Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.sys.apple2.programmer > #2959 > unrolled thread

ProDOS 2.4

Back to article view | Back to comp.sys.apple2.programmer

Contents

  ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-08-17 00:01 -0700
    Re: ProDOS 2.4 Michael J. Mahon <mjmahon@aol.com> - 2016-08-17 17:53 -0500
      Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-08-17 22:43 -0700
        Re: ProDOS 2.4 dempson@actrix.gen.nz (David Empson) - 2016-08-18 17:47 +1200
          Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-08-17 23:06 -0700
            Re: ProDOS 2.4 David Schmidt <schmidtd@my-deja.com> - 2016-08-18 07:16 -0400
              Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-08-18 07:24 -0700
                Re: ProDOS 2.4 dempson@actrix.gen.nz (David Empson) - 2016-08-19 10:38 +1200
                Re: ProDOS 2.4 gids.rs@sasktel.net - 2016-08-18 22:33 -0700
                  Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-08-19 14:05 -0700
                    Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-05 19:42 -0700
                      Re: ProDOS 2.4 Hugh Hood <hughhood@earthlink.net> - 2016-09-05 22:12 -0500
                        Re: ProDOS 2.4 Hugh Hood <hughhood@earthlink.net> - 2016-09-06 00:05 -0500
                          Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-06 08:40 -0700
                            Re: ProDOS 2.4 Michael Pohoreski <michael.pohoreski@gmail.com> - 2016-09-06 09:37 -0700
                              Re: ProDOS 2.4 gids.rs@sasktel.net - 2016-09-06 13:26 -0700
                            Re: ProDOS 2.4 David Schmenk <dschmenk@gmail.com> - 2016-09-06 14:08 -0700
                            Re: ProDOS 2.4 Oliver Schmidt <ol.sc@web.de> - 2016-09-07 03:23 -0700
                              Re: ProDOS 2.4 Michael Pohoreski <michael.pohoreski@gmail.com> - 2016-09-07 13:37 -0700
                                Re: ProDOS 2.4 Kelvin Sherlock <email@domain.com> - 2016-09-07 20:11 -0400
                                Re: ProDOS 2.4 Michael J. Mahon <mjmahon@aol.com> - 2016-09-07 19:33 -0500
                              Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-07 17:18 -0700
                                Re: ProDOS 2.4 Oliver Schmidt <ol.sc@web.de> - 2016-09-08 01:37 -0700
                                  Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-08 08:49 -0700
                                    Re: ProDOS 2.4 Oliver Schmidt <ol.sc@web.de> - 2016-09-09 02:57 -0700
                                      Re: ProDOS 2.4 Steve Nickolas <usotsuki@buric.co> - 2016-09-09 06:06 -0400
                                      Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-09 08:29 -0700
                                        Re: ProDOS 2.4 sicklittlemonkey <nick.westgate@gmail.com> - 2016-09-15 18:21 -0700
                                Re: ProDOS 2.4 David Schmenk <dschmenk@gmail.com> - 2016-09-08 07:06 -0700
                                  Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-08 09:25 -0700
                                    Re: ProDOS 2.4 David Schmenk <dschmenk@gmail.com> - 2016-09-08 12:31 -0700
                                    Re: ProDOS 2.4 Oliver Schmidt <ol.sc@web.de> - 2016-09-09 04:16 -0700
                                      Re: ProDOS 2.4 John Brooks <jbrooks@blueshiftinc.com> - 2016-09-09 09:02 -0700
                                        Re: ProDOS 2.4 ol.sc@web.de (Oliver Schmidt) - 2016-09-11 12:56 +0000
                                  Re: ProDOS 2.4 Oliver Schmidt <ol.sc@web.de> - 2016-09-09 02:09 -0700
                              Re: ProDOS 2.4 Michael J. Mahon <mjmahon@aol.com> - 2016-09-07 19:33 -0500

Page 1 of 2  [1] 2  Next page →

#2959 — ProDOS 2.4

I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.

Notes on new features (1-6 are same as Beta1):
1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS clock, & GSOS quit
2) Apple ][, ][+, and un-enhanced //e compatible
3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][ & ][+
4) Greeting text and LC error text converts from lower to upper case if CPU==6502
5) Tested on emulators to run on ][+ and IIGS
6) Can be copied to system folder and renamed to P8 to use with GSOS
7) SystemDeath message system added at LC bank 2, $D001-$D04F
8) Minor size and space optimizations to fit SysDeath + 6502 without increasing the file size

Known issues:
-none-

-JB
Twitter: @JBrooksBSI

[toc] | [next] | [standalone]

#2965

John Brooks <jbrooks@blueshiftinc.com> wrote:
> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> 
> Notes on new features (1-6 are same as Beta1):
> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS clock, & GSOS quit
> 2) Apple ][, ][+, and un-enhanced //e compatible
> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][ & ][+
> 4) Greeting text and LC error text converts from lower to upper case if CPU==6502
> 5) Tested on emulators to run on ][+ and IIGS
> 6) Can be copied to system folder and renamed to P8 to use with GSOS
> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> 8) Minor size and space optimizations to fit SysDeath + 6502 without
> increasing the file size
> 
> Known issues:
> -none-
> 
> -JB
> Twitter: @JBrooksBSI
> 

Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
made that "force messages to upper case" should be an option separate from
the CPU option.
-- 
-michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com

[toc] | [prev] | [next] | [standalone]

#2969

On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> John Brooks wrote:
> > I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > 
> > Notes on new features (1-6 are same as Beta1):
> > 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS clock, & GSOS quit
> > 2) Apple ][, ][+, and un-enhanced //e compatible
> > 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][ & ][+
> > 4) Greeting text and LC error text converts from lower to upper case if CPU==6502
> > 5) Tested on emulators to run on ][+ and IIGS
> > 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > increasing the file size
> > 
> > Known issues:
> > -none-
> > 
> > -JB
> > Twitter: @JBrooksBSI
> > 
> 
> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> made that "force messages to upper case" should be an option separate from
> the CPU option.
> -- 
> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com

Agreed. I'm out of space at the moment though so a single test is controlling both uppercase boot text and forcing selection of the 40-column quit code as the 80-column quit code requires enhanced ROMs, 65c02, and a //e or newer keyboard.

