#ls2 (irc2p) | #ls2 I2P Protocol Development Discussion - Next Meeting Monday, October 9th 6:30 UTC

orignal ahash: Alice's 32-byte router hash (for messages 3 and 4 only; not included in the message)

orignal what does it mean?

zzz thats the signature calculation

orignal I know but what does this pharse mean?

orignal is it included to calculations for message 3 or 4? or it's not inluded to message 3 or 4

orignal how about 1?

zzz it means you only use it in the signature calculation for messages 3 and 4. And it's not included in message 3 and 4.

orignal why not in 1?

orignal 1 includes Alice's ident hash

zzz it's not really necessary to include it in the signature calculation for 1 and 2, because Alice is signing it anyway. There's no point in alice signing her own hash

zzz 0) Hi

zzz hi

eyedeekay hi

orignal hi

Xeha hi

Xeha :D

orignal so not ahash in 1

zzz right, I'll fix it up to be more clear

zzz what's on the agenda for today?

eyedeekay last) ongoing go-i2p progress report

orignal SSU2

orignal as usual

zzz 1) is SSU2

zzz how about 2) release status

orignal fine

zzz and 3) send RIs to connected peers status

zzz and 4) go-i2p

zzz 1) SSU2

orignal 5. relay testing status

zzz ok that's 5)

zzz 1) SSU2

orignal so I was busy with release and other personal tasks

orignal so not too much progress

zzz I'll give a quick update from my side

orignal it just works

zzz I put some info about SSU2 in our release notes and pointed them to my forum for how to enable

zzz we're up to at least 10 each java and i2pd

orignal 10 more?

zzz no, 10 each total

orignal not impressive however it's enough

zzz we were at about 5 each before the release

zzz we can always push it harder if we want more

orignal no need for now

zzz it's enough for regular testing but not nearly enough for peer test and relay testing

zzz I updated my peer test code to get rid of the 'role' field as we agreed

orignal for relay it's enpough I guess

orignal yes I saw it in the specs

zzz I have peer testing enabled on one router

zzz I'm going to enable it in the code soon

orignal I don't have peer test code yet

zzz and push that change so we'll get a couple more

orignal only relay

zzz I've been promising for weeks to get back to relay. This week I mean it :)

orignal nice

zzz we;ll see

orignal and I will start publishing itags soon

zzz ok

orignal remind me again

orignal if you have SSU address you can publish introducers for SSU1 and SSU2 in the same address

orignal and I differetiate by ih

zzz that's the plan

orignal I need to understand how to parse your address

zzz not sure how it's going to work if there's a mix and match

zzz haven't gotten to that part yet

orignal then what are you going to do?

orignal if you are firewallede

zzz because say #0 is ssu 1, 1 is ssu2, 2 is ssu 1, not sure if that would work

zzz parsing might be messy

zzz have to think about it

orignal not for me

orignal I create two addreses anyway

orignal when parse

zzz I had some stuff coded and forgot about it then it crashed my router and had to yank it out

zzz will know more later this week

zzz enough about 1) ? anything else on SSU2 ?

orignal no

orignal everybody is asking

orignal when

zzz i2p-projekt.i2p/spec/proposals/159 right at the top

zzz 2) release status

zzz 20% of net updated so far; no sign of bigly yet; haven't heard any complaints

orignal fine from my side

orignal no issues

zzz eyedeekay, anything to add? anything todo? android? maven? packages?

eyedeekay i2p.firefox successful in-net automatic update happened this cycle, which is when I was prepared to consider it ready to move "out of beta."

eyedeekay I communicated with a guy who uses it paired with the IzPack installer alongside the i2p.firefox browser profile and dealt with several issues in that configuration before the release as well.

eyedeekay release was delayed because my first implementation of AndroidSAMSecureSession depended on the ability to draw an alert over another application, which requires a special interactive/gated permission of it's own.

eyedeekay I changed it to use the NotificationArea to remove that step and avoid a repeat argument with Google

eyedeekay It was released Friday night after that issue was resolved

eyedeekay Those are my notes from both

zzz I would have recommended you not be coding/testing new features after the release, but ok

zzz if you don't tell me something is out, I don't know and I can't tweet it

zzz anything else on 2) ?

eyedeekay Nothing else from me

zzz 3) send RI to connected peers

zzz I have it coded and tested

zzz for SSU 1/2 and NTCP2

zzz will be checking it in soon

orignal why do we need it for NTCP2?

