps some kind of contact-info-related digital object.

* BFFs

* Recent acquaintances

* Close family (reasonable people)

* Extended family (think annoying newsletter emails)

* Co-workers (cell number included)

* Clients (cell number definitely not included)

* Spammy business contacts (vendors)

* Travel (something I could hand out with temp hotel and phone info, etc.)

* Japanese (my wife is from Japan, we have lots of friends here and visit there often)

I could also imagine handing out specific cards at various social functions. For example, at a writing club I used to attend, I'd want to provide a link to my portfolio online. I'd never ever want anyone else to read that embarrassing stuff though. In another case, I used to attend a board game club, and it would've been nice to hand out cards with links to my profile on boardgamegeek, meet-up profile, etc.

It seems like this would work best if the exchange format was simply vCard or some other open format. You get to control who gets what at the point of exchange, and they can consume it into whatever system they like. There's no after-the-fact futzing with tags or permissions or contact management; it's just a system to control how you present yourself. The whole idea is probably DOA if you can't interface with the rest of the world's contact software. Maybe you allow connections if both people in the exchange are users of your service, but you wouldn't want to take it for granted.

As an aside, this got really interesting when we considered the "hot girl in a bar" situation. Let's say she gets asked for contact details many times, it's too loud to hear names or whatever, so she might need to snap a picture of the guy at any point of exchange, then review and modify permissions later. But that seems kinda fussy as a system.

See Boston Review (Mar 2014) editorial on pseudonyms, with responses by Frank Pasquale, Bruce Schneier, Richard Stallman, Evgeny Morozov & others, http://bostonreview.net/forum/reed-hundt-saving-privacy

> If any of you can figure out how to make this work, I'd definitely pay for it.

You have two versions of your business card. One goes to anyone and has a junk email. The other goes to wanted contacts, and has an email that you pay attention to.

Exactly. For most people, that's probably enough. I couldn't see the business case for creating a whole social network out of the concept.

For myself though, I could easily see creating 9 or 10 cards for different situations. It'd be nice to have an easy way to manage all that.

What you need is an "premium" email at a "premium e-mail provider". :)

You signed up as Joe Doe and plug your basic info: website url, physical address, your real email address, phone number cell phone, etc. All this can be in multitude quantity, like your email: office job 1, office job 2, personal. All this info is confidential and only seen to you.

Then your conference comes in. You click "create addresses" and system is randomly generating your email addresses, like this:





Then for each email you can setup few options, like:

- forward this email to office job 1

- forward this email to office job 2

- forward this email to my personal email

- if someone emails me at this email, send them autoreply with short profile: my homepage url, my pshysical address, my office phone, but no cell phone.

- if someone emails me at this email, send them autoreply with my personal cellphone.

When forwarding your email you can have optional header, such as extra info you added initially: "this is my Word Expo 2014". Then when 2016 comes in and you receive an email, header tells you the origin of it.

You can setup any email to hold on mail and not forward it at all and just notify you once a week/month/once there is X emails awaiting.

At any point you can dispose an email or simply set it to auto-expire. Or change its settings for that matter.

You have two options when it comes to your presentatione. You can go high tech and print (or request print) of business cards with QR code where each code is your CRF card with different email of course. In this scenario you could have different color of qr code representing different email setup and just memorize: green qr code - potential investors, yellow qr code - potential employees, blue qr code - cute blonde that i think likes me!

Or even simplier solution: order yourself a few thousand business cards with your email address like this: "joe-doe_______@email.com" and then just write down the email number after the pre-printed part of your email address that will match particular premium email #. And if someone asks about it, just tell them a perfect excuse: "oh yeah i get so much spam that once a year i have to create a new email for myself".

Someone building such a system could call it similar to gmail so your emails look normal, like maybe pmail.com (for "premium" mail) or something like it.

I think the problem is that this is very niche market. I am sure for a right solution you would pay even $99 per year, but again it would be hard to market and even harder to make some money long-term off of it :)

You can actually do this with gmail itself: johndoe+x@gmail.com is a valid alias for johndoe@gmail.com, for any value of x. So you can make up whatever set of custom addresses you want; they all go to the same account, and filtering/forwarding/etc. can be done with standard gmail filters (not sure about autoresponses; you'd probably need a third-party client for that).

Of course this plus-sign behavior is a pretty well-known fact about Gmail, so it'll be obvious that you're bucketing your email in this way. Anyone can easily read off the 'base' address johndoe@gmail.com, so you'd need to set it up so that mail to this address is heavily restricted and anyone wanting to actually reach you would have to use johndoe+secret@gmail.com or whatever. But I think this would basically also be true of any custom service you could invent having similar functionality.

But gmail basically does this now with the +: joe-doe+1456212@gmail.com, joe-doe+_____@gmail.com, etc.

Exactly. Anyone know this "trick" so I can envision people that didnt have the response back in a day or two, just remove the + part thinking exactly what you said -- my email went to some less-important folder.

I, for example, always remove the + part. Not sure why, just a behavior. Even if wrong, still it is what it is.

Besides, I don't think you can discard your email. So if you had joe.doe+aaa@gmail.com and month later you decided you don't want anymore email from this account, you would have to setup rule for this email to go to spam, which is not a perfect scenario. Better would be an auto-response that e-mail expired. Something like PlutoMail is doing.

> If any of you can figure out how to make this work, I'd definitely pay for it.

I don't give out business cards. I tell people to google me and send me an email or a tweet or whatever they end up finding.

This filters out everyone who doesn't have a really good reason to contact me because they just won't care enough to invest the 30s of extra time it takes to type "swizec" into google and click a link.

I think the main purpose of business cards today is to serve as a physical reminder of who you met, not to provide you contact info which can be found elsewhere.

Memory is not the best keeper of information at conferences where you might meet tens of people.

Well, based on the thread I created a quick mockup of the app I envisioned: http://invis.io/9K181IPES

If anyone cares to contribute (or develop), feel free.

A social network used explicitly for exchanging contact information would fill a perfect niche for me.