I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it would work on ][, ][+, and un-enhanced //e, but that's a larger effort.

-JB

[toc] | [prev] | [next] | [standalone]

#2970

John Brooks <jbrooks@blueshiftinc.com> wrote:

> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> > John Brooks wrote:
> > > I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > > 
> > > Notes on new features (1-6 are same as Beta1):
> > > 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> > > clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> > > 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> > > & ][+
> > > 4) Greeting text and LC error text converts from lower to upper case
> > > if CPU==6502
> > > 5) Tested on emulators to run on ][+ and IIGS
> > > 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > > 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > > 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > > increasing the file size
> > > 
> > > Known issues:
> > > -none-
> > > 
> > > -JB
> > > Twitter: @JBrooksBSI
> > > 
> > 
> > Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> > made that "force messages to upper case" should be an option separate from
> > the CPU option.
> > -- 
> > -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> 
> Agreed. I'm out of space at the moment though so a single test is
> controlling both uppercase boot text and forcing selection of the
> 40-column quit code as the 80-column quit code requires enhanced ROMs,
> 65c02, and a //e or newer keyboard.
> 
> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.

I'd already done that a few years ago - my Quitter.System (written with
a friend, and he got permission from Randy Bird) includes a modified
Birds Better Bye which if I remember right works on a ][+. It added
several extra keys (e.g. to read the current drive again), and showed
two columns of files, still using 40-column mode.

I don't have time to dig it out at the moment, but once I set up my IIgs
again I can refresh my memory.

-- 
David Empson
dempson@actrix.gen.nz

[toc] | [prev] | [next] | [standalone]

#2972

On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> John Brooks wrote:
> 
> > On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> > > John Brooks wrote:
> > > > I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > > > 
> > > > Notes on new features (1-6 are same as Beta1):
> > > > 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> > > > clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> > > > 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> > > > & ][+
> > > > 4) Greeting text and LC error text converts from lower to upper case
> > > > if CPU==6502
> > > > 5) Tested on emulators to run on ][+ and IIGS
> > > > 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > > > 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > > > 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > > > increasing the file size
> > > > 
> > > > Known issues:
> > > > -none-
> > > > 
> > > > -JB
> > > > Twitter: @JBrooksBSI
> > > > 
> > > 
> > > Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> > > made that "force messages to upper case" should be an option separate from
> > > the CPU option.
> > > -- 
> > > -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> > 
> > Agreed. I'm out of space at the moment though so a single test is
> > controlling both uppercase boot text and forcing selection of the
> > 40-column quit code as the 80-column quit code requires enhanced ROMs,
> > 65c02, and a //e or newer keyboard.
> > 
> > I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> > would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> 
> I'd already done that a few years ago - my Quitter.System (written with
> a friend, and he got permission from Randy Bird) includes a modified
> Birds Better Bye which if I remember right works on a ][+. It added
> several extra keys (e.g. to read the current drive again), and showed
> two columns of files, still using 40-column mode.
> 
> I don't have time to dig it out at the moment, but once I set up my IIgs
> again I can refresh my memory.
> 
> -- 
> David Empson

That would be great! Did it fit in $300 bytes like Bird's Better Bye?

If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
 1) The GSOS quit
 2) The universal Better Bye quit

Doing this may also reduce the ProDOS file size by a block or two as a bonus.

-JB

[toc] | [prev] | [next] | [standalone]

#2973

On 8/18/2016 2:06 AM, John Brooks wrote:
> On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
>> John Brooks wrote:
>>
>>> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
>>>> John Brooks wrote:
>>>>> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
>>>>>
>>>>> Notes on new features (1-6 are same as Beta1):
>>>>> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
>>>>> clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
>>>>> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
>>>>> & ][+
>>>>> 4) Greeting text and LC error text converts from lower to upper case
>>>>> if CPU==6502
>>>>> 5) Tested on emulators to run on ][+ and IIGS
>>>>> 6) Can be copied to system folder and renamed to P8 to use with GSOS
>>>>> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
>>>>> 8) Minor size and space optimizations to fit SysDeath + 6502 without
>>>>> increasing the file size
>>>>>
>>>>> Known issues:
>>>>> -none-
>>>>>
>>>>> -JB
>>>>> Twitter: @JBrooksBSI
>>>>>
>>>>
>>>> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
>>>> made that "force messages to upper case" should be an option separate from
>>>> the CPU option.
>>>> --
>>>> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
>>>
>>> Agreed. I'm out of space at the moment though so a single test is
>>> controlling both uppercase boot text and forcing selection of the
>>> 40-column quit code as the 80-column quit code requires enhanced ROMs,
>>> 65c02, and a //e or newer keyboard.
>>>
>>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
>>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
>>
>> I'd already done that a few years ago - my Quitter.System (written with
>> a friend, and he got permission from Randy Bird) includes a modified
>> Birds Better Bye which if I remember right works on a ][+. It added
>> several extra keys (e.g. to read the current drive again), and showed
>> two columns of files, still using 40-column mode.
>>
>> I don't have time to dig it out at the moment, but once I set up my IIgs
>> again I can refresh my memory.
>>
>> --
>> David Empson
>
> That would be great! Did it fit in $300 bytes like Bird's Better Bye?
>
> If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
>  1) The GSOS quit
>  2) The universal Better Bye quit
>
> Doing this may also reduce the ProDOS file size by a block or two as a bonus.

