Let's dig into how open source could KO the Silicon Valley chat silos

There's never been a better opportunity for the world to start untangling itself from the giant Silicon Valley data harvesters than now. Last week, we revealed a plan to embed open-source chat into three quarters of the world's IMAP servers. And this may be an important development. Maybe. Google, Yahoo!, Apple and Microsoft …

[Anonymous Coward]

Quick - call a sociologist

Speaking as one who runs dovecot and prosody for imap and xmpp respectively, some kind of combination seems a good idea, providing OTR etc can be part of the mix. Thunderbird's chat, for example, doesn't allow OTR. The way we use imap and xmpp is exactly as described in the article - at times, chat-like activity takes place over email. Chat (pidgin client) used as a file transfer mechanism also takes issues away from mail.

But I can't help thinking that the real research should be into why people herd themselves towards closed ecosystems. That's the non-technical issue that seems to need solving.

[philthane]

Re: Quick - call a sociologist

Advertising and herd mentality. Once a few people are convinced by the advertising they 'invite' their contacts to join them. OpenSource companies tend to have small advertising budgets.

[John Brown (no body)]

Re: Quick - call a sociologist

"Advertising and herd mentality. Once a few people are convinced by the advertising they 'invite' their contacts to join them. "

Yeah, remember when GMail was launched? Only a select few were allowed to sign up, but they got a certain number of "invites" to give out. And then those invited people got more invites to give out. Lots of publicity, limited and gradual supply. We saw people quite literally begging to be sent GMail invites on Usenet newsgroups, in forums and message boards etc. And for what? An email address, at least one of which they already had anyway.

[Andraž 'ruskie' Levstik]

Re: Quick - call a sociologist

Ease of use. That's the simple one - As someone who has been running dovecot and prosody for a while - I more or less gave up on xmpp - it's very clunky it doesn't do stuff as easily as it can - certain things like message history is lackluster etc...

Since I'm a gamer I've started using discord and I can put it in simple terms:

It's easy to setup, it has oth group chats and servers - basically it's an improved IRC.

Need to share a file with a person or group - attach it to the message. It's there.

The feel of it is not that far from IRC as well. Of course if you prefer you can go with a more "chat" like feel as well. It gives you that.

There is no "setup an account on some xmpp server" - it's simple - create account, use existing email, start using it.

The user experience is what counts now-a-days. This is the main reason people go to something. They try it out and start using it. Then invite friends.

An example from a friend that only recently started using discord - "This is what I like about WhatsApp - I can tell when my friend saw the message."

Honestly even users themselves often don't know what that killer feature is that they want.

[DropBear]

Re: Quick - call a sociologist

Interesting, considering for me the killer feature (on other channels like SMS or email) is absolutely nobody being able to tell whether I saw a message or not - disabling "message was opened" replies even if requested is the first thing I do in any email client. Plausible deniability, people. YES, I INSIST.

[Andraž 'ruskie' Levstik]

Re: Quick - call a sociologist

I personally dislike that feature as well.

Other thing that annoys me is the mutability of the messages themselves. This is what I like about email and IRC - you post it - it's there - can't delete it or change it.

No write something - look at it 5 minutes later - it changed to something different on gone completely.

Why I have bitlbee-discord on as well - that shows me the full trail.

[Whitter]

Re: Quick - call a sociologist

In forum software of yesteryear, posts are often editable for a short while so typos can be corrected. There is usually a flag that the post has been edited and sometimes a history. Other platforms allow editing at any time until the next post is made (at which point all previous posts in that thread become non-user-editable), or a combination of the two. Or edit whenever and so what - its just a forum board.

It seems odd that FB didn't take any notice of who people like to edit/maintain their posts on a public(ish) thread-based conversation model given all the prior history. Probably just because they didn't need to, so less work.

[DF118]

Re: Quick - call a sociologist

"Interesting, considering for me the killer feature (on other channels like SMS or email) is absolutely nobody being able to tell whether I saw a message or not - disabling "message was opened" replies even if requested is the first thing I do in any email client. Plausible deniability, people. YES, I INSIST."

