Quote:
Originally Posted by Minuous
Good to know those bugs have been fixed, but really the broken versions never should have been inflicted on users without testing in the first place.
|
Well, if that helps, they have been inflicted with testing
The floppy drive issue was timing-related. It did not happen in all configurations. I seem to remember that floppy-related timing specifics was an elusive factor also in the Commodore days, with few engineers familiar with this aspect.
The GetHostVar bug did not manifest itself in a "normal" (default) emulation environment. The Classic Support ADFs were also not affected (they do not have emulation additions). But indeed, if you took a 1.3 or 3.1 ADF from the emulation, and used it on a real Amiga, or with an emulated one without the (virtual) I/O board, it would crash.
I am not trying to minimize these issues. I agree that it would have been desirable to catch them during beta testing. I also wish to have seen this thread before, rather than after somebody sent me a link to it. But, I think that we still do have a good internal feedback/escalation loop in place, whereby we try to address bugs fast, and release updates just as fast. It lacks transparency (a detailed revision history), but that too should be there in one of the next updates.
Fronts that badly need more work include the KX Light portion (the Linux-based boot code), and that's why it's still called "experimental". I am somewhat relieved that I did not have to address this here
I hope that we can keep working hard and improve during 2018 and beyond, also with feedback like yours.