That's the version that is on the ADTPro distribution disk's ProDOS 1.9 
code.  You could yank it from there.  Also:
http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq

[toc] | [prev] | [next] | [standalone]

#2974

On Thursday, August 18, 2016 at 4:16:25 AM UTC-7, schmidtd wrote:
> On 8/18/2016 2:06 AM, John Brooks wrote:
> > On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> >> John Brooks wrote:
> >>
> >>> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> >>>> John Brooks wrote:
> >>>>> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> >>>>>
> >>>>> Notes on new features (1-6 are same as Beta1):
> >>>>> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> >>>>> clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> >>>>> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> >>>>> & ][+
> >>>>> 4) Greeting text and LC error text converts from lower to upper case
> >>>>> if CPU==6502
> >>>>> 5) Tested on emulators to run on ][+ and IIGS
> >>>>> 6) Can be copied to system folder and renamed to P8 to use with GSOS
> >>>>> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> >>>>> 8) Minor size and space optimizations to fit SysDeath + 6502 without
> >>>>> increasing the file size
> >>>>>
> >>>>> Known issues:
> >>>>> -none-
> >>>>>
> >>>>> -JB
> >>>>> Twitter: @JBrooksBSI
> >>>>>
> >>>>
> >>>> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> >>>> made that "force messages to upper case" should be an option separate from
> >>>> the CPU option.
> >>>> --
> >>>> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> >>>
> >>> Agreed. I'm out of space at the moment though so a single test is
> >>> controlling both uppercase boot text and forcing selection of the
> >>> 40-column quit code as the 80-column quit code requires enhanced ROMs,
> >>> 65c02, and a //e or newer keyboard.
> >>>
> >>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> >>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> >>
> >> I'd already done that a few years ago - my Quitter.System (written with
> >> a friend, and he got permission from Randy Bird) includes a modified
> >> Birds Better Bye which if I remember right works on a ][+. It added
> >> several extra keys (e.g. to read the current drive again), and showed
> >> two columns of files, still using 40-column mode.
> >>
> >> I don't have time to dig it out at the moment, but once I set up my IIgs
> >> again I can refresh my memory.
> >>
> >> --
> >> David Empson
> >
> > That would be great! Did it fit in $300 bytes like Bird's Better Bye?
> >
> > If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
> >  1) The GSOS quit
> >  2) The universal Better Bye quit
> >
> > Doing this may also reduce the ProDOS file size by a block or two as a bonus.
> 
> That's the version that is on the ADTPro distribution disk's ProDOS 1.9 
> code.  You could yank it from there.  Also:
> http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq

I downloaded and looked at Quitter 4, but it is limited to 2 columns of 16 DIR/SYS files where the BetterBye in P8 supports scrolling through >256 DIR/SYS files.

Anyone know if a 40-col scrolling >256 file BetterBye has been made before?

-JB

[toc] | [prev] | [next] | [standalone]

#2976

John Brooks <jbrooks@blueshiftinc.com> wrote:

> On Thursday, August 18, 2016 at 4:16:25 AM UTC-7, schmidtd wrote:
> > On 8/18/2016 2:06 AM, John Brooks wrote:
> > > On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> > >> John Brooks wrote:
> > >>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> > >>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> > >>
> > >> I'd already done that a few years ago - my Quitter.System (written with
> > >> a friend, and he got permission from Randy Bird) includes a modified
> > >> Birds Better Bye which if I remember right works on a ][+. It added
> > >> several extra keys (e.g. to read the current drive again), and showed
> > >> two columns of files, still using 40-column mode.
> > >>
> > >> I don't have time to dig it out at the moment, but once I set up my IIgs
> > >> again I can refresh my memory.
> > >> David Empson
> > >
> > > That would be great! Did it fit in $300 bytes like Bird's Better Bye?

Yes. I rewrote some of the code to make it smaller, and added a few
features (details of which are vague about 25 years after I last worked
on it). Quitter works by replacing the in-memory ProDOS quit code, so
sticks within the $0300 size limit (it can also reinstate the GS/OS quit
code if you run Quitter again).

> > > If so, my vote would be to dump the simple ProDOS quit option and go
> > > from 3 Quit selectors to two:
> > >  1) The GSOS quit
> > >  2) The universal Better Bye quit
> > >
> > > Doing this may also reduce the ProDOS file size by a block or two as a
> > > bonus.
> > 
> > That's the version that is on the ADTPro distribution disk's ProDOS 1.9
> > code.  You could yank it from there.  Also:
>>http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq
> 
> I downloaded and looked at Quitter 4, but it is limited to 2 columns of 16
> DIR/SYS files where the BetterBye in P8 supports scrolling through >256
> DIR/SYS files.

I assume you mean the P8 2.0.x 80-column quit code (which I assume Apple
wrote - it looks quite different to Birds Better Bye).

It may be feasible to adapt my version of BBB to do paging instead of
two columns. I never needed more than two columns worth of SYS files
plus directories, but often exceeded the original one column limit.

> Anyone know if a 40-col scrolling >256 file BetterBye has been made
> before?

-- 
David Empson
dempson@actrix.gen.nz

[toc] | [prev] | [next] | [standalone]

