> AFAIK the current implementation is also lacking on ways to relay the > device limitations from the driver to the user, ie. max sample rate with > the selected amount of channels, max sample depth for non-streaming > devices.
This should be taken into account and while not immediately implementable, it should be kept in mind. A good example of this is pulseview that really could use this to say "if N channels are selected, max sample rate is f_sample_max/N". However, I suspect this will be very difficult in the general case, especially with complex devices offering multiple modes, channels, channel types, triggers, etc. Before trying to solve this, I suggest people find some examples of complex devices and try to write down the rules on a piece of paper... if even possible. > > Creating a new block that provides data to the sigrok pipeline would not > > only be completely decoupled from all other blocks, it would also not > > matter which language it's written in. That's where I'd like to go. > > This is the better approach indeed, but I usually prefer to build > small evolutions of on top of well used legacy stuff. That's where > immediate huge value is and I personally cannot embark myself in a > long journey anyway. Besides, nothing stops us to have a packet based > protocol that can be used via stdin/out or tcp or even mqtt if want to > be super fancy. TBH I'm a bit lost here since I don't know what the "block" is specifically. It's not a device driver, is it? I looked at it briefly and it looks like this "block" would have to be able to provide a bit more than the RPC functions and sr_dev_driver struct - a new loader would have to be implemented (and while this could be hacked together as a device driver, I'm not sure if it would be long term viable). > So, if you don't object loudly, i'll move forward and try something out. I'm not the one to give you a go ahead as I'm not the autor of the maintainer. However, I'd be happy to collaborate and have a look at a proposal and try to figure out the best approach. Best regards, Ladislav > > Cheers, > -- > Steve Schnepp -- S pozdravem Ladislav "Krakonoš" Láska http://www.krakonos.org/ _______________________________________________ sigrok-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/sigrok-devel