That idea is fantastic.

You can use Facebook to accomplish everything you can do with your social network. Make your profile 2-3 sentences and your current city, allow people to request to see email, and just ignore the feed.

The advantage of course is that there's about a billion people already on it.

I actually figured out a domain for this idea before I gave up. The problem is I can't figure out how it scales correctly. It ideally would A) have strong network effects B) you would get contact data through your interfaces of choice (iphone, email address book, gmail, android, etc) and would rarely if ever interact with a web gui for the service. It should also have a circles concept so that I can give different people different details.

those all seem to rule out the various business models. hope someone smarter than me figures it out.

You need to find a set of people who are connected and connectors, and who have this kind of problem in spades. Then build in sufficient virality so they all start using it, and as you point out, multi-platform presence. Many folks have tried to replace the business card and failed, but I've seen cards with custom URLs rather than contact details on them; maybe that's one way to go. (QR codes require too much friction.)


Limits the network effect dramatically. Best I could come up with was some sort of power user tools or something.

Cobook sort of seemed to be going in this direction, but they got acquired by FullContact:


Something like the .tel domain? http://telnic.org/ You don't get to run your own server, all .tel domains are run by the same company. You just set up yourname.tel with your contact info. Example: http://mark.tel/

http://mark.tel (and about.me mentioned in another comment) is a prime example of a problem with using given names/last names to form URLs to share contact information. Say you get http://smith.tel - this is great for you but bad for every other Smith out there. http://mark.smith.tel doesn't solve it either.

OTOH using some sort of central-issued IDs (NI, kennitala, PESEL, SSN, INSEE, passport number) like http://87120402424.pl.tel feels a bit dystopian...

everyone should be issued an IPv6 address at birth

That was called Plaxo.

It's still around and does exactly what this thread is asking for. http://plaxo.com

Are microformats still a thing? Because business cards as a social network sounds like a great idea.

This (and sister comments to mine) are super interesting, we're working on this problem space - with a professional focus - and there are a bunch of folks in the consumer side too. Happy to chat to anyone interested in solving this sort of problem.

I had high hopes that http://about.me would shape up to be this, but ... nope.


Why isn't it? Simply too little adoption, or are they doing something wrong?

> a rolodex social network

As someone who recently came across an actual rolodex, with both typed and hand-written notations throughout, this idea sounds like something I'd seriously consider signing up for.

Agreed. I actually enjoy email but would really like a great App/SaaS that would be a good rolodex.

> and signing up for newsletters and/or updates on various subjects

I still think RSS trumps email for this...

Agree, but I like having everything in one place. I don't get a lot of email, so that may change, but I have a filter and separate inbox[1] set up for stuff I read, including RSS.


Isn't this basically the same as your email contacts list with a free text field?

No, because people have too damn many email addresses.

I typically have multiple email addresses at all times, with one constant (my gmail) and the others revolving around my employer, my school, and other affiliations. And which email I give you depends on how I know you (I'll give classmates and potential employers the university one, etc). So even though I might have someone's corporate address, if they change companies, I'm SOL. With this service, I can click your "request email" button, and you tell me the most convenient address to reach you at.

> I have absolute control of my inbox with filters, labels, and signing up for newsletters and/or updates on various subjects.

I see you are not using gmail.

We'll build it. Don't worry.

Didn't tent.io try this?

As this article alludes to: the heart of email's longevity, the thing that prevents it being closed by a single entity, is being a classical IETF protocol: federated, decentralized, open and interoperable.

The first two of those properties arise from being based on the DNS for SMTP endpoint discovery.

This is why every protocol needs to specify that it uses the DNS, and how.

And that is why I get so worried that the draft HTTP/2 editors so steadfastly refuses to do so.

It would be useful to draw a dependency graph of IETF and W3C protocols that characterizes each one on their degree of divergence from classical IETF values. Identify the point(s) in time where history took a wrong turn, revert to that point, fork and start over with new protocols. It can't be any less disruptive than new protocols which have a non-classical values.

Are there modern use cases for NNTP? Could UUCP be used for sneakernet or bluetooth intermittent mesh applications? What's the verdict on WebDAV, CalDAV and CardDAV as neutral protocols for sharing data, contacts & calendar?

Yes - I have used NNTP as an eventually-reliable pub/sub message bus in an unstable mesh environment. The specific use being transport of per-device usage/billing & monitoring data.

The implied Gossip algorithm of the server-to-server subprotocol was well suited to the intermittent availability of links in that particular network. Messages (articles) were PGP verified. NNTP is channelized by design, and servers like Diablo have built-in expiry management. It's also easy to monitor with a lightweight news reader.

This was only a few years ago. I was pleased with it, at the time - simple to build from readily available parts.

Thanks for that encouraging data point.

We need a neutral, replicable client-side data store with standardized interfaces for automated analysis/filtering/reading. This could be an offline archive of docs, web browser history or online/cloud bookmarks, along with _private_ annotations about content and authors (e.g. killfiles, quality ratings).

If the datastore is open (standard & source), there could be healthy competition among proprietary apps to use the datastore, but not control it. NNTP could predictively replicate (via home wifi) a subset of this cache to mobile based on personal calendar, reducing public network usage.

What's good about this approach is that it would surpass the usability of mobile apps + public cloud, thanks to the private cache reducing the impact of public network latency. With neverending hijinks on net neutrality and web standards, it may be worth exploring NNTP pub/sub between home server/laptop and mobile devices.

So basically, CouchDB + PouchdB/TouchDB/Cloudant ? I'd love to see something really emerge in this domain (CouchDB is already mature).

We are ways off, but my personal goal for http://hood.ie is to be able to allow people to build gmail/gdocs competitors without having to do any heavy lifting. We could use your help, though :)

> Are there modern use cases for NNTP?

After the death of Google Reader, I've switched to news.gwene.org, which operates a RSS to Usenet bridge. It solves a lot of the problems that other RSS readers are still dealing with, since the Usenet people figured this stuff out back in the 80s.