As always, Mr. Munroe has a strip for that...

https://xkcd.com/1886/

[Anonymous Coward]

Re: Quick - call a sociologist

Wrong kind of social scientist. To determine what/why/how/etc people do things one particular way over another dozen different ways, you need an anthropologist. Sociologists do organizational or societal level analysis. I bless going out into the field with my mother, an anthropologist, and seeing how you collect such information and use it to create an ethnography for later use in creating processes and tools people will actually use rather than misuse and/or destroy. Saved me tons of grief. I slept better at nights as well.

[Warm Braw]

Why people herd themselves towards closed ecosystems

The main reason is that open ecosystems cost money to run and in turn charge their customers. I pay actual money for my IMAP email service. Closed ecosystems can give their services away in exchange for monetising their users - but only because the closed system prevents the users from leaving.

The other problem with open ecosystems is that they can be buggy and unreliable. IMAP was a real mess for much of its early life and finding compatible combinations of client and server implementations was not easy. Much easier to provide a reliable service if you control it end to end.

Of course, open systems are more likely to persist (sometimes well past their sell-by date: IPv4, SMTP...), but a lot of these closed ecosystems are short-term fads so their longevity isn't an issue.

This may not be desirable in terms of the overall good of mankind, but it's unanswerable in the marketplace.

[Aitor 1]

Re: Why people herd themselves towards closed ecosystems

well, yeah but I want ease of use.. and slack is SO much better than IMAP, etc..

[Doctor Syntax]

Re: Why people herd themselves towards closed ecosystems

"a lot of these closed ecosystems are short-term fads so their longevity isn't an issue."

It depends what's in them. If it's every idle thought that wafted through idle heads then no problem. If it's somebody's only collection of photographs of their deceased parents (yes, I know) then it's a bigger problem. If it's a long piece of collective research then it's real pissing-off time.

Anyone remember geocities?

There may be merit in paying for what you get.

[P. Lee]

>open ecosystems cost money to run and in turn charge their customers.

There's that, which is probably the main reason.

There is also the reason that the open source crowd try to do things properly, usually using a new TCP port which then fails by default at every firewall. We build an open protocol with a proprietary port.

Closed providers build proprietary protocols over an open port.

IMAP is going to fail over corporate firewalls where facebook messenger succeeds.

The one way this could succeed is through phone use. These days, phones bypass the corporate firewalls so we could use any protocol we fancy. What we do need to do is to enable presence information without enabling spam.

Personally, I think we need a handshake protocol for swapping certificates over email. We need to hold addressbooks in email, so email servers can run the handshakes without needing the clients to be online. Once we've identified "friends" (essentially we all have to run our own CAs) we can communicate presence info without spam problems. If someone's email get's hacked, we revoke the cert and issue a new one.

[Doctor Syntax]

Re: >open ecosystems cost money to run and in turn charge their customers.

"Closed providers build proprietary protocols over an open port....

These days, phones bypass the corporate firewalls so we could use any protocol we fancy."

And so we do things the convenient way. Once again convenience tramples security. More and more stuff gets done in the browser - more and more ways to deliver malware.

[Anonymous Coward]

I'm not paying attention ..

.. until I hear from Whispersys on this.

Open is one thing, secure is quite another.

[Daedalus]

Geeks don't get it

The drones like their shiny baubles. They don't care about open source v. proprietary.

They want cool. Or is it dope these days? What-ev-errrrr.

[cantankerous swineherd]

email is pure poison, so I'll pass on this ta.

[Anonymous Coward]

WhatsApp, or something like it, was inevitable

Telcos charged for SMS per message, or offered plans of X messages per month that could not keep up with chatty millennials. Apple solved the problem for its customers with iMessage, but strangely Google chose not to bundle something similar with Android. I'll bet they're kicking themselves daily they didn't, since WhatsApp ended up filling that void for Android users.

Of course now telcos (at least in the US) compete on data, and all but the most bargain basement plans start with unlimited calls and SMS/MMS. If iMessage and WhatsApp went away tomorrow most people wouldn't care, they'd just go back to texting now that it doesn't cost them anything extra.

Sure, the apps are more secure but the average person doesn't know or care about that.

[JLV]

>If I send you an email, and you reply to me, we've created a trust relationship.

No, we haven't.

Or at least I don't see it that way to the point of wanting to chat with you, which I consider, socially, to be something else.