XMPP creator's "Locker" -- personal social media aggregation platform | Hacker News Reader

XMPP creator's "Locker" -- personal social media aggregation platform

github.com

132 pointsTerretta34 comments

I like this idea a lot. I wish the guy the best of luck.

In my opinion, somebody is going to have to do something -- probably coding but perhaps an embedded mixed app -- to start tearing down all these walled gardens companies are creating. It's hurting both people and innovation when tons of folks get lock-in on Facebook, or Twitter, or whatever will come next. Decentralizing it, making the person the hub, is not only the natural state of things, but if it works will be enormously disruptive (in a good way). If I didn't have a zillion things on my plate right now this is something I'd like to participate/help with.

"Making the person the hub" sounds more like these projects; rather than pulling in your data from numerous sources, you push.

Dave Winer's reallysimple.org: http://scripting.com/stories/2011/01/05/upcomingTheMinimalBl...

Tantek Celik's falcon: http://tantek.pbworks.com/w/page/21743425/Falcon

ThinkUp(http://thinkupapp.com/) does something similar, though with a focus on analysis. It retrieves your data from different sources (facebook, twitter) and indexes it, so you have a copy you control. Once you have a copy of your data you can present it anyway you want, see who you interact with the most, etc.

This is what I want, and I'd love to work on a project like this:

One place to store my blog, pages, messages (long and short), emails, save my IMs and SMSs, pictures, albums, ebooks, docs, files of any kind, all in one place. My whole web presence in one place instead of spread all over the web. Like my own VPS but public, or at least where I can control its levels of privacy.

or perhaps myspace can adopt this idea, well, at least they have the right domain. Perhaps Mozilla? or Ubuntu? they sure would be great hosts, and this would increase traffic to their properties a thousand fold plus other ways of monetization.

Checkout tonido (http://www.tonido.com). You can do everything you have asked for except the email.

Jeremie Miller is also working on TeleHash, a P2P protocol based on JSON and UDP: http://www.telehash.org/

I came in here because I wanted to comment that I think that this has a lot of potential, alas I'm not the first one to quip that. Given how many people already wrote that they like the idea or could imagine working on such a project I'd wager that there's huge growth ahead. It would be incredibly awesome if this project would gain lots of traction.

Mozilla Labs has something vaguely similar (albeit only for messages) called 'Raindrop' (http://mozillalabs.com/raindrop). I played around with that when it came out but didn't like it too much.

> Contexts - A context is a place where I have data about myself, such as an account on a site or service, or in some desktop app, on my phone, or even from a device.

This part doesn't feel appropriately named. The place a data is stored should be a 'Location' or a 'Destination'. A 'Context' might be a better name for where that data is used to express a bigger idea than just itself.

For example, I would look to my phone as a Destination for my contacts and these contacts showing up next to an event i'm going to with others as a Context. This feels more intuitive to me. I know I'm splitting hairs, but would love your opinions.

All that aside, really love this idea! Happy to see people becoming more aware of these issues and finding interesting solutions around them.

A context appears to be a venue of participation from which information is pulled, say, like your FB wall. Note that FB prohibits API users from actually storing much of the data accessible while a user is logged in and operating with a website through the FB API, so your Location/Destination metaphor breaks down when the app has to deal with intermittent access or ephemeral information deriving from these "contexts."

Phones are pretty much in the "apps" part of the scheme, just an access device.

We are also struggling with the word Context. The thought was "a Connector is a dumb pipe and it needs a Context in which to run" (e.g. the auth credentials, etc.). Not perfect, and possibly something that will change in the not too distant future, but it works for now.

The whole project is still very young, but we are incredibly excited about all the cool apps that will be created on top of it.

How about "smart pipe" (in contrast to a dumb pipe)?

:).

It's not really a separate pipe, its's simply the bits of information that tell the dumb pipe where/how to run. For example, the Google Docs connector knows generically how to connect to a hypothetical person's Gdocs account, but the auth credentials provide the information that is required to connect to a real person's account.

A "smart pipe" is actually a lot more like the Collections. They bring data in from sets of similar Connectors (dump pipe) and collect, de-dupe, and normalize it.

I'm excited about project. Hope I'll get rid of custom scripts that download contact data and place into repository.

Don't really understand what this is for or why it's valuable. Would anyone care to explain?

I believe it's a self-hosted mashup application for all your social networks. Correct me if I'm wrong. I think it's a pretty neat idea, although most normal people are only connected to 1 or 2 social networks. I myself connect to Facebook for friends and family, and LinkedIn for professional contacts. I don't see the point of sharing personal stuff on LinkedIn too.

