Re: Differences in ProDOS versions

[shack@deimos.frii.com (Randy Shackelford)]

In article <4vpspd$17nb@mail.pittstate.edu>,
Bryan K. Johnson <bkjohns1@news.pittstate.edu> wrote:
>Nathan Mates (nathan@visi.com) wrote:
>: In article <32124F97.7E18@baldcom.net>,
>: Edhel Iaur  <drsmooth@baldcom.net> wrote:
>: >Just what is the difference between all those versions of ProDOS?  The 
>: >only one I can find is the addition of AppleTalk stuff, the improvement 
>: >in the "BYE" code, and the way versions 2+ require enhanced machines.
>
>: >BTW, in these days of Apple not caring much about the //, why did Apple 
>: >bother to update ProDOS to version 2.0.3?
>
>:    Each release of GS/OS also seemed to have an associated ProDOS
>: 8 update. No, I haven't done any file comparing to see if they
>: changed anything other than the version number.
>
>ProDOS is updated so that the dates will change correctly, because when
>a certain year is reached the date reverts back to 1985 or something
>like that.  So you had to patch ProDOS or get a later version so it
>would keep the year correctly.  I don't know much of the details on how
>it calculates the date other than it uses the day and month to figure
>out the year.

You can't mix ProDOS 8 versions since GS/OS has to patch some stuff into 
ProDOS 8 when it loads. For instance, an extended quit call has no effect 
unless you run ProDOS 8 from GS/OS. The locations to be patched are naturally
version dependent so GS/OS checks the ProDOS 8 version and gives the sliding
apple with "Wrong OS version" if it's not the one it expects.

As far as the year idea above goes, this only applies for //e's with Thunder-
clock type clock cards. ProDOS 8 reads the year from the IIgs clock. The 
deal with figuring out the year from the table based on the day of the week is 
due to the Thunderclock's not supplying the year.
-- 
Randy Shackelford                                 US Air Force officer
shack@frii.com                                    and Apple aficionado