Thanks for the heads up, and sorry about that -- caught us right at the wrong moment

We just had some downtime (~30 minutes) as we were migrating the main production server temporarily to backup storage as we installed new hardware for the mapping server.

We're configuring a new RAID setup this afternoon, so once that's ready, we'll probably have just a couple minutes of downtime as we switch back over to that.

We try to keep things up to date on our twitter so everyone can have a heads up as to to any potential downtime -- @elqntdotorg

Just ran through some tests this evening w/ Tor and everything still works great, albeit a little slow (to be expected when using Tor).  

Let me know if there are any issues on your end when traversing w/ Tor.  

Update in follow-up post.


Great!

Makes sense, we'll work to clean things up re: typography.  Thanks for the feedback!  

NAT traversal shouldn't be a problem, barring a blocked port.  Basically what happens is there is what's called a STUN server that helps each individual peer know their public IP and port.  The ID that the user assigns itself (done automatically), becomes the route to communicate with that user.  All communication is done directly peer-to-peer via JSON calls.  We have our own schema for how those JSON calls are structured.

We're working on a method to handle persistence, but haven't yet implemented this.  We already have the functionality basically spec'd, but aren't yet sure with what release it makes the most sense for.  

= Version Changelog Archive =
*This is linked to in the first post*

v1.1 - minor release (7.27.13)

- Substantial Back-End Refactoring In Preparation for Major Functionality Inclusions
- Analytics Opt-Out
- Mobile UI Improvements (Will come into play more as mobile Chrome supports WebRTC)

v1 (7.19.13) --
- As a buyer, find local sellers of Bitcoin
- As a seller, list either an exchange area or a specific location for which you would like to sell Bitcoins within
- early currency I18n (still needs work)
- All data + communication is conducted directly p2p utilizing WebRTC -- the only data our server collects is peer Id's (volatile - i.e. once you leave we remove the ID from memory)
- Mobile compatibility w/ Chrome (still mostly broken, but working to clean this up very soon)

Our core consideration with the entirety of this suite of services is on anonymity and sadly all of the major current mobile OSs are a bit dubious when it comes to anonymity + privacy (in large part due to impositions by the carriers, and following that chain up, by the respective governments that have jurisdiction over them).  

That being said, we're sticking with web based at this point.  We may do an Android app, but more likely would be a custom Android build which strips all 'phoning home', along with all user data collection and is purpose built for privacy surrounding communication, financial transactions, and user data.

We're also eyeing Ubuntu One, but it's still too early to make any conclusions.

Thanks! 

Any thoughts on functionality improvements or feature requests?

It's basically identical in regards to networking typography.

The way in which we're different is that our servers only do two things:

1) Maintain a list of active peers (ie - peers currently with browser windows open) by client generated ID ONLY.  
All of the rest of the data (price, rate, location, chat, etc.) is directly transmitted between peers and never touches our servers.  It's also all encrypted peer-to-peer to help against surveillance.  We also have checks in place to prevent ID spoofing.  

2) Serve the necessary static HTML + JS files.  We've built the exchange in Ember.js so structurally it's basically a self-contained browser based app that only communicates between the mapping server (soon to be in-house as well, today possibly...), peer server, and other peers -- all encrypted.

Lastly, we've open sourced both the peer server and Elqnt itself to give everyone a chance to dig into the code and audit it for security vulnerabilities and also contribute back.

Well shoot...could be a couple things:

We've been doing some pretty heavy server modification over the past couple of days with our new in-house server, so you might have caught the site in one of those periods.  I try to keep our twitter -- @elqntdotorg -- updated regularly with a heads up to downtime.

Also, if you're using anything other than Chrome or Chromium, we've blocked the exchange due to compatibility.  WebRTC is implemented in the stable release of Chrome currently making it much easier to maintain on our end.  FF has it implemented in the nightly release, but some stuff is still not quite stable, so we're holding on support until things have stabilized.  Doubt support for anything else (besides maybe Opera) is coming anytime soon.  Most of this support hinges on which browser has implemented WebRTC and to what degree of effort we have to apply to make it work in our system w/ browser quirks.

Lastly, give it another shot.  Should be back up, if not, could be a bigger issue.

I'm excited to announce Elqnt! - An anonymous, P2P, Bitcoin Exchange -- https://secure.elqnt.org/

Keep an eye on this post, as we use this for the most up-to-date info, with subsequent posts being used to answer question/discuss, highlight certain info, and to house an archive of content.

== Overview ==

Elqnt is a decentralized + anonymous + peer-to-peer Bitcoin exchange.

White Paper -- https://git.thebody.io/elqnt/whitepaper/blob/master/whitepaper.md

== Updates ==
12/8/13
We've recently taken a completely new tact with the technologies that have been developed thus far -- resulting in a massively different service, but with the result of providing a substantially more beneficial platform to all those involved.

We've written up a white paper to detail how the new incarnation of the exchange will function -- https://git.thebody.io/elqnt/whitepaper/blob/master/whitepaper.md

== Donate ==
We are intentially not actively seeking institutional investment due to the conflict of interest around user data collection.  We believe this data is yours and yours along; thus, we want to place ourselves in a position of only having 'useless' data such that even if compelled, we hold nothing of value re: identifiable user data.
Because of this, the project has thus far been funded primarily by the individuals involved, but we will need donations or revenue to continue meaningful development going forward.

Please donate if you like this idea and want to see it happen!


