[chirp_devel] Comments on a UI rewrite

[Dan Smith]

> I just tried MCP-4A on my Window laptop, there's no options in it to set the baud or see what it's using.  Also, while the CP210x virtual comport driver properties allow you to change the port's baud from 9600 to, say 57600, there didn't seem to be any change in cloning speed by doing so, nor could I see anything relevant in the USBPcap traces I took during a couple of tests.

Yeah, changing that default won't do anything for any non-DOS app from this century. In order to see what rate MCP is using requires something like portmon.

> Honestly, I prefer the D72's clone driver, as a full read or write operation takes <20 seconds total with it.  That said, the driver was broken for several years due to a bug relating to VFO filtering (eventually fixed in #1611).  While it works great now, anyone who tried to use it during the time it was busted are likely still using the live version, whether out of lack of trust or sheer habit.

Okay, well, it sounds like maybe the best thing to do would be to just not expose the live driver for that radio in the UI. It might be useful for someone using chirp as a library to control the radio, but it sounds like general purpose usage makes it pretty pointless. When you were using the old UI, did you know you could hit Esc to stop it loading the whole thing and then refresh certain regions to avoid it doing all the stuff in the background? Moot for the new interface though, of course ... :)

> The live version under the existing UI is ok too.  It takes a few minutes to do the initial read, but then each edit gets instantly applied/rejected.  My concern was mainly that the new UI greatly increases the amount of time to sync the changes back for the live driver, which some people might not be happy about :)  However, if you can add the ability to just sync the dirty memories edited during the same session the radio was read during, that'd probably make it a non-issue.

Certainly for the live-only radios, although on my D700 the "upload" is pretty quick. My D710 will do 115200 but I haven't lugged a laptop out to the Jeep to try it on the new interface yet.

> That said, regardless of the impact to this specific radio/driver, the new UI still seems like a positive thing overall.

Yep. I spent some time trying to make the new "treat it like a clone radio" interface even more seamless with the ability to save an "image" and such, but I think it's going to take a more substantial refactor to make that happen properly.

Anyway, thanks for the feedback on the D72 issue and its relation to this effort.

--Dan