These are chat archives for ethersphere/orange-lounge

12th
Sep 2017
holisticode
@holisticode
Sep 12 2017 00:00
second diamond diagram you mean?
Viktor Trón
@zelig
Sep 12 2017 00:00
yes
which is step 2 on the time dimension
holisticode
@holisticode
Sep 12 2017 00:02
ah so the diamonds reflect time vertically, i.e. each step corresponds to one diagram, top first, last bottom
?
in other words the complete protocol comprises all three diamonds above?
Viktor Trón
@zelig
Sep 12 2017 00:04
correct:
  • the movements on one plane can be considered instantaneous
  • any systematic time delay in effect depicted by steps
  • maybe clearer if the diamond are drawn as "planes" using dashed dvertical ilines to connect the same entities
a few i drew on my drawing
holisticode
@holisticode
Sep 12 2017 00:06
ok I think I understand
Viktor Trón
@zelig
Sep 12 2017 00:07
not sure im serious about this but
holisticode
@holisticode
Sep 12 2017 00:13
it's just a different (non-standard) way to visuualise
not sure it holds as soon as we want to provide diagrams with witnesses though
holisticode
@holisticode
Sep 12 2017 00:21
I finished re-reading your paper (commented on PR)
Viktor Trón
@zelig
Sep 12 2017 00:21
no it wont work
holisticode
@holisticode
Sep 12 2017 00:21
There are a couple of things I still don't understand
I believe the process could be to produce more diagrams
this will validate my understanding, making eventual needs for the paper explicit (like it happened with the chequebook above)
even if not every diagram needs to make it into the paper
Viktor Trón
@zelig
Sep 12 2017 00:23
after the diamond is explained with the dimensions, the idea goes through
and processes can jus t be numbered list of sets of arrows:
Viktor Trón
@zelig
Sep 12 2017 00:29
1) issue: A -> cheque -> B
2) cash: B ->cheque->W(A)
3) clear?: W(A) -> transfer -> W(B)
[ 4) notify: W(A) ->withdrawal->A, W(B) ->deposit-> B ]
once you get whats what this is better, drawing diamonds is unnecessary and annoying
holisticode
@holisticode
Sep 12 2017 00:31
yes
Viktor Trón
@zelig
Sep 12 2017 00:34
and this way you can describe formally an entire process involving swear and swindle too
aaah
Viktor Trón
@zelig
Sep 12 2017 00:40
in fact this could be used as a dsl to describe all processes of a sw3 instrument / game
Viktor Trón
@zelig
Sep 12 2017 00:50
which means implementations can be just generic and use the dsl resource to drive the client logic to deploy contracts if needed, registration, swapping, ... litigation
thanks @holisticode
Viktor Trón
@zelig
Sep 12 2017 07:57
@jmozah please address all issues raised by @lmars re ethereum/go-ethereum#14382
thanks @lmars for the review ;)
Viktor Trón
@zelig
Sep 12 2017 09:49
stand-up at 3pm CEST
Aron
@homotopycolimit
Sep 12 2017 13:01
that's now
Zahoor Mohamed
@jmozah
Sep 12 2017 13:05
:)
@lmars I understand your concern regarding the waitgroups usage in the chunker. There is a seperate task i am working on which will fix this both in chunker and in pyramid chunker.
@zelig do you want to me fix the “concurrancy control” task in pyramid chunker PR or do you want me to do it separately?
Viktor Trón
@zelig
Sep 12 2017 13:09
there are other issues raised in the PR
let's fix those first
then start working on the concurrency control, if the PR still open and we reviewed your new addition we will sneakily add it to the PR to accelerate the merge process
Zahoor Mohamed
@jmozah
Sep 12 2017 13:11
Yeah.. will adress those
Viktor Trón
@zelig
Sep 12 2017 13:11
FYI friday is gonna be review day for the go team, so I hope i manage to push our open PRs
Lewis Marshall
@lmars
Sep 12 2017 13:12
@jmozah yep understood, just thought I'd lay my thoughts down whilst I was looking at it, I suspect other reviewers would say the same :grin:
Zahoor Mohamed
@jmozah
Sep 12 2017 13:13
@lmars sure..
holisticode
@holisticode
Sep 12 2017 16:18
@zelig @homotopycolimit @lmars this is the PR I was talking about in the standup
fixes issue ethereum/go-ethereum#14872
Travis CI build failure above seems to only occur for go1.9.0 and for items unrelated to this PR (whisper)
I am pretty confident about this PR in general, but I fixed several issues (mentioned in #14872) of inconsistent behavior in manifest handling
as changes affect manifest handling, it affects swarm behavior directly (my first of this kind), thus review is appreciated :)
thanks
holisticode
@holisticode
Sep 12 2017 17:06
@lmars currently uploading a snapshot fails for me
most of the nodes in the backend seem to start
INFO [09-12|12:07:33] started node b269d2d55131dcb0c334030a2c9f01df72f59c5143c2558d56f4ab3d1c17614e84954c05598f5b6623a82c0cf1f6f4ec145cc5b81bb3aed8bf5788080eccf97a: true
but in the frontend I get:
500 Internal Server Error
responseText:
"one 6d7331750df133e021465db235de38870231c6cc0be542df4b254e9eeb727055a55ea2705b2930862cde9e0b933494634861918e96677cd2628eab2bc8427f4a is not up↵"
indeed that ID seems to not have been started in the backend (no log for it)
holisticode
@holisticode
Sep 12 2017 17:12
I am uploading a snapshot of 30 nodes
the backend logs 29 "started node <id>"
suggesting that the missing one reported as error is indeed not started
Lewis Marshall
@lmars
Sep 12 2017 17:13
hmm ok, do you want to send me the snapshot and I can take a look? I'm about to head offline for the evening
holisticode
@holisticode
Sep 12 2017 17:13
I don't think this is urgent; you could have a look for it tomorrow
Lewis Marshall
@lmars
Sep 12 2017 17:13
(snapshotting should work, not sure why it wouldn't)
holisticode
@holisticode
Sep 12 2017 17:13
by email?
Lewis Marshall
@lmars
Sep 12 2017 17:14
sure, or stick in a gist?
holisticode
@holisticode
Sep 12 2017 17:14
I was going to do that
hang on
Aron
@homotopycolimit
Sep 12 2017 18:10
Did you see the raiden network visualiser? https://jsfiddle.net/UloPe/3jy8dwn3/2/show/
holisticode
@holisticode
Sep 12 2017 18:12
no - and it's awesome
thanks
it's using D3 in 2D
the major difference seems to be that it doesn't update in real-time (as it looks)
which means the network can be loaded in one go, eliminating performance issues
holisticode
@holisticode
Sep 12 2017 20:53
@zelig what do you think of this diagram
chequebook.gif