[MiNT] ()

[Joerg_Westheide@su.maus.de (Jörg Westheide)]

Hi Petr!

PS>I still consider the DATE and TIME cookies to be completely OK, just
PS>like _CPU or _IDT. They would not slow your system down, would not
PS>break any compatibility and you wouldn=b4t even notice they are there.
I disagree.
They may have some advantages but the concept is broken!

If I understand this
MS>If you need the time from an interrupt, hook into _hz_200 or similar and
MS>count ticks, and sync that local variable with Gemdos=b4 view of date/time
MS>(eg. on startup, or whenever it is possible).
PS>this is already done and was the easier part. Now one has to ensure that
PS>if user changes the date/time by calling a GEMDOS or XBIOS function the
PS>routine in interrupt must notice that.
right, then you only have to catch the calls modifying the time/date.
Have you had a look at my TraPatch?
It's easy to catch these calls by using it!

Bye

                Joerg

PS:  You can find TraPatch at http://home.pages.de/~westheide