#2977

On Thursday, August 18, 2016 at 8:24:28 AM UTC-6, John Brooks wrote:
> On Thursday, August 18, 2016 at 4:16:25 AM UTC-7, schmidtd wrote:
> > On 8/18/2016 2:06 AM, John Brooks wrote:
> > > On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> > >> John Brooks wrote:
> > >>
> > >>> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> > >>>> John Brooks wrote:
> > >>>>> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > >>>>>
> > >>>>> Notes on new features (1-6 are same as Beta1):
> > >>>>> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> > >>>>> clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> > >>>>> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> > >>>>> & ][+
> > >>>>> 4) Greeting text and LC error text converts from lower to upper case
> > >>>>> if CPU==6502
> > >>>>> 5) Tested on emulators to run on ][+ and IIGS
> > >>>>> 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > >>>>> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > >>>>> 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > >>>>> increasing the file size
> > >>>>>
> > >>>>> Known issues:
> > >>>>> -none-
> > >>>>>
> > >>>>> -JB
> > >>>>> Twitter: @JBrooksBSI
> > >>>>>
> > >>>>
> > >>>> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> > >>>> made that "force messages to upper case" should be an option separate from
> > >>>> the CPU option.
> > >>>> --
> > >>>> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> > >>>
> > >>> Agreed. I'm out of space at the moment though so a single test is
> > >>> controlling both uppercase boot text and forcing selection of the
> > >>> 40-column quit code as the 80-column quit code requires enhanced ROMs,
> > >>> 65c02, and a //e or newer keyboard.
> > >>>
> > >>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> > >>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> > >>
> > >> I'd already done that a few years ago - my Quitter.System (written with
> > >> a friend, and he got permission from Randy Bird) includes a modified
> > >> Birds Better Bye which if I remember right works on a ][+. It added
> > >> several extra keys (e.g. to read the current drive again), and showed
> > >> two columns of files, still using 40-column mode.
> > >>
> > >> I don't have time to dig it out at the moment, but once I set up my IIgs
> > >> again I can refresh my memory.
> > >>
> > >> --
> > >> David Empson
> > >
> > > That would be great! Did it fit in $300 bytes like Bird's Better Bye?
> > >
> > > If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
> > >  1) The GSOS quit
> > >  2) The universal Better Bye quit
> > >
> > > Doing this may also reduce the ProDOS file size by a block or two as a bonus.
> > 
> > That's the version that is on the ADTPro distribution disk's ProDOS 1.9 
> > code.  You could yank it from there.  Also:
> > http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq
> 
> I downloaded and looked at Quitter 4, but it is limited to 2 columns of 16 DIR/SYS files where the BetterBye in P8 supports scrolling through >256 DIR/SYS files.
> 
> Anyone know if a 40-col scrolling >256 file BetterBye has been made before?
> 
> -JB


I revised the Prodos 8 v1.9 bye selector to allow 64 pages of 16 files which is 1024 files, which I thought would take long enough to scroll.  This takes $4000 bytes of memory and can be expanded to $7000 bytes for 1792 files max.

The folder shows up as "XY" on non-mousetext computers and works with 40 or 80 columns.

[toc] | [prev] | [next] | [standalone]

#2980

On Thursday, August 18, 2016 at 10:33:33 PM UTC-7, gid...@sasktel.net wrote:
> On Thursday, August 18, 2016 at 8:24:28 AM UTC-6, John Brooks wrote:
> > On Thursday, August 18, 2016 at 4:16:25 AM UTC-7, schmidtd wrote:
> > > On 8/18/2016 2:06 AM, John Brooks wrote:
> > > > On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> > > >> John Brooks wrote:
> > > >>
> > > >>> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> > > >>>> John Brooks wrote:
> > > >>>>> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > > >>>>>
> > > >>>>> Notes on new features (1-6 are same as Beta1):
> > > >>>>> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> > > >>>>> clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> > > >>>>> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> > > >>>>> & ][+
> > > >>>>> 4) Greeting text and LC error text converts from lower to upper case
> > > >>>>> if CPU==6502
> > > >>>>> 5) Tested on emulators to run on ][+ and IIGS
> > > >>>>> 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > > >>>>> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > > >>>>> 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > > >>>>> increasing the file size
> > > >>>>>
> > > >>>>> Known issues:
> > > >>>>> -none-
> > > >>>>>
> > > >>>>> -JB
> > > >>>>> Twitter: @JBrooksBSI
> > > >>>>>
> > > >>>>
> > > >>>> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> > > >>>> made that "force messages to upper case" should be an option separate from
> > > >>>> the CPU option.
> > > >>>> --
> > > >>>> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> > > >>>
> > > >>> Agreed. I'm out of space at the moment though so a single test is
> > > >>> controlling both uppercase boot text and forcing selection of the
> > > >>> 40-column quit code as the 80-column quit code requires enhanced ROMs,
> > > >>> 65c02, and a //e or newer keyboard.
> > > >>>
> > > >>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> > > >>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> > > >>
> > > >> I'd already done that a few years ago - my Quitter.System (written with
> > > >> a friend, and he got permission from Randy Bird) includes a modified
> > > >> Birds Better Bye which if I remember right works on a ][+. It added
> > > >> several extra keys (e.g. to read the current drive again), and showed
> > > >> two columns of files, still using 40-column mode.
> > > >>
> > > >> I don't have time to dig it out at the moment, but once I set up my IIgs
> > > >> again I can refresh my memory.
> > > >>
> > > >> --
> > > >> David Empson
> > > >
> > > > That would be great! Did it fit in $300 bytes like Bird's Better Bye?
> > > >
> > > > If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
> > > >  1) The GSOS quit
> > > >  2) The universal Better Bye quit
> > > >
> > > > Doing this may also reduce the ProDOS file size by a block or two as a bonus.
> > > 
> > > That's the version that is on the ADTPro distribution disk's ProDOS 1.9 
> > > code.  You could yank it from there.  Also:
> > > http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq
> > 
> > I downloaded and looked at Quitter 4, but it is limited to 2 columns of 16 DIR/SYS files where the BetterBye in P8 supports scrolling through >256 DIR/SYS files.
> > 
> > Anyone know if a 40-col scrolling >256 file BetterBye has been made before?
> > 
> > -JB
> 
> 
> I revised the Prodos 8 v1.9 bye selector to allow 64 pages of 16 files which is 1024 files, which I thought would take long enough to scroll.  This takes $4000 bytes of memory and can be expanded to $7000 bytes for 1792 files max.
> 
> The folder shows up as "XY" on non-mousetext computers and works with 40 or 80 columns.

This sounds like the closest fit so far. Can you send me the source or a link?

I'm not sure what improvements Apple made to the original BBB, or what Apple changed between P8 1.x and 2.x, so there's some archeology to do along the way to make sure the new program selector handles errors and compatibility issues at least as well as P8 2.x.

