See also: IRC log
<scribe> Scribe: AdrianHB
<danmiller> these recent openssl ASN.1 parser issues worry me about the crypto-conditions encoding. But hopefully OER is less crazy to work with
<dimitri> https://bluejeans.com/795795755/
If you are dialing in for the call it's just started
Stefan: quick update.
... some major changes to reference implementations
... lots of contributors to crypto-conditions spec now
stable
... time to update the code to match
... we've also used crypto-conditions for signing the
notifications
... actually sped things up (uses some built in crypto
libs)
... internal feedback has been good
... nicely abstracted so can use as drop in lib
... esp for multi-sig
... will be a lot of uses outside of just ILP
... seen some inbound interest for others
... still some issues about naming (some people aren't
happy)
... want to be careful that we don't make it confusing
... lots of people now using them so don't want to change
prefixes
... cc are also more general than sigs so don't want to use
sigs
Stefan: working ona writeup on
how we think it should work
... considers some of the ideas on the mailing list (thanks
contributors)
... one of the biggest changes is an idea taken from how
routing works in TCP/IP stack
... routing not done by sender
... most efficinet routing is to route for local
environment
... will have good knowledge and can cache routes
... so proposal is to do incremental routing
... reliability of routes can be determined by ledgers as they
have some data on connectors they use
... difference from Internet is trying to avoid an equivalent
of IANA
... proposal is to use Interledger to pay to advertise
routes
... there is a prototype which are in PRs on connector, ledger,
and sender etc
<xavier> "will have good knowledge and can cache routes"
Stefan: early days though so consider unstable and please give input
<xavier> I like where this is going.
<xavier> would like to discuss later (tomorrow?) on this channel
<xavier> same for this "reliability of routes can be determined by ledgers as they have some data on connectors they use"
<xavier> justmoon: +1
stefan: correction - neighbouring
connectors will maintain this data
... intent is to keep functionality in connectors
<xavier> "neighbouring connectors will maintain this data" good for me
<justmoon> dimitri: Woo, crypto-condition HYPE! :)
dimitri: is now using
crypto-conditions as a base for all transactions and works
perfectly
... what would be a great demonstrator for us would be to run
with another ledger and prove connectivity
... what is the effort to get this up?
<xavier> I'm looking forward to the write up. hafta go now but will be on this channel later/tomorrow
evan: in theory, in the connector code you just add a ledger plugin
<justmoon> xavier: Cool! Thanks for stopping by!
evan: it's a bit old and so there are some hard-coded assumptions in other places
<xavier> justmoon: sorry now is night time on my side of the planet
evan: need to clean those up
dimitri: can I get some help as we try this
evan: yes
stefan: I got a lot of calls
about why we weren't at Consensys. Thanks for those that were
there who were talking about ILP
... we know that we need to do better with outreach but are
trying to balance adding noise to the group with some quick
early progress
... we think we have gotten to a a good level of maturity where
we can be more proactive about getting more input
stefan: working on design of test
suite and trying to find the ways it could break and document
those
... also starting to write and implementation of that that will
enable a data driven test suite
... quote a few different language implementations being talked
about so this is becoming a crucial req
... lots of exploits caused by edge cases in crypto algos so
need to test for standard behaviour
... hope to share test plan soon
stefan: no updates for a while.
Now uses latest package versions but also switched ledger
notifications from webhooks to websockets
... websockets are a little more lightweight
... the change removes reliability of delivery
... the realization was that ILP already accounts for lost
notifications
... this also made the demo much faster
... next change will be to implement the new routing logic
which should increase speed significantly esp for large test
nets
<justmoon> as AdrianHB said, gotta run! thanks all!!
evan: we've been discussing the
diff between the ledger and the wallet
... ledger is lower level and wallet implements some
"application layer" stuff on top
<emschwartz> * Custom condition logic
<emschwartz> * Custom memos (as objects)
<emschwartz> * Websocket notifications
<emschwartz> * Webfinger address resolution (just so users don’t have to type ledger account on login)
<emschwartz> * Simple, unchanging API for sending payments (don’t want to include f-b-sender)
https://github.com/interledger/specs/blob/gh-pages/0001-interledger-architecture.md
<emschwartz> * Simple API for pathfinding (needed for normalizing amounts in sender’s currency)
<emschwartz> * All of these would ideally work with any ledger
evan: these are the functions of
the application layer protocol that are needed
... the idea is that developers can get a lib that implements
this and building on ILP is easier
RRSAgent make minutes
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: AdrianHB Inferring ScribeNick: AdrianHB WARNING: No "Present: ... " found! Possibly Present: danmiller dimitri emschwartz evan https jamesyoung justmoon stefan xavier You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 04 May 2016 Guessing minutes URL: http://www.w3.org/2016/05/04-interledger-minutes.html People with action items:[End of scribe.perl diagnostic output]