[Coco] Does anyone have the documentation for MLBASIC 2.0?

[Andrew]

Regarding image size on scans:

For a manual like this, there's no reason to scan in color - it was 
likely done up on a 9-pin dot matrix printer or a typewriter back in the 
day, and so anything above 300 dpi is wasted; heck, you could probably 
get away with 150 or 200 dpi, maybe even less.

Scan as black-and-white; first try it as "line art", and adjust the 
contrast and brightness settings - this will get you a small scan 
because it will be a 1-bit black and white image. If that doesn't work 
out, go for an 8-bit greyscale scan, again adjusting settings for the 
best image.

Once you have things adjusted, scan a page in a preview mode (I'm most 
familiar with XSane on *nix), then adjust the scan area to pick up just 
what is needed and no more. Once that is done, then you can scan the 
individual images in - keeping each page aligned just like it was 
originally so that scan area is the same.

 From there, there are plenty of command line tools to convert the 
series of images over to a PDF.

You won't get an OCR'd searchable PDF this way, but you should be able 
to keep it down to a reasonable size - probably 50 meg or less (though I 
don't know how many pages the manual is).

You might try to convert it to djvu format, too - which usually gives a 
smaller file than PDF.

I'm personally fine with not having an OCR'd and searchable manual; 
those manuals were pretty easy to flip thru to find the info you needed. 
That said, if the ultimate goal is preservation (and perhaps uploading 
to archive.org or such) - then scanning with the goal of OCR and getting 
that working well should be paramount. But a quick and usable scan can 
and should be done first, just to get something out there first.

That's my 2 cents, anyhow. Cheers.

-- 
Andrew L. Ayers
Glendale, Arizona
https://github.com/andrew-ayers
http://www.phoenixgarage.org/