-JB

[toc] | [prev] | [next] | [standalone]

#3037

On Friday, August 19, 2016 at 2:05:22 PM UTC-7, John Brooks wrote:
> On Thursday, August 18, 2016 at 10:33:33 PM UTC-7, gid...@sasktel.net wrote:
> > On Thursday, August 18, 2016 at 8:24:28 AM UTC-6, John Brooks wrote:
> > > On Thursday, August 18, 2016 at 4:16:25 AM UTC-7, schmidtd wrote:
> > > > On 8/18/2016 2:06 AM, John Brooks wrote:
> > > > > On Wednesday, August 17, 2016 at 10:47:33 PM UTC-7, David Empson wrote:
> > > > >> John Brooks wrote:
> > > > >>
> > > > >>> On Wednesday, August 17, 2016 at 3:53:17 PM UTC-7, mjm...@aol.com wrote:
> > > > >>>> John Brooks wrote:
> > > > >>>>> I built ProDOS 2.4 RC1 today and sent it to volunteers for testing.
> > > > >>>>>
> > > > >>>>> Notes on new features (1-6 are same as Beta1):
> > > > >>>>> 1) All ProDOS functions are now 6502 code except /RAM, 80 col quit, GS
> > > > >>>>> clock, & GSOS quit 2) Apple ][, ][+, and un-enhanced //e compatible
> > > > >>>>> 3) Fixed aux-mem test at 01/0400 which caused a flashing period on ][
> > > > >>>>> & ][+
> > > > >>>>> 4) Greeting text and LC error text converts from lower to upper case
> > > > >>>>> if CPU==6502
> > > > >>>>> 5) Tested on emulators to run on ][+ and IIGS
> > > > >>>>> 6) Can be copied to system folder and renamed to P8 to use with GSOS
> > > > >>>>> 7) SystemDeath message system added at LC bank 2, $D001-$D04F
> > > > >>>>> 8) Minor size and space optimizations to fit SysDeath + 6502 without
> > > > >>>>> increasing the file size
> > > > >>>>>
> > > > >>>>> Known issues:
> > > > >>>>> -none-
> > > > >>>>>
> > > > >>>>> -JB
> > > > >>>>> Twitter: @JBrooksBSI
> > > > >>>>>
> > > > >>>>
> > > > >>>> Since unenhanced //e's have 6502's, but are fully dual-case, a case can be
> > > > >>>> made that "force messages to upper case" should be an option separate from
> > > > >>>> the CPU option.
> > > > >>>> --
> > > > >>>> -michael - NadaNet 3.1 and AppleCrate II:  http://michaeljmahon.com
> > > > >>>
> > > > >>> Agreed. I'm out of space at the moment though so a single test is
> > > > >>> controlling both uppercase boot text and forcing selection of the
> > > > >>> 40-column quit code as the 80-column quit code requires enhanced ROMs,
> > > > >>> 65c02, and a //e or newer keyboard.
> > > > >>>
> > > > >>> I'd like to see the 'nice' 80-column quit (Better-Bye) rewritten so it
> > > > >>> would work on ][, ][+, and un-enhanced //e, but that's a larger effort.
> > > > >>
> > > > >> I'd already done that a few years ago - my Quitter.System (written with
> > > > >> a friend, and he got permission from Randy Bird) includes a modified
> > > > >> Birds Better Bye which if I remember right works on a ][+. It added
> > > > >> several extra keys (e.g. to read the current drive again), and showed
> > > > >> two columns of files, still using 40-column mode.
> > > > >>
> > > > >> I don't have time to dig it out at the moment, but once I set up my IIgs
> > > > >> again I can refresh my memory.
> > > > >>
> > > > >> --
> > > > >> David Empson
> > > > >
> > > > > That would be great! Did it fit in $300 bytes like Bird's Better Bye?
> > > > >
> > > > > If so, my vote would be to dump the simple ProDOS quit option and go from 3 Quit selectors to two:
> > > > >  1) The GSOS quit
> > > > >  2) The universal Better Bye quit
> > > > >
> > > > > Doing this may also reduce the ProDOS file size by a block or two as a bonus.
> > > > 
> > > > That's the version that is on the ADTPro distribution disk's ProDOS 1.9 
> > > > code.  You could yank it from there.  Also:
> > > > http://mirrors.apple2.org.za/ground.icaen.uiowa.edu/Mirrors/uni-kl/utilities/quitter.bsq
> > > 
> > > I downloaded and looked at Quitter 4, but it is limited to 2 columns of 16 DIR/SYS files where the BetterBye in P8 supports scrolling through >256 DIR/SYS files.
> > > 
> > > Anyone know if a 40-col scrolling >256 file BetterBye has been made before?
> > > 
> > > -JB
> > 
> > 
> > I revised the Prodos 8 v1.9 bye selector to allow 64 pages of 16 files which is 1024 files, which I thought would take long enough to scroll.  This takes $4000 bytes of memory and can be expanded to $7000 bytes for 1792 files max.
> > 
> > The folder shows up as "XY" on non-mousetext computers and works with 40 or 80 columns.
> 
> This sounds like the closest fit so far. Can you send me the source or a link?
> 
> I'm not sure what improvements Apple made to the original BBB, or what Apple changed between P8 1.x and 2.x, so there's some archeology to do along the way to make sure the new program selector handles errors and compatibility issues at least as well as P8 2.x.
> 
> -JB

While writing the new MLI $65 Quit command program launcher for ProDOS 2.4, I've discovered a couple things:

1) ProDOS has a bug where partial reads of directory files can return an EOF ($4C) error. Section 4.5.3 of The ProDOS 8 Technical Reference Manual states:

"The end of file error ($4C) is returned if and only if zero bytes were transferred (trans_count = 0)."

The above statement is true for non-directory files, but a read of a directory will return EOF for partial reads if the read request would extend past the last byte of the last block of the directory.


2) Copy II Plus overwrites the ProDOS $EE ID byte at $D000 LC bank 2 along with the rest of that page. Copy II uses LC Bank 2 $D000-$D0FF to hold a 'No DOS on disk" message which is written to T0,S0 if/when a 5.25 disk is formatted for DOS 3.3 using Copy II.

