On the state of stratum protocol | Bitcointalksearch.org

MaxDZ8

hero member

Activity: 672

Merit: 500

Thank you for the link! It sucks to read that!

Personally I always considered stratum a bit brittle. It's not like I require something to be fully extensible to be "future proof" but at least a version string on connection setup would have been nice.

I guess there goes my plan to take a look at cryptonight.

In the meanwhile, I connected my prototype miner to the server and it indeed gets confirmed logins as usual. Huh

I went back to the code of the legacy miner and found out I forgot to dump initiate_stratum. D'oh!
I now dump receive line. Easier and more accurate.

Thank you very much you got the ball rolling.

cayars

full member

Activity: 168

Merit: 100

https://bitcointalksearch.org/topic/m.7773643 #136 sums it up pretty well.

Carlo

MaxDZ8

hero member

Activity: 672

Merit: 500

I don't like the amount of sarcasm implied there Sad

Any gem you feel to share?

cayars

full member

Activity: 168

Merit: 100

Want to talk about deviation. Take a look at the stratum implementation used by the CryptoNight coins.

MaxDZ8

hero member

Activity: 672

Merit: 500

Hello, I've just dumped some data from a p2pool server (MYR qubit in case this is relevant, albeit I don't think so).
It has come to my attention mining.notify got an ID. It is therefore no more a notify but a request.
It also seems to me the miner isn't confirming it (obviously, since it branches on the string name first).

So it seems to me the whole stratum protocol has gone awry big way and now everyone does whatever (s)he pleases while writing servers.

I'm in the process of writing an educational miner and I would like to understand this better.

I think everybody knows now the official stratum specs by Slush are way outdated but this is taking it to a new degree... of nonsense as far as I am concerned.

What would you suggest to do?

Topic: On the state of stratum protocol (Read 796 times)