Miro Kropáček

But how would you explain the first post's "fix" was to tweak some CT60 stuff?

I guess the ultimate proof that it is really some kind of HW issue/corner case would be to reproduce it under TOS. I think in one of the threads somebody...

I guess this is still relevant and hasn't been fixed?

Can you test the latest build to see whether it is still there? As you say, most likely it is.

> I think the only solution is to identify which drivers are abusing of keeping the interrupts disable for too long and add the brilliant solution implemented by @czietz for...

>> Is there a performance impact from keeping asserts in release builds? > The performance impact of keeping assert in release builds would be minimal. Just my two cents: it...

That looks more like an Aranym issue to me. I'd imagine that Aranym keeps that file open all the time for exactly that - realtime update from the Atari side....

Just to be on the safe side, can you try to replicate the issue in Aranym? V4SA still contains some bugs (sometimes not very obvious), so just to rule something...

Wouldn't this be fixed also by pointing the script to installed mintlib's timezone database? What files does it check for?

Definitely, that's pointless. I'm just worried whether `--with-libstdcxx-zoneinfo=no` doesn't disable something in libstdc++ for good, i.e. whether all the time-related functions will still work as expected, just relying on underlying...