Great. The main page looks exactly like gmane, but all was explained by http://gwene.org/about.php.

Sample web interface to NNTP messages, http://read.gwene.org/group/gwene.com.nytimes.nyt.rss.books

I wonder how long they keep articles? Could be useful for podcasts.

> And that is why I get so worried that the draft HTTP/2 editors so steadfastly refuses to do so.

Can you elaborate on what you mean here, for those of us in the dark?

The main power for me in email is that I actually own it. I have my own domain. While I use Slack, Google Hangouts, Hipchat and a bunch of other services, none of them replace email. It is standardized and despite social media/chat services living and dying over time I have had my domain and email for 15 years now. I can't think of many other services I can say that about.

It is also a good medium for non urgent communication that paper mail used to serve. The problem people see with email is actually not a problem at all with email, it is with how it is sometimes abused. My boss sometimes sends me an email and then prods me via Slack if I did not read it in 5 minutes. That is what these chat/message services are replacing... The short term action required requests that were formerly served with a phone call.

What I find interesting about people who say email is bad is they they almost all have some vested interest in another communication method, especially something proprietary...

Email will outlive everyone commenting here because it works. I run my own server so I know the NSA don't have direct access to content (although I always take note of any inbound messages that are flagged as not having used TLS, or where the other address is gmail etc.), I can make disposable addresses, addresses specific to websites (to identify sites that sell/leak your address), I can run my own spamfiltering that doesn't invade my privacy, I can DKIM sign my messages and have a provable way that only I sent the message, I can use PGP for any private information, I have a set of filters to classify email so I don't even need to spend that much time dealing with it, and I can access it from anywhere I can get an SSH client. No service does that.

I don't understand the point about college students viewing email as stale. I'm in college and email is used more than ever. I used to get tens of emails a day from various student groups (now they get sent to spam), and it is pretty much the easiest/fastest way for groups to communicate with each other.

Sure, email isn't "sexy" anymore, but that doesn't mean people of my generation don't appreciate it.

From the spammer's perspective, e-mail is stale because you're sending everything to the spam folder rather than clicking, sharing, and buying.

Other people use it though and do click through. I just make a conscious effort to remove clutter from my life.

This brings up a point I think about a lot:

Is email our last success in popularizing an open and federated standard?

Maybe you can count OAuth, but IMO i have low confidence that we'll in the near future be able to collaborate on an open protocol so that many benefits of email such as control without vendor lock-in can be enjoyed.

We have too many entrenched interests by the main players. I have been working briefly on improving the exchange of trust/reputation data online, but it seemd for us that there was no alternative to a proprietary system if you wish to see widespread adoption.

EDIT: I guess Bitcoin has good potential.

> Is email our last success in popularizing an open and federated standard?

If you mean one with a distinct and separate S2S/federation protocol (SMTP) from the main C2S protocol(s) (POP/IMAP), probably.

OTOH, it could just be that separate S2S protocols are less favored in the first place -- the web is newer than email, open, and federated (in that web servers -- but doesn't have a separate S2S protocol. A web server that needs information from another web server to do its job uses HTTP just like any other client.

> Is email our last success in popularizing an open and federated standard?

Isn't the Web younger?

You are right, https://en.wikipedia.org/wiki/Email quotes 1982, https://en.wikipedia.org/wiki/World_Wide_Web 1989

Either way, my point still stands that besides the rapid development of the "internet", we have yet to see another standard emerge (besides OAuth and Bitcoin)

1982 is when SMTP was defined. Non-SMTP email on the ARPANET goes back to 1971.

SOAP? It was almost a good idea.

It was a great idea, badly implemented. I feel confident that we'll rediscover SOAP, or something like it - it's still pretty much the best option for RPC in typed languages, which are coming back into fashion. (Maybe Thrift is better, but it's morally the same thing IMO).

I don't think that entrenched interests are not the problem here. I think it's that a small private team can always out flank a standards committee. This difference is amplified when it's not clear exactly what needs to be built and a lot of iterative design is needed.

Email was one's passport and identity. Before Facebook became a true alternative for verifying one's identity on the web, the email address was how one accomplished serious things on the Internet. Want to verify a bank account? Email. Amazon? Email. Forums? Email. Even Facebook in the early days? Email.

Looking at Facebook's sign up page right now, and it seems that email is still required for registering a new account.

The thing is, almost every Internet service still requires an email address to sign up, and that ranges from mobile games to ecommerce shops. Some services provide the alternatives of allowing users to sign up via Facebook/Twitter/Google+; but in order for the users to get a Facebook/Twitter/Google+ account, they'll still need to sign up using an email address. Besides, almost all services that allow social network sign-in gives their users the option to sign up with email as well.

The services that do not allow email signups are few and far between -- like Medium.com, for example, but as said before, in order to get a Facebook or Twitter account, the user would still need an email address. Even mobile only apps like Whatsapp still appear to require an email address to sign up for their online support site.

> Email was one's passport and identity.

Email is not an identity. Most people have multiple email addresses, and some people share some addresses.

A person can have multiple identities, and not all identities are associated with specific people.

Email is not a person's unique ID, although sometimes it's used to serve that purpose. The ID layer of internet is not established yet. I'm not sure if it's possible in a decentralized system. Centralized management may be necessary.

Agree, but "identity" and "unique ID for a person" are two different concepts.

Last I checked Gmail allowed me to create an account without providing an email address or phone number.

When was this?

Currently, you need to provide a code that is texted to a mobile phone for some level of proof of identity.

I have registered an account last week or so without any codes beyond captcha. It surprised me, too, because for quite a while you needed a mobile.

Maybe the requirement is based on some heuristics? IP address, country, other info provided by the user, and so on?

Gmail IS email.

A lot of services today uses the cellphone number as a more unique identifier.

