James Hilliard
James Hilliard
> Can you try with [ecca06c](https://github.com/elementary/granite/commit/ecca06c52f09f6d931d6442800de0034322c9abe) Haven't tested but from a quick review that's probably going to cause issues, when cross compiling we have to [override the default `g-ir-compiler`](https://github.com/buildroot/buildroot/blob/2023.02/support/misc/cross-compilation.conf.in#L14) with...
Something like this seems to work: ```diff diff --git a/lib/meson.build b/lib/meson.build index 1716cf7e..a937928b 100644 --- a/lib/meson.build +++ b/lib/meson.build @@ -106,6 +106,14 @@ install_data( if get_option('introspection') # typelib generation isn't automated...
> You have 5 pull requests open, that seems to be around the same theme...could you please close the ones no longer relevant. I closed the ones that are no...
> I think I wrote that callbacks are information only. Well since `callback_data` can't be used to inject a response on timeout when using a queue...this provides the ability to...
> It would be very illogical to wait for the timeout, when you know the device is not going to respond ! Well you mentioned that "rs485 prescribes a delay...
> I do not think that has anything to do with a timeout callback. The timeout callback would be where we would inject our fake response as opposed to `callback_data`....
> That is too late, it needs to be before the timeout to have effect. Depends on what sort of effect you are looking for I suppose. If you merely...
> You do not mix broadcast and non broadcast on a single client. What's the issue with that?
> at call level broadcast is already implemented as pr modbus standard, slave=0 What is `broadcast_enable` used for then?
> The modbus protocol is a request/response protocol, so a response is always expected meaning the parameter is implementing a non standard feature. AFAIU the spec doesn't really mandate that...