i'm trying to use the pre-built INDI packages on ubuntu 13, as well as kstars/ekos.

i understand the configuration for the mount is basically a serial port device. but the "default" video capture is /dev/video0 - this makes no sense for my camera (SBIG).

i have gotten the SBIG software working (and even modified it a bit) and - like the QHY8 drivers I played with months ago - the drivers look for the correct USB device ID and then connect using libusb. So I am not sure that there is a "device" per se, hence it's not clear what sort of configuration the SBIG INDI driver requires.

also, longer term i want to do the following from (Unix) command-line:
1) capture images (i can already do this from the SBIG sample apps)
2) send and receive commands via LX200 (i can do this using any software that can send/receive serial commands)

i want to prototype a solution that captures short images, plate-solves them, moves the mount in RA, captures more short images and plate-solves them (again) and thus build a limited mount model so that i can experiment with something similar to the ASA MLPT (or Software Bisque ProTrack) feature. so the software would then repoint the mount to the object of interest, and supply guiding corrections blindly (with no guide scope!) based on the RA/DEC drift track derived from the multiple plate solves.

it would be better if i could do these using INDI so that my solution is not hard-coded for my particular hardware (I can easily prototype the above using the SBIG sample app and by banging bytes over the serial port) but i don't know where to start.

You are right, KStars not suppose to send Video port option over to CCD devices, I fixed that issue now in KStars. But SBIG driver uses the port to specify the device type in a call made to SBIG Library. The default value is "sbigusb0".

I would highly recommend you look at the recent discussions we had in the INDI mailing list on this very topic. (some discussions like this sourceforge.net/p/indi/mailman/message/31646023/). There is now a development branch libindi-align that is tackling this issue so that a generic high level alignment layer is made to be compatible with all INDI mounts.

Excellent. I just revisited this and it works great (I was using your repository and just blindly updating).

SBIG camera, AP mount.

I am not sure if the AP driver uses all of the features of the AP command set. For example use of RCAL instead of RSYNC, and the variable-rate dual-axis tracking. Since INDI now works great for me, I will port all of my code to it.

Additional question - is there any idiot's guide to writing command-line INDI applications? e.g. how do I capture an image; or how do I command a mount slew. etc.

Finally: I found this Fortran code for mount modeling. The code compiles cleanly on Linux (the source is GPL) and also cleanly converts to C via f2c. I would like to add this capability somewhere... but I don't know what sort of module it is.

I will write a tutorial soon hopefully for common command line examples. The INDI alignment layer is actually now ready and in production, and I plan to use it very soon for a 1-meter telescope. Please take a look at it.