Checkout Stripe.

and this, gentlemen, is why you should create actual protocols, not applications.

applications die, protocols stay.

if your software solution use the web protocol, you're already limited by it. that's why I hate 99% of the internet techs.

tell that to app.net

I got a free account and spent an evening or two trying to figure out what app.net was supposed to be, and I gave up. I get the feeling it was supposed to be something more than the ghost town Twitter-clone it appears to be, but I couldn't figure it out. Email's purpose, on the other hand, is pretty obvious.

I still use pine :)

I use gmail so that I get good operation cross device but it's heavily filtered so I only see a fraction of total email on my phone, but I can search everything.

I then pop everything off using fetchmail and process all emails down to zero once or twice a day using pine (either in Terminal or irssi connectbot on my xperia).

This suits me not only from a day to day perspective but also because if gmail locks me out for some reason I can easily route around it and still have a full

backup of my email history.

Thanks very much for this post. I'm doing a full pop3 download of around 3 years of mostly email newsletters from my gmail account now.

In order to enable pop3 though, I had to change a default gmail security setting. Have you had any issues at all since leaving pop enabled?

Using Kmail, I'm having to fetch each month's worth of email at a time. Can't find a setting anywhere so assume that is a Google thing.

Want to verify a bank account? Email. Amazon? Email. Forums? Email. Even Facebook in the early days? Email.

This is a big part of why email continues to thrive. So many services have email baked in (e.g. a new WordPress install sends you an email). There are some services that let you choose between email and SMS, like plane reservations and banking alerts, but 95% of the time, any notification will come through email.

Given that, there is no way you could eliminate email without cutting off all those services in the process. Any new protocol to replace email would have to be a drop-in replacement for anything that currently sends out email or at least coexist peacefully alongside it.

You would think that if anyone could accomplish that it would be Facebook or Twitter, but I haven't seen any integration like that so far (e.g. get your plane reservation update or Amazon shipping confirmation by Twitter DM or Facebook message).

What if someone designed a new email protocol/etc from scratch without any reference to the existing one. Could it be made better? Can you even define what better is?

I seriously think that what we have is great. Send this text from here to there. It's so simple that almost anything can be done with it. If only social networking had followed the model of a simple protocol with distributed servers, we would have a serious alternative to Twitter's API restrictions and Facebook's privacy invasions.

That said, if we could redesign email, my quick wishlist would be:

* Better encryption. No surprises that a HN comment thinks we should encrypt metadata as well as contents, and have something better than the way we have to exchange GPG keys at the moment.

* Better security. Easier to guarantee that the ostensible sender of a message really did send it. Also, help prevent the viral problem of malware, which spreads through inboxes protected by rubbish passwords.

Actually, that's exactly what we are trying to do with AirDispatch [1].

We started out intending to "fix" email with a better protocol, and ended up creating something much greater.

The core problems that we tried to solve (to help define "better") were:

- Security: encrypting and authenticating your messages

- Control: the sender hosts the message, so they can edit or delete them before the receiver sees them

- Flexibility: all messages are, in fact, just key-value stores, so any type of data can be sent on the protocol (whether it's Mail, IMs, Vines, Instagrams, or whatever the flavor of the week social network is)

I think it's a great product, and we are going to start the developer release of the client (Melange [2]) tomorrow.

[1] http://airdispat.ch

[2] http://github.com/melange-app/melange

Much as it pains me to mention DJB; that sounds very similar to the Internet Mail 2000 model. http://cr.yp.to/im2000.html

Good catch! We actually completely developed AirDispatch without realizing that it had many of the same tenants as IM2000.

I haven't seen any IM2000 implementations, so I'd love to check it out if you know of any.

Do you not like IM2000, or have other feelings towards it? I'd love to hear your feedback.

Any reticence to mention it is likely due to the fact that djb is a bit of a polarizing personality.

I figured it had to do with something like that. However, do you or inopinatus have any thoughts on the system as a whole?

I think it is an interesting idea that has yet to gain traction, let alone interoperable implementations, possibly due to the cantankerous origins. May have difficulty penetrating the enterprise, where control of policy trumps most other considerations. And as with any multi-sided business model, it requires a community segment with unmet needs to provide initial critical mass. I suggest trying to make the protocol less binary and more like a classical line-oriented exchange, it'll improve uptake amongst implementers. And I wish you well; every effort to reinvent email deserves a red hot go.

There are certainly things I would fix. Require all servers to be 8-bit clean, and default text to UTF-8. Sort out the rendering of HTML email, either using something that's more amenable to embedding in a fixed screen frame, or some more extensible standard to allow different types of document to be attached with dispositional settings. (In fact this is something I'd want to solve in general - embedding objects in webpages has long been a mess, to the point where we've resorted to specific <VIDEO> and <AUDIO> tags rather than trying to solve "embed another file in this web page" in a general way).

As another poster said, crypto everywhere and a permission system, from DNSSEC on down. Every assertion would carry a cryptographic chain of its authority.

Some kind of multi-user solution would be great. Mailing lists are a hack. To some extent NNTP filled this role though - maybe it's just a case of using the right protocol for the right job.

So yeah, it could definitely be made better - very few protocols are perfect. But it's done pretty well considering.

> In fact this is something I'd want to solve in general - embedding objects in webpages has long been a mess, to the point where we've resorted to specific <VIDEO> and <AUDIO> tags rather than trying to solve "embed another file in this web page" in a general way

Like <embed> and <object> tags in HTML?

Yeah. I think the introduction of <video> and <audio> make it pretty clear that those have failed.

I don't think it could, or should, be created by a single entity or brand. All of the foundational protocols were done by the IETF or the W3C.

Of course. Lots of ideas in this direction. But be warned, thinking about new protocols is the first place engineers like us go without thinking about what getting adoption would really look like (which would be near impossible + need a backwards compatible beachhead imo).

That's what Google Wave was intended to be, among other initiatives.

There are some cases when encouraging shorter messages is very beneficial.

I can never imagine how would a university/its various student organizations get their messages through to students, without email. I went on a full-year exchange to Chile, where the email service simply is useless and everybody(including university officials) seems to rely on Facebook, which to me was simply crazy. Why would something like FB be used for any serious business? How would they suppose that everybody has a FB account and likes to use it? That was a truly horrible experience, especially for a Social Network avoider like me.

