RE: [N8VEM: 1461] Re: CP/M for the SBC

["Andrew Lynch" <lyn...@yahoo.com>]

Thanks Bill!  I looked it over and it is quite nice!  

I think we really need a common N8VEM and mini-N8VEM SBC default software
release.  James has made lots of great improvements already and it is a
logical starting point at least for me.  Especially since the default XMODEM
in the 23 May 2008 release is just plain broken on many SBCs.

James, would you like to give this CCP a try on your mini-N8VEM?

This is very encouraging!  We need some pot stirring to get the ideas
churning!

Thanks and have a nice day!

Andrew Lynch
 


> -----Original Message-----
> From: n8...@googlegroups.com [mailto:n8...@googlegroups.com] On Behalf Of
> Bill Beech (NJ7P)
> Sent: Sunday, December 21, 2008 10:20 PM
> To: n8...@googlegroups.com
> Subject: [N8VEM: 1461] Re: CP/M for the SBC
> 
> 
> Andrew,
> 
> I shipped you my "nearer to original" CCP.asm file directly.  Feel free
> to look it over and compare it with the CCP source in the current load
> distro.
> 
> The movcpm that shipped with the JadeDD controller did work, but I much
> rather just change the CPM size and reassemble the source.
> 
> I agree a tool that would just build the ROM image is the way to go.  I
> have downloaded but not looked at the exchange software for Linux.  I
> hope that takes us part of the way there.
> 
> Can't wait for the parts to get here!
> 
> Bill
> 
> Andrew Lynch wrote:
> >
> >
> >> -----Original Message-----
> >> From: n8...@googlegroups.com [mailto:n8...@googlegroups.com] On Behalf
> Of
> >> Bill Beech (NJ7P)
> >> Sent: Sunday, December 21, 2008 4:01 PM
> >> To: n8...@googlegroups.com
> >> Subject: [N8VEM: 1452] Re: CP/M for the SBC
> >>
> >>
> >>
> >>
> >> Andrew Lynch wrote:
> >>
> >>>
> >>>> -----Original Message-----
> >>>> From: n8...@googlegroups.com [mailto:n8...@googlegroups.com] On
> Behalf
> >>>>
> >> Of
> >>
> >>>> James Moxham (Dr_Acula)
> >>>> Sent: Sunday, December 21, 2008 12:19 AM
> >>>> To: N8VEM
> >>>> Subject: [N8VEM: 1424] Re: CP/M for the SBC
> >>>>
> >>>>
> >>>> Eg N8VEM SW003 from May 23rd.
> >>>>
> >>>> The further back you go in the files archive - the more "pure" the
> >>>> version of CP/M becomes.
> >>>>
> >>>> It depends on what you want to do with it. Bits I've added include
> >>>> using a 27C256 instead of a big eprom, swapping drive A and B,
> putting
> >>>> a format program on the eprom instead of other programs, putting a
> >>>> different version of xmodem on the eprom, adding calls in cp/m to
> >>>> drive a keyboard and 20x4 LCD display, making all the opcodes Z80
> >>>> instead of 8080 and bypassing a call to an IDE routine that isn't
> >>>> actually implemented yet but which was making a warm boot last much
> >>>> longer than it should have.
> >>>>
> >>>> I suspect all these things breach the sanctity of CP/M with some
> awful
> >>>> kludges.
> >>>>
> >>>> On the other hand, I can start soldering up a board from scratch,
> >>>> finish soldering it 35 minutes later, and within just 5 minutes have
> >>>> wordstar and mbasic running on the board.
> >>>> [AJL>]
> >>>>
> >>>>
> >>> Hi James! Thanks!  During development of peripherals, which you've
> been
> >>> doing a great job at, the software has to get tweaked and changed to
> >>>
> >> make
> >>
> >>> things work.  That is expected and the whole intent of the software
> >>>
> >> release.
> >>
> >>> There really is no software standard as that is up to the individual
> >>>
> >> builder
> >>
> >>> to decide what they want in their environment.
> >>>
> >>> I think we should use your improved software baseline as the new
> default
> >>> software.  It will probably have to go through some additional tweaks
> as
> >>>
> >> we
> >>
> >>> go.  Certainly the Disk IO support for IDE and eventually FDC will
> have
> >>>
> >> to
> >>
> >>> be functioning in the CBIOS.  With lots of builders digging into the
> >>> software there is no doubt we'll find bugs like the XMODEM UART
> problem
> >>>
> >> that
> >>
> >>> plagued us early on.  I'm sure there are more bugs waiting to be
> >>>
> >> discovered
> >>
> >>> too.
> >>>
> >>> Ideally any builder customization to CP/M ought to be restricted to
> the
> >>> CBIOS code.  Is there some way to tweak the new baseline to use the
> >>>
> >> default
> >>
> >>> CCP and BDOS code from the original source?  If it has to run at boot
> >>>
> >> up,
> >>
> >>> the "BOOT" and/or "WBOOT" CBIOS routines might be a suitable home.
> >>>
> >> Probably
> >>
> >>> the CBIOS code should be separated from the CCP and BDOS code.
> >>>
> >> Combining
> >>
> >>> CCP+BDOS+CBIOS was just an ugly hack I made early during TestPrototype
> >>> development.
> >>>
> >>> http://www.cpm.z80.de/download/cpm2-asm.zip
> >>>
> >>>
> >> Andrew,
> >>
> >> This "original source" has about 512 bytes of code near the front of
> the
> >> CCP that is NOT in the original CPM 2.2 distribution.  That is what I
> >> was going to replace with a CCP assembler code I have here.  I can
> build
> >> it up, get it working, then load it up for you to say yea or nae to.
> It
> >> will give us more BIOS room, which we will soon be needing...
> >>
> >> Bill
> >>
> >
> > [AJL>]
> >
> > Hi Bill!  That source is probably not the actual original CP/M 2.2
> source as
> > I think CP/M was written by DRI in PL/M and compiled on a VAX if I
> recall
> > correctly.  That source is a "reconstructed" or reverse engineered from
> a
> > loaded binary image I suppose.  Someone got out the disassembler!
> >
> > What is at the beginning of the source that is not original?  Has it
> been
> > patched by someone?  What does it do?  It looked like a bunch of utility
> > routines when I looked at it and I did not realize it was a modified
> CP/M
> > source.  I learn something interesting everyday!
> >
> > If the CP/M CCP+BDOS source can be reduced in size that would be good.
> I
> > did not mess with it much though as I was more interested in getting the
> > CBIOS to work and once that happened I just left it alone.
> >
> > The nice thing about the assembler source is that it is relatively easy
> with
> > a small patch to the origin and move CCP+BDOS just like MOVCPM does.
> > Unfortunately MOVCPM only works for like one computer (MDS-800, I think)
> and
> > being able to move CP/M in memory is handy when you have a growing or
> > shrinking CBIOS.
> >
> > What would be neat though would be a builder customizable CP/M binary
> ROM
> > image generator tool.  Run a menu and select which IO you have, which
> SBC
> > you are running, what ROMs, etc and it sets the conditional assembly
> flags
> > and then a fully custom CP/M CBIOS is created.  That sounds like a
> fantasy
> > but it is not too far fetched.
> >
> > Thanks and have a nice day!
> >
> > Andrew Lynch
> >
> > [snip]
> >
> >
> > >
> >
> >
> 
> 
>