This bank 2 $D000-$D0FF use not only changes the ProDOS LC bank 2 ID byte from $EE to $01, which is nonfatal, but it also overwrites the new ProDOS 2.4 System Death handler and program launcher data.

To workaround this conflict between Copy II Plus and P8 2.4, I am patching CopyII Plus so it does not write to LC Bank 2 $D000-$D0FF (a one byte patch in the first page of UTIL.SYSTEM).

Without the patch, Copy II Plus will hang if there is a ProDOS SysDeath call, and the quit program selector will not work, requiring a reboot.

If anyone knows of other programs which use $D000-$D0FF of LC bank 2, I'd like to know.

Thanks!

-JB
Twitter: @JBrooksBSI

[toc] | [prev] | [next] | [standalone]

#3039

On 9/5/2016 9:42 PM, John Brooks wrote:
>
> 2) Copy II Plus overwrites the ProDOS $EE ID byte at $D000 LC bank 2 along with the rest of that page. Copy II uses LC Bank 2 $D000-$D0FF to hold a 'No DOS on disk" message which is written to T0,S0 if/when a 5.25 disk is formatted for DOS 3.3 using Copy II.
>
> This bank 2 $D000-$D0FF use not only changes the ProDOS LC bank 2 ID byte from $EE to $01, which is nonfatal, but it also overwrites the new ProDOS 2.4 System Death handler and program launcher data.
>
> To workaround this conflict between Copy II Plus and P8 2.4, I am patching CopyII Plus so it does not write to LC Bank 2 $D000-$D0FF (a one byte patch in the first page of UTIL.SYSTEM).
>
> Without the patch, Copy II Plus will hang if there is a ProDOS SysDeath call, and the quit program selector will not work, requiring a reboot.
>
> If anyone knows of other programs which use $D000-$D0FF of LC bank 2, I'd like to know.
>
> Thanks!
>
> -JB
> Twitter: @JBrooksBSI
>

John,

AppleWorks uses $D000-$D0FF of the second language card bank in main 
memory for what it calls its 'sub-host' routines. {Actually, it uses all 
the way up to $DFFF}.

I know that upon startup, AppleWorks saves at least a portion of the 
$D000-$DFFF in its 'Desktop Memory', and upon quitting, it copies it 
back to its original location.

What I don't know (without doing a little disassembly), is how much of 
that area is saved and restored. It might _not_ be the whole 4K.

I should have some time a little later this week if you'd like for me to 
find out.

BTW, nice work with this '2.4'. I've enjoyed following it.





Hugh Hood

[toc] | [prev] | [next] | [standalone]

#3043

On 9/5/2016 10:12 PM, Hugh Hood wrote:
>
> AppleWorks uses $D000-$D0FF of the second language card bank in main
> memory for what it calls its 'sub-host' routines. {Actually, it uses all
> the way up to $DFFF}.
>
> I know that upon startup, AppleWorks saves at least a portion of the
> $D000-$DFFF in its 'Desktop Memory', and upon quitting, it copies it
> back to its original location.
>
> What I don't know (without doing a little disassembly), is how much of
> that area is saved and restored. It might _not_ be the whole 4K.
>

John,

OK, I fired up Ewen's BrkDown and checked. AppleWorks on startup is 
saving (and then restoring upon exit) the $300 bytes from $D100-$D3FF in 
LC Bank 2. The rest of $D0xx is also used, but neither saved nor 
restored. I'm pretty sure that AppleWorks could be modified without much 
problem to do that, if it's important.

FWIW, earlier versions of AppleWorks saved/restored the ProDOS quit code 
in LC Bank 2 to/from disk, and then read it back from disk and restored 
it upon exit. The later versions just saved/restored it to/from 'Desktop 
Memory' rather than to disk.

I hope this helps you.





Hugh Hood

[toc] | [prev] | [next] | [standalone]

#3045

On Monday, September 5, 2016 at 10:05:49 PM UTC-7, Hugh Hood wrote:
> On 9/5/2016 10:12 PM, Hugh Hood wrote:
> >
> > AppleWorks uses $D000-$D0FF of the second language card bank in main
> > memory for what it calls its 'sub-host' routines. {Actually, it uses all
> > the way up to $DFFF}.
> >
> > I know that upon startup, AppleWorks saves at least a portion of the
> > $D000-$DFFF in its 'Desktop Memory', and upon quitting, it copies it
> > back to its original location.
> >
> > What I don't know (without doing a little disassembly), is how much of
> > that area is saved and restored. It might _not_ be the whole 4K.
> >
> 
> John,
> 
> OK, I fired up Ewen's BrkDown and checked. AppleWorks on startup is 
> saving (and then restoring upon exit) the $300 bytes from $D100-$D3FF in 
> LC Bank 2. The rest of $D0xx is also used, but neither saved nor 
> restored. I'm pretty sure that AppleWorks could be modified without much 
> problem to do that, if it's important.
> 
> FWIW, earlier versions of AppleWorks saved/restored the ProDOS quit code 
> in LC Bank 2 to/from disk, and then read it back from disk and restored 
> it upon exit. The later versions just saved/restored it to/from 'Desktop 
> Memory' rather than to disk.
> 
> I hope this helps you.
> 
> 
> 
> 
> 
> Hugh Hood

Thanks Hugh.

In the 2.4 testing done so far, it appears that the vast majority of ProDOS programs do not use LC Bank 2 $D000 page, but the few that do are popular (Copy II, Appleworks, ProSel's Cat Doctor).

While patching a handful of apps is a viable option for P8 2.4, and IMO, a worthwhile trade to get a universal improved ProDOS, it's definitely not ideal. I think I'll take another look through the ProDOS memory map and see if I can find another way/place for the larger 6502 code/data that would have less impact on applications.

One possible option is to allow ProDOS to load code/data from the boot volume after startup, but this comes with it's own tradeoffs.

I suspect that future use of ProDOS will be largely on emulators with lots of drives attached/available (ie, Smartport remapping feature is high-value and boot HD partition is likely), or for broad compatibility (6502 ][,][+, //e) for universal apps like ADT Pro or retro development.

