TomCh

OK, I've figured it out... Firstly I added some logging to AppleWin (latest modern code + with either the transplanted beta2 or 1.00 EDS "optimisation"). The EDS optimisation only triggers...

Summary: - TZ disk5i.nib had track-11, sector-4's address and data fields split by ~0x171 0xFF nibbles (due to 0x1A00 padding & ring buffer) - beta2 had an optimisation that when...

>Of course all the other TZ .nib images should be checked too I added a quick bit of code: for an inserted NIB, check all tracks that the first D5...

>Does it make sense for modern AppleWin to support these badly formed NIB's, or should it just identify them an put up an error? If the NIB images are read-only...

>I mean in general, what was used to create the ~1000 Asimov NIB images ? I guess it was [Super Saltines Transcopy](https://macgui.com/kb/article/75). Also [EDD](http://yesterbits.com/topics/edd-image-project/) and [I'm fEDD up](https://brutaldeluxe.fr/products/apple2/imfEDDup/index.html) create them...

>how did you generate those track layout tables without combing through each one by hand? (Or did you?) I just hacked something into AppleWin to produce that. Should be easy...

>It'd be interesting to check all the NIB's (eg. on Asimov) to see how many other potentially problematic images there are I knocked up a quick bit of python to...

Thanks for confirming this, Marc. I think Witch Trail (which looks great, btw), is the first title ever to use both Mockingboard and Mouse!

This is an interesting question. We are currently discussing this topic offline, and will follow-up with a response soon.

@audetto is gradually splitting up the code to separate the platform-agnostic from the (Windows) platform-specific code. btw, have you tried @TomHarte's [Clock Signal](https://github.com/TomHarte/CLK). This works on Linux.