I have some questions about the infamous 83+ OS 1.17 that came out during Summer 2004 for a short-lived period of time on calcs sold in some stores (I think it never got available for download anywhere and the rumors about it being buggy is probably why it's so hard to find anywhere):

What was so bad in particular about it? I remember hearing things such as how it was unstable, but other people were saying that it didn't run ASM programs or apps. Apparently it seemed even worse than OS 1.13 and 2.5xMP (and perhaps the fake 2.71MP prank). Was it true and if it's unstable, then what bugs in particular were introduced? WikiTI lacks information on what bugs were present in specific OSes.

I would have tried in WabbitEmu, but WabbitEmu just displays the Waiting OS screen when I try to drag the 8xu file on the screen, and I am kinda scared of trying it on a real calc.

I have a couple friends with it.
One of them won't receive anything but lists it seems.

Ah that might explain it. I only remembered that it never came out on TI website and it was really short-lived. By the time it was noticed on MaxCoderz (the old board), it was already gone and no one seemed to want to stick with it. The LCD issues might explain it.

By the way didn't TI switch a 2nd time to different LCD drivers later? I think I remember a lot of 2004 TI-83+ calcs that couldn't run old ASM games properly, but older and newer ones could run them fine. Or could it have been the user not reporting his OS version and that 1.17 was at cause like you say?

Also WikiTI only reported some OS changes, not bug fixes nor the bugs themselves. It's mainly French sites that reported problems about OS bugs (most notably 1.13, which was very unstable). I think 1.13 was even worse than 2.53 and 2.55MP. In OS 1.13, if you went in the PRGM editor then scrolled the PRGM list when it had over 50-60 programs in it, it crashed halfway through then caused a RAM clear (or worse). I think 1.13 was when they changed the way the PRGM menu scrolls so that it's much faster, but it was totally butchered.

TI buys the LCD driver from commercial manufacturers. The calcs get whatever chip compatible with the same command set happens to be cheapest on the market. It's the same way with flash chips. TI has bought them from half-a-dozen different companies. I'm the only person who's ever noticed because I'm the only person who reads the datasheets, and there are no assembly programs that will break if the flash chips are different. (Although, IIRC, either TI-boy or zStart will use the fast write command if it's available.)