-update- From the looks of it, Locker only supports Facebook right now...

I believe this is a decentralized social media platform - "Diaspora" but by someone who knows what they are doing.

XMPP? Knows what they are doing?

Does not compute.

Using document markup as a packet based message protocol is just perverse.

These type of comments are pointless. If you have a point to make then make it.

I think the protocol is actually quite nice, I have an aversion to anything XML but that's just a personal preference. Have you used it in development? It works quite well... I know it sucks for binary data but that's not what it's designed for.

edit: I made my comment before axod edited in the last line "Using document markup as a packet based message protocol is just perverse." I somewhat agree with his point so ignore my reply.

Yes I've written XMPP libs at the base level. It's mental. There's 15 hoops to jump through just to say "hello!", forms must be filled out in triplicate, several nested stanzas must be entered into. It's absolutely ridiculous at the protocol level.

I agree it's probably pointless, but I find it odd that XMPP is now apparently 'cool'. I thought it was commonly thought to be an abomination of the highest order.

deal with it, it's established at the RFC level so at this point I'd prefer XMPP to <insert proprietary protocol here>

Come on, some credit is due.. XMPP protocol is in production all over the place.

Windows is in production all over the place.

Cobol is in production all over the place.

Autocracy... okay, there was that whole Egypt thing, but you get the idea.

Agree. The phrase "from the creator of XMPP" never fails to make me cringe, for a variety of reasons.

It's a tribute to the simplicity of IRC that you can write a channel logging bot in like 20 lines of shell (https://github.com/acg/logbot).

Why do you say that the author knows what he's doing? If your statement is based his work on Jabber (now called XMPP), then I am somewhat concerned. Several design decisions make Jabber difficult to implement with no benefit to the use cases. My previous comments about Jabber design decisions are here: http://news.ycombinator.com/item?id=2071652

> Why do you say that the author knows what he's doing?

Jeremie Miller wrote a popular open source IM server and created a specification for the protocol which has been widely implemented in several different languages (Erlang, C, Python to name a few) for both client and server.

Many of the functions solved by XMPP (instant messaging, presence, federation) fit nicely into the distributed social network problem domain.

So that's why I have confidence that he knows what he's doing. Nits like "I don't like different XML namespaces" more or less miss the forest for the trees.

I agree that my previous rant was about nits, but these nits add up and act as a barrier to entry to development in the space.

You mention that XMPP is widely implemented on the server, but only a few of the implementations are actually any good. To make matters worse, some of the best server implementations are written in fringe languages (Erlang for example).

I think that XMPP would be more successful if it had been more developer friendly.

Have you looked at telehash (http://www.telehash.org), a much more recent project by the same guy?

I'm impressed by how elegant the protocol is. I hate XMPP, but telehash gives me faith.

A bit more information is available here: http://www.readwriteweb.com/archives/creator_of_instant_mess...

It looks like it is a framework for pulling together information stored by a lot of different apps and networks. You can then use that information in different ways.

Even though it just has contacts now I could imagine it getting much more advanced. Something like "Have I already visited the top links today shared by my friends?" could be another interesting starting point.

It will be interesting to see if something like this can gain traction.

There is not really much there yet, but the idea seems to be aggregate all your information related to interaction/social-networking/communication in one place. One place under _your_ control. That's feature number one.

Then, it could present this information in a way that is actually comprehensible by mere humans. Keeping track of contacts/addresses/events/news/photos/videos/conversations spread up between Facebook/Hacker News/lesswrong/Flickr/YouTube/LinkedIn/blablablablabla is not easy.

It may be of much less interest to you if you like to just forget about what you've done in the past (as it is practically done outside the digital world). Or if you do have an actually working sync-solution cross all your communications means.

Besides having a better overview of my digital self through this, I think I'd use it to come back and convince myself regularly that yes, I also have said so immensely stupid things in the past.

I think it is a great idea -- but to make it useful, it requires a tremendous amount of work.

I'm not sure where all of this is going to end up, but to me it reads like a rehash of the thick/thin client see-saws that the market has been supporting since the dawn of client-server computing. Where personal information is both stored and contributed will be a big implementation choice as more people generate content about themselves on the Internet.

Imagine if an open source toolkit allowed you to connect yourself with other wordpress (or whatever) users with all the usual facebook type widgets. Would you use facebook at all?

That investment that Goldman made in Facebook will ultimately prove to be one of the more boneheaded investments in recent times. Just my prediction.

Why this rather than Camlistore? Are there major differences in their use cases? I really want to write a "platform" on top of either Camlistore or Locker.

Reply on news.ycombinator.com