|
Hello,
BUT Companion app works flawlessly, confirming the base protocol is correct. Goal - what I don't know:
Pairing in Elgato app is partially working: I saw in Elgato app new network device appeared 1x (blinks - shows for couple of milliseconds) and immediately disappears.
Also I saw first button on webUI has changed to full black image - so Elgato app sends request to my app and it rendered first button. I I noticed after pairing it as "Add Network Device" it added new device under Preferences - Devices 2x:
My ideas:
Could you please help me and point me where I should take a look? |
Replies: 1 comment 17 replies
|
I don't know how much I will be able to help here, its going to be a lot of guesswork. My approach to reverse engineering something like this is to start by using a wireshark capture of the real thing and effectively re-play the packets. Once that works I move on to try to generate packets correctly. Usually I do this when there is no prior code to base it on though, but maybe this would be a good approach for certain packets which you are unsure of. And the TCP nature will complicate it, because of the cora headers which will need generating correctly
Possibly, but I don't think this will matter. those bytes are for a checksum of the firmware, I cant see why it would care what they are.
Yeah probably this. But figuring out what is the challenge. One thing to figure out is whether it is the network dock connection that is being closed or the secondary device. I have a feeling there were a few commands that this library doesnt implement as it has no need for them (same with some values inside of messages), but I couldnt say what they were.. Hopefully they can be at least discovered by what report ids it sends and we dont have an answer for |



I don't know how much I will be able to help here, its going to be a lot of guesswork. My approach to reverse engineering something like this is to start by using a wireshark capture of the real thing and effectively re-play the packets. Once that works I move on to try to generate packets correctly. Usually I do this when there is no prior code to base it on though, but maybe this would be a good approach for certain packets which you are unsure of. And the TCP nature will complicate it, because of the cora headers which will need generating correctly
Possibly, but I don't think this will matter. those bytes are for a checksum of the firmware, I cant see why it wou…