People who think that email can die are the people who don't understand technology. Email is simply sending text from one user to another. There is nothing more simple than that. Therefore it probably won't die as long as we use text interfaces to our machines. What cute interfaces you put on top or machine learning features, that's all up to the marketing department. But all cool social networks and chat Apps can't do better than simply sending text.

Email will always be around, but I must say that "Email killers" are really going to be successful.

For example, I don't remember the last time my team sent an internal email that wasn't a forward from a client. We use Slack. Exclusively. We organize projects around it, sales efforts, everything. It has the async nature of email, the separation of topics like email, and the search power of email. It also means that none of us ever feel like we're out of the loop.

Unfortunately, the "nobody owns" feature of EMail is something Slack lacks and probably always will. Email is a federated, decentralized protocol of distributed mail servers that anybody can set up. Any email server can participate with any other email server. That's something that Slack will never have (it does interface with XMPP and IRC, but the service itself is centralized.)

It is hard to overstate how critical this fact is to Email's future. If you look at most of the canonical Internet protocols (SMTP, HTTP, DNS, XMPP, IRC, even BitCoin) - they have this fundamental feature in common. Even though they're part of the application layer of the Internet, they've become fundamental protocols that other technologies rely on. They are part of the infrastructure of the Internet.

I say this is someone who just started using Slack and loves it. I still can't see it replacing email, though.


I actually think these guys are interesting exceptions. Yes, they are still around, but they've largely been eclipsed by proprietary solutions. Why? Why did Twitter storm to success XMPP was ostensibly the same thing but with all of these good qualities (open, federated, decentralised, etc.). Still to this day messaging is a battleground for new entrants touting completely opaque solutions. It seems like there's something more to success than just being open, federated and decentralised.

XMPP is one of those things that billions of people use without knowing it. Facebook Chat and GChat for a long time (and maybe still now) were actually just branded XMPP solutions - Facebook (annoyingly) shut off federation, keeping in the spirit of their walled garden approach, but GChat still interoperates with other XMPP servers if I'm not mistaken.

Edit: GTalk shut off federation for XMPP extensions like multi-user chat and Jingle (Voice-Over-IP).

As for IRC - yes it has been largely eclipsed by proprietary solutions. I chalk this up to the lack of a compelling, truly simple web front-end for it. For instance, Slack seems to imitate a lot of the feature of IRC - it would be easy enough for it to just be IRC underneath, and maybe it is. But as far as I know nobody ever wrote a true web client for IRC that was compelling enough for the mainstream.

Internally it's not xmpp, but instead an xmpp facing compatibility layer.

Interestingly WhatsApp is also a XMPP-based system, although totally closed and likely incompatible with other clients.

XMPP's federation is great, but it's a real pitas to scale up a single domain (long-lived connections, implementations typically assuming one node, dialbacks). The gtalk team had a really hard time with it. Http is a lot easier to scale.

Twitter originally had a chat interface. It's where most people I knew, used to use it - via pidgin. I think it was an AIM bot.

> I must say that "Email killers" are really going to be successful.

I remember people saying that in about 1998. Email killers have been successful, but not at killing email.

How disrupted will your company be if Slack is purchased by Yahoo and is EOL'd with 30 days notice?

To do that to email, you'd have to dismantle DNS.

> We use Slack.

I'm saddened and proud that every time I read that sentence, I initially think the person is using Slackware.

I'm pretty firm in my belief that one of the great marketing moves of the 21st century is convincing people that email is not a social network. It's almost on par with making people believe that diamonds are romantic and a necessary part of the marriage ritual.

I think this has been helped by the general lack of innovation in the email space. From pretty basic mail, we ended up with a few (very surprisingly few) email clients and very little advancement on the original theme outside of html formatting and huge inboxes.

Lots of people dump on Microsoft, but one of the huge upsides of exchange is the tight integration of mail and calendar. From a conversation you can immediately schedule actions. Invites are even sent out over SMTP if I'm not mistaken. Getting a calendar to integrate well with gmail was one of the major accomplishments of web-based email, yet it seems like repeating this anywhere else is an accomplishment comparable to discovering cold fusion.

There's also been pitifully little work done in improving the experience of managing email and calendar servers. Managing spam is still a tremendous problem and all this adds up to most places, if they aren't using Exchange, just buying corporate Outlook.com or gmail accounts for their employees.

The problem of course is that for any serious advancement to really work, everybody (both client and server) have to move to support the advancement.

But one lesson to be learned from Facebook and G+ is that email can be replaced by an easier to use and friendlier system. There's a possibility of disruption, but it's obviously not in anybody's particular interest to keep reinventing email+otherstuff in this kind of highly centralized way. If Facebook goes down, there goes a huge chunk of the global communication system. At least with email I can be pretty sure my message is going to arrive at the destination at some point.

Another lesson to be learned is that social networks like Facebook are actually just a combination and integration of two (or three) common things that used to be all over the web: a personal website and email. You get a profile (which does a good enough approximation of the personal homepages of the web 1.0 days but actually a bit more like ) and people can message you (and more recently IM you). Basically a global presence you don't have to put much effort into to manage and a way to contact you. More importantly Facebook offers you various levels of control over who can see your presence and who can message you. Spam is almost unknown in Facebook's version of email.

So when I see distributed social network efforts like Diaspora, and all this talk of authentication and protocols and whatnot I wonder why we're not really using and extending the distributed infrastructure we already have. Even if we improve it in some way that makes it no longer work with the old email network, it won't be the first time a better internet service replaced a previous one (WWW replaced gopher for example) -- there's no reason two competing distributed messaging services can't run in parallel.

I'm convinced that Facebook took over messaging because it has one killer feature the 90s didn't: real names.

Contact points on Facebook are discoverable by real name based on the social graph, which is kind of a first in the history of communication. If someone mentions a John Doe in conversation, I can get in contact with the correct John Doe with a very high probability of success and no effort.