zzz I'm doing it about every 30 minute, starting after 30 minutes of connection uptime. I believe that's more or less what we agreed to at the May 9 meeting 3 weeks ago

orignal it's not used there

orignal we agree about different thing

zzz for java i2p there's some problems with having the RI expire for a connected peer

orignal we send it before relay request or peer test if we didn't send it for 30 minutes

zzz it's discussed on the gitlab ticket. I also put the patch there

orignal so do you want me to send through NTCP2 every 30 minutes?

zzz so yes, I decided to do it on all transports

orignal no problem I will add it

zzz I think it might be helpful but let me do some more investigation first before I ask you do the work

orignal like block or I2NP messager?

zzz for NTCP2 I use block because there's no length issue

zzz for SSU 1/2 I use I2NP so I don't have to worry about the length

zzz I can handle either incoming

orignal I meant NTCP2

zzz anything else on 3) ?

orignal fine, will add block

orignal no

zzz you can if you want, or wait until I bring it up again

zzz 4) go-i2p status

eyedeekay go-i2p progress slowed down last week mostly on account of playing catchup with Android, but I'm happy to announce that I *finally* finished both refactoring the common library and fixing the tests without breaking either

eyedeekay The refactor of the crypto primitives is started but not far along

zzz is it checked in and merged?

eyedeekay The common libs yes, the crypto stuff not yet

zzz we really need to find you some help

eyedeekay I'm going to replace the netDB parsing tools in the reseed-tools server with the common structures in go-i2p to remove that duplicate work and when I know for sure that's ready I'll tag a new version of reseed-tools

eyedeekay Yeah some help would be nice but I'm just going to keep at it and keep learning, I'll get faster as I have more success

Xeha will go-i2p become a "mainline" lib? like ipsam/libsam3?

Xeha *i2psam

eyedeekay That's the goal, but go-i2p is a router/library not a SAM library

eyedeekay Go already has 2 or 3 pretty nice SAM libraries

Xeha im aware, im asking regarding it being "the official go library/router" :)

Xeha thanks

zzz if you're looking for reuse, splitting it into a "core" lib and a "router" lib makes sense

eyedeekay That is part of how it was structured when I got it

eyedeekay Taking the good with the bad there

zzz defining your API boundaries early can also help a lot

eyedeekay That's all from me this week, more same time next week

zzz great. if you're making progress maybe it will attract others

zzz anything else on 4) ?

zzz 5) relay testing

zzz you're up orignal

orignal yes

orignal sec

orignal we need an SSU2 with itag

zzz sure, cant test without alice bob and charlie

zzz have any thing to propose?

orignal I have bob ready and working

orignal and Alice

orignal still don't have Charlie

orignal basically publishing itag

zzz ok, so you want me to start on charlie first?

orignal no, just asking where you are

orignal I'm working on that part as well

zzz I'm half done with everything.

zzz I'm still a month away from everything coded and tested on the testnet.

orignal that's fine

zzz I could put up something untested on the live net sooner, e.g. charlie-only

orignal but you peer test is ready?

orignal I can start with it I guess

zzz yes, peer test fully tested on testnet

zzz I just need to flip a switch in the code to enable it on live net for people running dev builds

zzz it's a good idea to just code charlie. I need to split up the work somehow into smaller pieces.

zzz will let you know

orignal do you publish it in caps?

zzz I keep trying to start with Alice but the state management is hard. I need to start somewhere easier

orignal you mean for relay or peer test?

zzz switch flip is for peer test. state management issues is for relay

zzz I publish caps for relay/testing, but it's really just SSU 1 caps

zzz I don't know if we need different caps for ssu2

orignal got it

orignal ofc not

orignal if your adrress is capable it's capable

orignal but see the problem now

zzz so there's no way to tell if it's enabled or not, which is fine if we get it all done before we turn it on by default

zzz unless we use router version

orignal I don't know which your router supports relays now

zzz right

zzz I assume we'd use router version, but maybe we need different caps

zzz a little bit messy

orignal so let me know when you are ready with ssu2 relays

zzz will do

orignal I will exclude this caps from SSU2 when parse your SSU

zzz ok, we'll figure it out

zzz anything else on 5) ?

zzz anything else for the meeting?

eyedeekay Nothing from me

zzz thanks everybody