While drive remap and universal ProDOS have gained in value, IMO the following features are less important now than when ProDOS was first authored:

1) 4x user-installed IRQ handlers managed by ProDOS
2) SysDeath pretty-print for unclaimed IRQs (and corrupt ProDOS buffers)
3) 8x files open at once across 8x devices
4) Apple II workstation card for Appletalk support
5) /RAM bells & whistles (Ramworks, Ramfactor, and IIGS /RAM5 are more useful/effective today)
6) Disk II driver bells & whistles (5.25 is now either the primary device, or is seldom used)

I am interested in feedback from other Apple II folks.

Thoughts?

-JB
Twitter: @JBrooksBSI

[toc] | [prev] | [next] | [standalone]

#3046

On Tuesday, September 6, 2016 at 8:40:24 AM UTC-7, John Brooks wrote:
> While drive remap and universal ProDOS have gained in value, IMO the following features are less important now than when ProDOS was first authored:
> 
> 1) 4x user-installed IRQ handlers managed by ProDOS

What's the common use case? 

i.e.
a) Whats the average # of interrupts people are using?
b) Whats the maximum # of interrupts in practice?

> 2) SysDeath pretty-print for unclaimed IRQs (and corrupt ProDOS buffers)

Is this really needed?

> 3) 8x files open at once across 8x devices

What's the common  use case?

i.e.
a)  What's the average # of files open people use?
b) What's the maximum # of files open in practice?


Without more meta-data and analytics it will be hard to determine what the right course of design & implementation is, IMHO.

[toc] | [prev] | [next] | [standalone]

#3048

On Tuesday, September 6, 2016 at 10:37:57 AM UTC-6, Michael Pohoreski wrote:
> On Tuesday, September 6, 2016 at 8:40:24 AM UTC-7, John Brooks wrote:
> > While drive remap and universal ProDOS have gained in value, IMO the following features are less important now than when ProDOS was first authored:
> > 
> > 1) 4x user-installed IRQ handlers managed by ProDOS
> 
> What's the common use case? 
> 
> i.e.
> a) Whats the average # of interrupts people are using?
> b) Whats the maximum # of interrupts in practice?
> 
> > 2) SysDeath pretty-print for unclaimed IRQs (and corrupt ProDOS buffers)
> 
> Is this really needed?
> 
> > 3) 8x files open at once across 8x devices
> 
> What's the common  use case?
> 
> i.e.
> a)  What's the average # of files open people use?
> b) What's the maximum # of files open in practice?
> 
> 
> Without more meta-data and analytics it will be hard to determine what the right course of design & implementation is, IMHO.



I don't think this kind of thing should be considered as common use, but rather future expansion.

Sample code has already been written for multitasking, and all that is needed is a couple of good programmers to expand upon it.

Ergo, we should be asking what kind of things would we want to see in Prodos that would allow us to, say, write a desktop operating system.

If someone were to take on the task of writing a multitasking OS, then yeah, we would need 4x IRQ's and 8 open files at once.

removing these from Prodos would greatly hinder someone wanting to write a powerful program.

I wrote a copy program, for personal use, that uses 3 open files.  One for the source directory, one for destination directory and one for the file to be copied.

But I still dream of the day I retire so I can write a full double-click OS under Prodos 8.

But for now, that is still just a dream.

[toc] | [prev] | [next] | [standalone]

#3049

On Tuesday, 6 September 2016 08:40:24 UTC-7, John Brooks  wrote:
> On Monday, September 5, 2016 at 10:05:49 PM UTC-7, Hugh Hood wrote:
> > On 9/5/2016 10:12 PM, Hugh Hood wrote:
> > >
> > > AppleWorks uses $D000-$D0FF of the second language card bank in main
> > > memory for what it calls its 'sub-host' routines. {Actually, it uses all
> > > the way up to $DFFF}.
> > >
> > > I know that upon startup, AppleWorks saves at least a portion of the
> > > $D000-$DFFF in its 'Desktop Memory', and upon quitting, it copies it
> > > back to its original location.
> > >
> > > What I don't know (without doing a little disassembly), is how much of
> > > that area is saved and restored. It might _not_ be the whole 4K.
> > >
> > 
> > John,
> > 
> > OK, I fired up Ewen's BrkDown and checked. AppleWorks on startup is 
> > saving (and then restoring upon exit) the $300 bytes from $D100-$D3FF in 
> > LC Bank 2. The rest of $D0xx is also used, but neither saved nor 
> > restored. I'm pretty sure that AppleWorks could be modified without much 
> > problem to do that, if it's important.
> > 
> > FWIW, earlier versions of AppleWorks saved/restored the ProDOS quit code 
> > in LC Bank 2 to/from disk, and then read it back from disk and restored 
> > it upon exit. The later versions just saved/restored it to/from 'Desktop 
> > Memory' rather than to disk.
> > 
> > I hope this helps you.
> > 
> > 
> > 
> > 
> > 
> > Hugh Hood
> 
> Thanks Hugh.
> 
> In the 2.4 testing done so far, it appears that the vast majority of ProDOS programs do not use LC Bank 2 $D000 page, but the few that do are popular (Copy II, Appleworks, ProSel's Cat Doctor).
> 
> While patching a handful of apps is a viable option for P8 2.4, and IMO, a worthwhile trade to get a universal improved ProDOS, it's definitely not ideal. I think I'll take another look through the ProDOS memory map and see if I can find another way/place for the larger 6502 code/data that would have less impact on applications.
> 
> One possible option is to allow ProDOS to load code/data from the boot volume after startup, but this comes with it's own tradeoffs.
> 
> I suspect that future use of ProDOS will be largely on emulators with lots of drives attached/available (ie, Smartport remapping feature is high-value and boot HD partition is likely), or for broad compatibility (6502 ][,][+, //e) for universal apps like ADT Pro or retro development.
> 
> While drive remap and universal ProDOS have gained in value, IMO the following features are less important now than when ProDOS was first authored:
> 
> 1) 4x user-installed IRQ handlers managed by ProDOS
> 2) SysDeath pretty-print for unclaimed IRQs (and corrupt ProDOS buffers)
> 3) 8x files open at once across 8x devices
> 4) Apple II workstation card for Appletalk support
> 5) /RAM bells & whistles (Ramworks, Ramfactor, and IIGS /RAM5 are more useful/effective today)
> 6) Disk II driver bells & whistles (5.25 is now either the primary device, or is seldom used)
> 
> I am interested in feedback from other Apple II folks.
> 
> Thoughts?
> 
> -JB
> Twitter: @JBrooksBSI

Hi John-

I test PLASMA, which install the VM in the $D100-$DFFF range and overwrites the 'BYE' code with its own. All looks good, even on my ][+. Kudos, a great job.

