Finally after weeks of rain I have been able to use my new (old) setup. A skywatcher newtonian(750mm focal length) on a neq6 with an eos 300d as my main camera. An upgrade for the camera is possible in the future but for now this old camera will have to do. Indiserver running on a raspberry pi. Wifi to laptop running kstars-bleeding.

I was having a few problems getting ekos alignment to actually solve. I managed to narrow it down to the FOV values used. Once I had those sorted it would solve very quickly.
The alignment module calculates a FOV from the image width/height and the ccd-pixel size. The eos 300D pixel size is 7.4 whereas the indi-gphoto driver is using 5 (hardcoded?) and the image width/height is read from the image and depends on whether you capture a small/medium or large image (the 300D method of binning?).

I managed to trick the alignment module in using the correct FOV (or something close enough) by changing the focal length setting in the EQMod driver.
Is there a better (correct) way of setting the parameters for my scope and camera so that the alignment module uses the correct settings all the time? E.g. does the gphoto-driver allow for a settings file where ccd-pixel size is declared? Does the gphoto-driver understand the difference between small/medium/large image and treat is as binning?

I made the property now writable for gphoto driver, can you please update both libindi and gphoto and check? Also, is there a way to detect pixel size in anyway? I don't think GPhoto reports that at all.

And no, there is no binning support in GPhoto driver.

I don't think so. Small/medium/large is only for jpegs, which is a 'forbidden' format for astronomy if you want to stack and process the images . Some high-end cameras have more than 1 raw mode if I remember correctly. SRAW mode or something like that, but as far as I know, SRAW images are generated by the camera processor using the original RAWs, so it is not completely like shooting with binning because you're still suffering the read noise in ALL the pixels (while in binning 2 you would suffer for example 1/4 of the total read noise).

Thanks. I'll update later today and will let you know.
The 300D definitely does not report the pixel size and I'm pretty sure other camera do not report this either.

Minaya,
I am aware of the 'need' to use RAW for imaging. However, the 300D image transfer is very slow (nearly 1 minute for a raw image). The use of the small jpeg in the alignment is much more convenient. I was wondering whether it is possible, or feasible to have the alignment module treat the jpegs as on camera binned.
Large (and raw) - 3072x2048 = 1x1, Medium - 2048x1360 = 1.5x1.5 and Small - 1536x1024 = 2x2.
This way the FOV calculation stays correct. It is a minor issue because the FOV calculation can be tricked by adjusting the telescope focal length in the EQMod driver accordingly.

This calculator helped me get my head around this
www.skyatnightmagazine.com/field-view-calculator

Image size F ratio: Focal Length: Field of view: Resolution:
large f5 750mm 1.73° x 1.16° 2.03"/pixel
large 2x2 binned f5 750mm 1.73° x 1.16° 4.06"/pixel
small f5 750mm 52.03' x 34.69' 2.03"/pixel
small focal length adjusted f2.5 375mm 1.73° x 1.16° 4.06"/pixel

Finally managed to update and check. Not sure I understand what property is now writable. I don't see any difference

The pixel size should be writeable.