adam_free2air has quit [Ping timeout: 268 seconds]
_whitelogger has joined #qaul.net
adam_free2air has joined #qaul.net
adam_free2air has quit [Ping timeout: 240 seconds]
adam_free2air has joined #qaul.net
adam_free2air has quit [Ping timeout: 268 seconds]
adam_free2air has joined #qaul.net
<spacekookie>
noracodes: So I haven't looked at your MR in a while and I'm note sure I understood the problem you were talking about on the zulip
<noracodes>
Yeah! So, I have two examples (on the commits referred to there) with essentially the same network topologies
<noracodes>
One of them - where the two routers are connected directly with a MemMod - works fine
<noracodes>
But the in the other one, where they're connected via a BroadcastMedium, one router keeps resending the same message over and over again.
<noracodes>
And I am kind of at my wits end figuring out why, tbh.
<spacekookie>
Hmm
<spacekookie>
Can I easily run this example?
<noracodes>
Yep! It's in the same place your example code was (it's just a modification of that)
<noracodes>
(clients/linux, I believe. just `cargo run` in there)
<spacekookie>
Alright, doing that right now
<spacekookie>
There is some heartbeat stuff that would resend the same packets but like, that's not in your base branch
<spacekookie>
Also most things in RATMAN would just crash if something failed to send, instead of retransmitting
<noracodes>
Hmm. It's possible I'm misunderstanding how to use the rounter API
<noracodes>
Should I rebase?
<spacekookie>
So one of the examples worked
<spacekookie>
Also rebasing to master might be a good idea
<spacekookie>
Not much should have changed but you never know
<noracodes>
Okay, I'll do that.
<noracodes>
On a slightly different topic, do you know what the state of the HTTP API is?
<spacekookie>
Not really. I pinged Jess about it earlier. I might have a look at that tomorrow an dsee what's missing and stuff
<spacekookie>
Considering how much of a pain the web UI has been so for I think maybe "just" building a native android one might be better, but that's gotta be after congress :P
<noracodes>
Okay! I'll work on my branch and touch base tomorrow?
<spacekookie>
Sounds good
<noracodes>
Yeah I'd like to have a cli as well and that should be pretty quick
<spacekookie>
Anyway, the example is workin now, not really sure what changed from a few weeks ago xD
<spacekookie>
But it means we can merge into master after the rebase and take it from there
<spacekookie>
It would propably be good to start writing some tests that use async+std and timeouts to test if messages get delivered and stuff
<noracodes>
Yeah! That's what I'm working on at the moment - promisifying the send API
<spacekookie>
Yea I saw that
<spacekookie>
I feel scared about merging that before getting a simple Android example working that we can test at Congress
<spacekookie>
BUt after that's done adding async-std to ratman and makin git all async will propably decrease the complexity quite a bit
<spacekookie>
So that'll be excitig!
<spacekookie>
noracodes: want me to just hit the rebase button on gitlab?
<noracodes>
Sure! I was pretty much going to do that but I have to wait until I'm home so, go ahead :)
adam_free2air has quit [Ping timeout: 268 seconds]