As for your list of useful features, I still have two machines with Workstation cards. And I even use them, once in a while. But, if it must go, I can always keep 2.03 on those machines.

Dave...

[toc] | [prev] | [next] | [standalone]

#3050

Hi John,

> > FWIW, earlier versions of AppleWorks saved/restored the ProDOS quit code 
> > in LC Bank 2 to/from disk, and then read it back from disk and restored 
> > it upon exit. The later versions just saved/restored it to/from 'Desktop 
> > Memory' rather than to disk.

Not wanting to appear as the smartass that's basically what I wrote before.
That approach was just too tempting to not have been taken by application
developers. Especially as calling MLI with LC bank 2 enabled works just fine -
and even returns with LC bank 2 re-enabled.

> While patching a handful of apps is a viable option for P8 2.4, and IMO, a worthwhile trade to get a universal improved ProDOS, it's definitely not ideal. 

Full ACK.

> While drive remap and universal ProDOS have gained in value, IMO the following features are less important now than when ProDOS was first authored:
> 
> 1) 4x user-installed IRQ handlers managed by ProDOS

My guess would be that two (instead of four) are fine.

> 2) SysDeath pretty-print for unclaimed IRQs (and corrupt ProDOS buffers)

It certianly doesn't have to be pretty but the situation should still become visible to the user.

> 3) 8x files open at once across 8x devices

I certainly wouldn't reduce that.

> 4) Apple II workstation card for Appletalk support

I wasn't aware that there's any Appletalk support inside ProDOS. Would you mind to elaborate?

> 5) /RAM bells & whistles (Ramworks, Ramfactor, and IIGS /RAM5 are more useful/effective today)

As already written in quite some detail an existing and empty /RAM signals to
a cleanly written application that it can make use of AUXMEM (by disconnecting
/RAM). If you remove /RAM support you'll find yourself breaking applications -
just like you did with using the first LC bank 2 page.

> 6) Disk II driver bells & whistles (5.25 is now either the primary device, or is seldom used)

If you were about to create a ProDOS variant targeting "uplevel" machines like
the IIgs I would maybe following you here. But as far as I understand you want
to bring the fully functional device mirroring of ProDOS 2.0.3 to "downlevel"
machines without 65C02. For those machines Disk II drives will forever stay the
natural and primary mass storage device.

Apart from that - although you may of course consider it as not relevant -
reusing the Disk II driver area would for sure break
http://adtpro.sourceforge.net/vdrive.html - saying:

"There are two versions of the serial driver: VSDRIVE.LOW and VSDRIVE. The
VSDRIVE.LOW version will install itself in the main memory of the Apple II,
reducing the amount of memory available to BASIC and other programs. The ProDOS
Utilities Disk will not work with this version of the driver. The VSDRIVE
version will instead overwrite the Disk II driver, making Disk II floppies
inaccessible to ProDOS. The ProDOS Utilities Disk is compatible with this
version, and it will stay resident even when the ProDOS QUIT code is active."

I'm BTW about to create the very same thing but using the Uthernet II card
instead of a SSC...

> I am interested in feedback from other Apple II folks.

As already written I'd look into removing:

1) Writing sparse files. Already existing spares files should be read correctly.

2) Converting storage types "back" when shrinking files.

Both changes should not cause "direct" incompatibilies but rather "just" waste disk space - for in my gut feeling rare scenarios.

3) True /RAM support. However a "pseudo /RAM" is necessary for AUXMEM
management - as described.

Just my two cents,
Oliver

[toc] | [prev] | [next] | [standalone]

#3051

On Wednesday, September 7, 2016 at 3:23:55 AM UTC-7, Oliver Schmidt wrote:
> As already written I'd look into removing:
> 
> 1) Writing sparse files. Already existing spares files should be read correctly.

Does anyone know what applications actually used sparse files?

[toc] | [prev] | [next] | [standalone]

#3052

On 2016-09-07 20:37:28 +0000, Michael Pohoreski said:

> On Wednesday, September 7, 2016 at 3:23:55 AM UTC-7, Oliver Schmidt wrote:
>> As already written I'd look into removing:
>> 
>> 1) Writing sparse files. Already existing spares files should be read 
>> correctly.
> 
> Does anyone know what applications actually used sparse files?

Davex Disk Images do.

> The first 512 bytes of a Davex archived volume are a header, 
described under "File Format" in this Note. After the header comes 512 
bytes for each block on the saved volume, from zero on up. For blocks 
that are unused, you can just set the file mark ahead 512 bytes instead 
of writing 512 zero bytes--this way the unused blocks do not take up 
disk space, so the resulting file is only a few blocks larger than the 
number of used blocks on the original volume.

> This file format is suitable for ProDOS, but it is less useful on a 
file system (such as AppleShare) that does not allow for sparse files. 
A Davex archived volume file on an AppleShare server always takes up 
more blocks than the original volume contains.

Anyhow, I was under the impression that sparsification only happens 
when gaps are created with setmark or seteof. Removing the sparse code 
and zero-filling those blocks isn't necessarily an improvement.

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | comp.sys.apple2.programmer

csiph-web