Re: Apple II IDE Interface software : your opinions and suggestions welcome

["Stephane Guillard" <stephane.guillard@steria.com>]

Hi,

After your overwhelming amount of suggestions (:-) I'm going to
finally make things the way I suggested. I will have a ROM soon, maybe
autobooting if I manage to cram all the vitals in 2,2K...

Anyone out there finally built the interface ?


> Hi,
> 
> Those who are interested in the latest release of the software for
my
> IDE interface can go to http://s.guillard.free.fr where in the Apple
> II IDE page there is now a link to C source and 4 binary releases
> (including various levels of functionalities).
> 
> The biggest binary will enable navigating in the FAT dir tree,
create
> / delete files, change their attribute, view their contents, write
> into them etc. (it's funny to actually manipulate a volume created
on
> the PC, add files etc. and then plug the HDD in a PC and see all has
> worked properly :-).
> 
> I've done 50% of the work to port back C functions to assembly (for
> size and speed). Please refer to the main() source in the C file to
> know which functionalities are available in the various binaries
> (which are only recompliations of the same project, with various
> defines).
> 
> Now I wonder how to split things.
> 
> The functionalities could be split this way :
> 
> - minimal feature set goes to ROM. Able to list / navigate dir
> tree, load / save / execute a binary file. This is sufficient to
store
> something to the hard disk, and run it. It enables loading something
> from a DOS3.3 or ProDOS floppy and store it to the hard disk, but if
> this software calls DOS3.3 or ProDOS, its calls wil lgo to floppy,
not
> hard disk (thus the patches mentioned below)
> 
> - RAM-loadable extension API (mainly the POSIX-like API already
> done, which includes the full fcntl set : open(), creat(), read(),
> write() and more, see in the C source), which is a necessary support
> layer for the next 2 parts
> 
> - external command set, loaded from the hard disk itself (rm, mkdir,
rmdir, ls, cp, mv, cat, od, etc.),
> relying on the extension API
> 
> - patches for the most used features of DOS3.3 and ProDOS, so that
> software saved to the hard disk, which manipulate files through DOS
or
> ProDOS, can be completely operated from the hard disk.
> 
> I will anyway split the software in the following parts in the next
> days, and this is where I would appreciate your opinions :
> 
> - ROM part. I can use 2 K + 256 bytes.
> 
> - relocatable residentable RAM part. How much RAM can be eaten by
this
> one so that it does not interfer too much with DOS3.3 ? with ProDOS
?
> Which base addresses would be nice under DOS3.3 ? ProDOS ?
> 
> - DOS 3.3 and ProDOS patches. Various options could be done here,
from
> having a patched DOS 3.3 or ProDOS kernel which loads from the hard
> disk, but which is not able anymore to operate with floppies
> (easiest), to patching these in such a way that the hard disk
> operating system intercepts the calls to DOS / ProDOS, recognizes
> something in the file name which indicates hard disk, and subsequent
> calls to this file are redirected to HDD etc. What would you
> appreciate / require / suggest ?
> 
> Regarding the external command set, there is no restriction because
> they will be ordinary executables relying on the extension API.
> 
> Thank you for your
> answers, which should be made public in this newsgroup so that we
can
> consolidate into a specification which suits all needs.
> 
> Regards,
> 
> -- 
> St�phane GUILLARD  - stephane.guillard@steria.com
> T(+33)(0) 472 523 540 - F(+33)(0) 472 523 555 - M(+33)(0) 609 417
840
> A1K1.3/8M+100M
> A2KB3.9Ecs2M/GVP-M68060-60+64M+4.3G+32x/CV64-3D+SD/A2065
>
A4K3.9/CSPPC233-060+CVPPC+128M+4.3G+32x/PicassoIV-Concierto-Paloma/AriadneII
> 

-- 
St�phane GUILLARD  - stephane.guillard@steria.com
T(+33)(0) 472 523 540 - F(+33)(0) 472 523 555 - M(+33)(0) 609 417 840
A1K1.3/8M+100M
A2KB3.9Ecs2M/GVP-M68060-60+64M+4.3G+32x/CV64-3D+SD/A2065
A4K3.9/CSPPC233-060+CVPPC+128M+4.3G+32x/PicassoIV-Concierto-Paloma/AriadneII