The PSTN, email, AIM, ICQ, MSN, etc. didn't offer that. They could have, but their communities developed with different norms. Facebook managed to get people to use their actual identities. That's a remarkable feat - not even Google could replicate it. But it's what made Facebook so useful and so addictive - it's about the actual lives of people I actually know.

Facebook does contain an analogue of "Web 1.0" personal webpages, but the action on Facebook is centered on the News Feed, which was sort of a new class of thing. Few people actually have profile information filled out; going to someone's Facebook page is just a way of filtering the News Feed to only content related to that person.

The real "Web 1.0" equivalent to Facebook would have been an RSS reader for all your friends' blogs. But for people to be comfortable blogging, they had to be able to (feel like they were) in control of their audience. Though Blogger, Wordpress, etc. supported user authentication, it would have been incredibly onerous compared to centralized identity. Which is more or less what Facebook became.

A truly open, distributed Facebook based on the "Web 1.0" would probably have looked something like an RSS reader that could authenticate to each friend's blog with OpenID. But it still would have lacked the ability to discover and search for people by real name that makes Facebook so useful.

Nah. I had the same conversations on Facebook, with the same people, and initially under much the same pseudonyms, as I had on MSN 5 years previously. Excessive insistence on real names is what killed google+.

Facebook succeeded, initially, by being livejournal but not letting people customize how their profiles looked. (News Feed really wasn't big in the early days). Everyone's page looked the same, and it was a good look; you immediately knew where to see interests, where to see favourite movies and so on. It also had better-integrated photo sharing functionality than any of its rivals.

And it was a much better interface for organizing parties than anything else: a big chain of email CCs gets unwieldy fast (and is irritating to anyone who can't make the party but still sees all the messages). AIM/ICQ/MSN/* never really handled groupchat properly. Jabber screws up if you want to sign in from more than one place. IRC gets the groupchat part right, but relies on everyone leaving their computer on the whole time, or doing something even nerdier like setting up a bouncer. Skype would eventually solve the problem, but it wasn't in the picture when Facebook started out.

Yeah, absolutely. The centralized index of people is as critical to Facebook's growth as DNS is to the Web. You don't have to remember to get to everybody's favorite tech discussion forum, you just have to remember news.ycombinator.com.

So I wonder how hard it would be to setup a DNS-like searchable index for an e-mail like replacement? Instead of having to remember guy@foo.com I just have to remember 'John Doe' he works for Foo Inc. and I can probably narrow it down from a list of other John Does. If not, I can always call him and ask him for his explicit address as a fallback.

> but the action on Facebook is centered on the News Feed

It really seems that, much more than Twitter has ever managed to be. People's News Feeds are really their microblogs. For many of my FB friends, I can go to their feed and read back a few months to see what's going on in their lives. Just as if they had been keeping a blog. But the convention on FB seems to be short posts, so I don't have to read long-form posts, but they also aren't restricted to a tiny string of characters.

I actually think the news feed reminds me quite a bit of the finger protocol and .plan files. (there's commenting and likes and such of course also).

So what core distributed communication and web presence technologies already exist that could approximate Facebook and what and how would they have to be modified to work at least as well as Facebook? What new things would have to be created from wholecloth?

Would it be possible to integrate all this mess tightly enough that a person could set up their own "Social 2.0" node without too much fuss?

>So I wonder how hard it would be to setup a DNS-like searchable index for an e-mail like replacement? Instead of having to remember guy@foo.com I just have to remember 'John Doe' he works for Foo Inc.

But Facebook's power is the social graph. I don't have to remember "he works for Foo Inc." and I can always find him even if he's long since moved on from Foo Inc. Rather, he ranks highly in the search results for "John Doe" because we have friends in common.'

What you're describing might be better for, e.g. professional contacts you meet at conferences (which seems like LinkedIn's territory) but for organic friend-groups based on high school, college, location, etc. then mutual friend count (or degrees of separation, if no mutual friends) is a damn good indicator of relevance.

Is there a way of doing a distributed social graph with some degree of privacy and no central authority?

And yet Snapchat took over messaging by reverting back to arbitrary usernames, AOL styles.

Snapchat bootstraps its social graph on top of the already-existing social graph of "people who have each other's cell phone numbers." Actually learning and typing in people's Snapchat usernames is possible but not common.

Exchanging phone numbers is already integrated into the social fabric of our society as a step in escalating a friendship or asking for / receiving the opportunity to date someone. It's a somewhat intimate act, and appropriate for a communication mechanism that's basically only used between people who are already friends.

Snapchat doesn't have the ability to discover acquaintances / friends-of-friends nearly as easily as Facebook does.

I also wouldn't say it "took over messaging." People don't really have actual conversations on Snapchat, they send photos of themselves making weird faces or Twitter-style status updates. New class of thing.

There's nothing that prevents a non-Microsoft calendaring system from supplanting it, other than on agreeing on the protocol.

Email has MIME, there are calendar information formats (ical, IIRC), and various MIME handlers. The primary block was that Microsoft owned the corporate desktop, and competitors, until Google came along, couldn't agree on interoperable standards.

Facebook and Google win by being Web based and "in the cloud". Facebook had its social graph, Google had the fact that many users were relying on it for email, and increasingly, companies are.

There are any number of other problems surrounding mail, with privacy, spam, true federation (spam means residential/consumer IP space is virtually always blocked), and access from multiple devices being prime among them. Diaspora is interesting (I'm on it and active), but very small and growing slowly if at all. I do very much hope that some federated system (Diaspora, Friendica, FreedomBox, Sandstorm.io) will emerge, but it's a long and slow process.

> There's nothing that prevents a non-Microsoft calendaring system from supplanting it, other than on agreeing on the protocol.

Google and Apple both have integrated calendar and email systems. They just use separate apps instead of having everything in one silo.

Right. And there's absolutely nothing wrong with that.