17ost3nMZoGHYmNud4Qer7fybb782UCwDH

== Contact Info ==

Twitter: @elqntdotorg
E-mail: stephen@elqnt.org

We would love thoughts, feedback, insight, and anything constructive to continually grow + improve the project!

It's finally here!  I just released the new version of Elqnt.org

This version is mostly a re-write (about 70% of the code) to use Ember.js.  You won't notice really any major functionality changes yet, but you'll definitely notice some significant UI changes w/ the seller + buyer workflows.  Bugs abound, but I'm knocking them out as fast as I can.

Also, big news, I've open sourced the entirety of the platform.  I'd definitely love if anyone is up for jumping in to help develop Elqnt (I've also got larger plans for the brand on the whole).  

The Elqnt Thread -- https://bitcointalk.org/index.php?topic=259659

Best,
Stephen

Thanks!  Definitely will.  The plan is to open source everything this week.  Once that goes live I'll do one more post to let everyone know. 

I'm also going to be starting up a new threat just for Elqnt so I don't further clog up this one

In follow-up to back-end security --

I host the web server and peer server in-house.
Analytics utilized (Piwik) are self-hosted as well, in-house.  I also plan to let users opt-out of ananlytics, but haven't had time yet to set this up.  
All JS is self-hosted as well, in-house.

Basically this means no external logging occurs, except for the tile server (for the time being) and any other ISP related logging.

Tile server uses ssl, but is currently hosted through MapQuest and serves Open Street Map tiles.  I'd like to bring this in house, but the hardware requirements are a bit beyond what my stuff can swing at the moment.  For details on the data they collect, check it out here -- http://developer.mapquest.com/web/info/terms-of-use
Specific section in their ToS is: ACCESS AND USAGE DATA > Usage Data

Thanks, really appreciated. 

Just integrated 2048-bit SSL today (made it a priority based on your comment). 

All web and WebSocket traffic are encrypted now.  This coverage also includes all peer-to-peer communication as well.

Thanks!  I'm super excited to be building something valuable for the community. 

Re: your points --

1. Thanks for the heads up!  In contact...

2. In process...

Yep 

I completely agree re: coagulation of liquidity, but with anonymity comes a certain challenge to trust.  At this point it's a physical exchange, and thus the burden of trust is on end-users, elqnt simply provides a means to find each other.  The issue with the method you mentioned is that as soon as you have a centralized organization producing any kind of 'credit ledger' or the like, they can soon expect to get a knock on the door for being an online money transmitter (as seen with other internally developed digital currencies used to exchange BTC).  Staying at the individual level with direct peer-to-peer exchanges keeps things below the regulations and issues around being considered an online money transmitter (which brings with it a removal of anonymity to comply).  Making this peer-to-peer exchange as smooth as possible is the goal and some form of distributed 'credit ledger' would be fantastic to layer on, but I'm not sure how this would work.

Regarding services like ripple and other exchanges:

The core issue I take with networks like ripple is that they break the benefit of anonymity, and without this, Bitcoin looses almost all advantages over fiat currencies.

I lost faith in Bitcoin for a period of time simply because if a 3rd-party can observe who the money is coming from when it comes in and who it goes to when it comes out, then decentralization and internal anonymization mean almost nothing.  The exchange in + out of Bitcoin must be anonymous in order for the decentralized nature to have benefit.  Once you loose anonymity, enforceable regulation is soon behind, along with taxation.

The prior is why I'm opposed to any layering of identification onto Bitcoin (much like what Google is working on internally, to produce 'clean' Bitcoins via layered identity) or tracking in/out of the platform itself through linked bank accounts at online exchanges. 

This ultimately led me to explore options that would fit within current regulations and thus WebRTC produced the most readily available solution, and now I see a possibility for Bitcoin to remain valuable through Anonymous exchanges conducted peer-to-peer physically. 

Currently it's a physical exchange only, much like Craigslist.  Elqnt provides a means to find exchangers in your area, which you chat with directly in the platform and agree on a price and meeting place to conduct the exchange.  It's up to the end users to handle the trust of the exchange, but my current conceptual work-flow to handle trust with these physical exchanges is the following scenario --

Two users agree on a price + location.  They meet at the location, cash is presented and verified.  The bitcoin holder makes transfer, which is then verified after a few minutes (grab a coffee + chat while you wait).  After verification, cash (or whatever) is given to the seller of Bitcoin.  If both parties are happy with the exchange and would like to follow-up with further exchanges, additional contact details might be exchanged for direct communication later. 

I'd love to implement a reputation mechanism, but I'm shying away from UIDs and accounts all together.  The vision is an exchange that is a completely open ecosystem which will solve the 'last-mile' problem for those wanting to exchange easily and anonymously with Bitcoin.

I've been working on a new peer-to-peer exchange that employs WebRTC for the past few weeks. 

It's pretty unique because it fully complies with FinCEN's regulations around exchange because all communication in the platform is conducted directly peer-to-peer and subsequent exchanges are conducted physically by the respective exchangers (I'd love to employ an online transfer in further iterations, but don't have a good solution yet). 

I just went live with an early alpha tonight if others would like to try it out and give feedback. 

My primary goal with the platform is give as much anonymity as possible, meaning that I self-host all files (no external js).  I'm stuck using an external tile-server (which means mapquest logs some data, including IP + timestamp), but I hope to get some capacity soon to bring that in-house as well so no external logging of users occurs (beyond ISPs).

The site is elqnt.org (pronounced eloquent).

Best,
Stephen

Nice.  We have adequate protection. 

bumping for any thoughts, questions, or feedback.