Steem SSE v4.0.0 released

[kodak80]

Version 4.0.0 has finally been released.

https://sourceforge.net/p/steemsse/blog ... -sse-v400/

https://sourceforge.net/projects/steems ... .0/v4.0.0/

[troed]

Version 4.0.0 has finally been released.

ST model: STFM
The difference with STF is bus error (crash!) on STE register LINEWID access (Colorz, Fuzion 77 & 78, Imagination, No Cooper...).

This sounds weird. Anyone knows what this is about?

[exxos]

Can only assume a register is accessed which is causing a bus error when it shouldn't, or the other way around, not very clear what it means.

[troed]

Can only assume a register is accessed which is causing a bus error when it shouldn't, or the other way around, not very clear what it means.

Yeah, but I would be _very_ surprised if that's an STF vs STFM difference. Ricoh vs IMP chipsets however I know of at least one such difference.

[exxos]

Yeah, but I would be _very_ surprised if that's an STF vs STFM difference. Ricoh vs IMP chipsets however I know of at least one such difference.

Yeah doesn't make sense saying differences between STF and STFM, basically same chipset, but like you say, different manufacturers and versions of chips yes , as we know the Shifter in the ST has some differences from later ones for example. But that doesn't mean the "ST" is different than the "STF" as it depends on chips used. So I would assume the modes added to steem and not really "correct", as its the chips themselves, not the version of ST itself.

[czietz]

The Hatari authors made the (also somewhat arbitrary) choice of emulating the ST with a Ricoh chipset and the MegaST with an IMP chipset. But they correctly attribute the difference in bus error behavior to the chipsets (and not the machine, like Steem):

https://git.tuxfamily.org/hatari/hatari ... Mem.c#n138
https://git.tuxfamily.org/hatari/hatari ... Mem.c#n149

[troed]

Not sure I like Hatari's version any better actually. It's like having WS3 being the default wakestate, I would've preferred it to be random, as on real hardware. Faking the "most compatible" setting is a step away from accurate emulation ...

Ricoh/IMP chipset wouldn't change between boot ups on real HW though, so that wouldn't work ...