Ironically, one of the biggest successes I've seen with Google Apps is a site which had formed from multiple company mergers, with separate MS AD/Exchange domains, which 1) couldn't be merged and 2) couldn't exchange calendaring information.

The solution (this a few years back) was to go to Gmail and Google Calendaring. Among the most successful and enthusiastically adopted mail/PIM migrations I've seen ever.

I think the main thing keeping email as the baseline for communication on the internet is it's cost. There has been no other service that can provide such a varied yet simple medium of communication - text, images, video attachments with history and an audit-able trail - for free.

The closest that anyone has come have been the big social providers with messaging applications which mimic email in many respects. Even then, they are copying the email model with a branded version - not replacing it.

It will take an entirely new and different protocol that simplifies communication with the same or better capability scope to displace email.

Slack is obviously not killing emails as they are advertising through lot of PR. They have 150K active users so far. However one thing that they are correctly going after is fantastic search abilities for our own data that would include all emails, chat, attachments etc. My feeling is that eventually communication client that excels in search would indeed surface to the top. This client would then drive communication standards of the future, including emails. So in essence, problem is not integration of desperate sources of communication, but ability to search them efficiently.

Webmail was the first and continues to be the most successful implementation of cloud. Ever.

I don't see cloud designs of now working on top of standards like email does. No wonder data tends to get stuck in silos these days.

Email and WWW provide everything proprietary services are providing people now, but with less flash. The problem is that so many people are moving to these proprietary services that it becomes very difficult to organize and communicate people without making that move. I needed to coordinate with multiple people, and one of them emailed me that we should start a group chat on Facebook; why not just cc them?

I tend to disagree. To me the WWW is the best thing on the internet.

But I do use the heck out of email ever since I quit all social networks a couple of years ago.

I somewhat disagree with the background data this article cites as reasons for why everyone wants email dead. I do not believe people like Slack, Asana, et. al. are trying to kill email, rather, I believe they are trying to offer services that help to reduce email's dominance over communications.

One thing that email does not provide for anyone is choice. The barrier to entry is running your own IMAP, SMTP, and possibly spam blocker servers, which almost no one wants to or even knows how to do anymore. It's also somewhat inefficient for conversations, just like regular snail mail is inefficient for conversations. You mail letters, not one-line replies.

So I tend to think of IMs, chat rooms, and other methods of communication on the Internet as simply compliments to email, not replacements for it. Nothing will replace it, because nothing needs to. It's email.

(sidenote: I'm pretty sure one big reason everyone uses email is because they were given an email account, and didn't have to search out the technology. You don't see people clamoring about the reduced use of Usenet in comparison to forums nowadays. Same can be said for texting, I personally got a phone one day and someone sent me a text message, that's how I started texting. Probably never would have thought getting text messages sent to a phone that I can just talk on would have been a useful enough communications platform to seek out and possibly pay for.)

>You mail letters, not one-line replies.

What stops you to send a one-line reply in email?

I'll rise to the bait.

http://incubator.apache.org/wave/ (first from Google, now Apache) was/is a brave try. The main problem with email is the lack of consistent formatting rules which means that it's difficult to keep track of structure.

But hey, it's a massive success. Worse sometimes really is better.

Email is still inherently a conversation with a person or people, and that is why it has value.

I agree. I think there is more that can be done to enhance it, and make it a more enjoyable tool. Email clients --and I am not talking about third party services a la Mailbox, who rely yet in another server-- are still very primitive.

I feel that gmail has held back development of good desktop clients (which is what we really need). My laptop has an almost empty 1TB hard drive yet I need to wait to download an image attachment every time I want to view it. Email should be built around abundant storage, not a dependency on bandwidth.

> Email should be built around abundant storage, not a dependency on bandwidth.

I'm sorry but this doesn't make sense to me. To cache all images you've ever received is sort of antithetical to the whole 'decentralized' nature of email, and the internet in a nutshell. The 'inconvenience' of having to re-download far surpasses the need to store every single trivial little thing on your mail client's cache. A few seconds is worth way more to me than taking up tens of gigs of images on my HD that I will never see again.

I disagree. A modern device is almost entirely composed of files downloaded from the internet. The binary for your web browser is just a local cache of an entity that exists in the cloud. You need a local cache because bandwidth and latency are too high to retrieve a new copy every single time you run the program unless you need an update. It seems a pity to not use storage to cache when it is available. Especially considering that to most people bandwidth is much more expensive than storage.

But your email is never decentralized. The email you've received stays on a central server, unless you delete it.

I feel the same way. In my opionion, Outlook peaked with 2010 but they are missing new ideas. There is SO much you could do if you would think just a little bit out of the box.

One could think of contextual information added to emails, like detecting invoice numbers, intellligent attachments (e.g. recognizing Dropbox links etc.) and so on. I really hope there will be some client like this in the future, if not, we'll build it :)

Well, for what it's worth, I agree with you.

I think this goes beyond email.

Similar to the way consumer applications gratuitously became big and slow as processor speed increased and magnetic drive space increased, today we see gratuitous use of increased network bandwidth.

As such if there is even a momentary problem with the remote network, then there will likely be a problem with your application. Many of today's applications have no "offline" mode. Network availability and performance is assumed. Just as back in the 90's it became assumed that the user would have a capable PC. A "slow" PC would not run the latest software.

Many of the applications I write for my own use require very little computing power... I develop on a modestly powered machine and so my programs can run fast from old hardware if need be (i.e., can't afford the latest gear). My approach to "apps" is to periodically download bulk data from the network to keep the application updated, and this usually is scheduled for off-peak times when costs and congestion are lower. Fast network speeds and inexpensive local storage are much appreciated when performing these updates. Not to mention the ease of obtaining an internet globally reachable IP address through today's "cloud hosting" providers.

