Hi Bart, Thank you for responding. > > Sorry, horribly newbie questions and if I should be asking it > > somewhere else then please tell me. > > Hi Thomas! No, that's perfectly fine. That's good because I've been watching the traffic and it mostly seems to be reviews for code that is almost, but not quite, entirely unlike what I'm interested in. I'm guessing I'm either not looking hard enough or that it is for the kernel level implementations below the libgpiod API? - note that I've never had to deal with kernel level Linux development before, I'm an applications developer IRL. > > 1. Have there been any releases of libgpiod v2 ? I can't see any tags > > in the git repo later than v1.6. > > Nope, libgpiod v2 is under development and will probably still be so > for a while. OK. Could you explain what gives with something like this then: https://packages.debian.org/buster/libgpiod2 which advertises itself as "libgpiod2", but on closer inspection says version 1.2-3, which I assume is the corresponding version number here: https://git.kernel.org/pub/scm/libs/libgpiod/libgpiod.git/ Are libgpiod2 and libgpiod v2 referring to different meanings of "2" ? > > 4. Should I even be trying to use libgpiod v2 yet ? > > Probably not, unless you have a good reason to (writing bindings or > whatever). It's not yet stable and is about to change again soon. LOL. Not sure what you mean by "bindings". A friend and I are porting MMBasic, a *very* niche BASIC interpreter that is usually at home on microcontrollers, to Linux. The language has an array of inbuilt commands for controlling digital and analogue pins, SPI, I2C, etc. and libgpiod seems like the natural match for implementing the former. A previous (Raspberry Pi specific) attempt by another developer floundered because (apocryphally) he was trying to treat the Pi as a microcontroller and programming to too low a level (directly to the Broadcom API) and finding that it was breaking on every O/S update. We were hoping to avoid a repeat of this by relying on a higher level API. I imagine you are busy, but would you care to offer an opinion as to whether we should persevere with API v2 (where my friend has had some success) or roll-back to looking at v1.6.x, assuming that it even provides everything we need ? Best wishes, Tom