But I will never understand today's "app" and "web API" trends. Developers of these "apps" utilize the costs of storage and bandwidth a different way. Storage is mainly remote (on someone else's machines, not yours), and hence controlled by some third party, not the user. And most of these applications I see are designed to make constant piecemeal downloads, minute by minute, hour after hour, day after day (even if the same information has been requested before) when they could just as well download all the data they will ever need once, and periodically update it during off-peak periods. As such, if the network is congested or down at runtime, the user must suffer it out. Of course, by making a call to the remote network every time the user takes some action, third parties are able to track everything a user does. That, and the whole app/api approach to development has value to some people perhaps, but surely it is not users. For example, when I watch YouTube videos, I have an application that downloads them first, storing them on RAM disk or inexpensive local media storage and then serving them over a local network. I can then play them using whatever device suits me (or store them on any other device if I wish). I never see any ads. I never use Javascript or Flash. And I never have playback issues (because among other things I have control over my local network). This approach seems like common sense but I feel like I am probably one of the only people out there who does this... and it is not difficult at all to do.

Indeed! One think --among others-- that Gmail has done very well is search, which is expected, as Google specializes on that.

Why are you downloading attachments that are images? Gmail will pre-load images for you if you let it. (http://gmailblog.blogspot.com/2013/12/images-now-showing.htm...)

Also you could open the attachment right inside Google Drive

Quite happy with Thunderbird.

I gave KMail a try and so far I've been pleased. There are lots of aspects Thunderbird is ahead, though.

In what ways are the current batch of email clients primitive, and how do you think they can move forward?

An interesting post on this subject by Aaron Seigo:


This article is good but it makes some comments that don't seem true to me.

> It's a globally addressable system that does not require any centralized system, which means it works everywhere and for everyone no matter what choice in service provider the individual user (or their organization) makes.

Isn't it the case that nearly all email is routed using DNS? Certainly all of mine is. While there are certainly arguments against centralized addressing systems, the advantage that email has isn't that its addressing system is uncentralized: it's that its addressing system is more libre than proprietary systems.

>These "open social networks" don't pass messages to the end user. They pass messages to servers which the user then consults.

How is this different from email? Most people don't have email addresses at a domain assigned to their normal computer; the vast majority of email is accessed over some internet protocol from a remote server, whether its IMAP, POP3, or HTTP. I've never used Diaspora, so maybe I just don't understand what functionality its missing.

There are some designs for decentralization of the DNS. The decentralized nature of e-mail on the higher level still is valuable.

> the advantage that email has isn't that its addressing system is uncentralized: it's that its addressing system is more libre than proprietary systems.

I think that the biggest advantages of e-mail are its decentralization + federation based on open protocols. These make e-mail extremely useful despite all of its shortcomings. Consider for example IM services. They multiply like mushrooms these days, yet most of them completely lack federation and therefore are crippled if you look at them globally. You can't communicate with users of other services at all (unless you register there as well which defeats the purpose of decentralization). XMPP attempted to solve it, but unlike e-mail most participants on the IM scene were too selfish to fix the situation, including Google who deserted.

> How is this different from email? Most people don't have email addresses at a domain assigned to their normal computer; the vast majority of email is accessed over some internet protocol from a remote server, whether its IMAP, POP3, or HTTP. I've never used Diaspora, so maybe I just don't understand what functionality its missing.

I think his main point is, that in e-mail you can have real push mechanisms in theory. I.e. you send a message and the other participant receives it through the e-mail service. In social networks (even decentralized ones), the delivery of such notifications is delegated to... e-mail. Which shows their intrinsic deficiency. They aren't implementing the full scope of the communication process. I.e. if you envision a better social network, it should be decentralized but also take care of the delivery of the interactions in one coherent experience in order not to rely on other external tools for that.

E-mail isn't fully decentralized most of the time, but it still can be (you can run your own e-mail server). You can run your own Diaspora pod the same way, but like above, it would still delegate notifications to another service.

I agree that federation is critical to email's value, but isn't that true of Diaspora as well? Again, I've never used Diaspora, but why is it notifying you through email? Everyone checks their Facebook without needing email notifications.

It's like a phone number or address.

Nice article! We totally agree and don't believe in those "we need to replace email with an app" approaches. For example, that's why we build our service StandupMail (http://standupmail.com) completly upon emails.

It would probably only take one innovation in email, something that Gmail could probably pull off, to make email lovable again.

I really think that if Google could help alleviate people's concerns with email (volume, spam etc), people would be quite happy with email again :)

Maybe add a little bit of useful cryptography even. Spam is already fixed by Google, and volume? How could Google fix that?

what if we limit email to 140 characters and make it the ultimate decentralized push notification system?

Elevator pitch: "It's like email but for messages"

Email is important, WWW is all

Bitcoin is better IMO.

But I agree that email is a great thing. I hope someone creates a sick email client that turns back the flow from Facebook to the email. I'm fine with Thunderbird + Enigmail, but I don't see the average person using it any time soon (or ever if unchanged).

That is, except for kittens.

Has anyone got an invite code for the new Email product Sortd.com?

I got an invite code from a coworker as I couldn't wait any longer to get in!

I've been using it for two weeks now and I have to say it's a must have app for all Gmail users ... It's has completely changed the way I work.

Facebook Messenger will eclipse email. Social network is the perfect spam filter.

Spoken like someone who hasn't been invited to play FarmVille 1,000 times. And yes, you can disable these notifications, but doing so isn't any easier or automatic than training an email spam filter.

Doesn't solve the "This order is wrong I need to send email to support@somerandomcompany.org" problem though.

I'm actually afraid that the internet is moving to this model, specifically with Facebook. How many times have you seen an ad that, instead of a URL, asks you to go to their Facebook page?

> I'm actually afraid that the internet is moving to this model, specifically with Facebook. How many times have you seen an ad that, instead of a URL, asks you to go to their Facebook page?

No more often than I used to see the same thing with an AOL Keyword instead of a URL. I don't think the internet is "moving to this model" so much as it is a model which periodically emerges when a particular portal is dominant enough with the commercially-relevant audience, and fades as the dominance of that single portal.

But it is centralized and closed. It will not eclipse email for anything serious and long-term.

Would you like to play Farmville?

Like AIM, Trillian, IRC did?

Wrong use case...

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact