What are your obstacles to running CP/M? - CP/M
This is a discussion on What are your obstacles to running CP/M? - CP/M ; On Apr 5, 7:39 am, Bill H wrote:
> On Apr 4, 11:51 pm, Bill wrote:
>
>
>
> > On Fri, 04 Apr 2008 23:32:01 GMT, no.s...@no.uce.bellatlantic.net
> > wrote:
>
> > > on any given day ...
-
Re: What are your obstacles to running CP/M?
On Apr 5, 7:39 am, Bill H wrote:
> On Apr 4, 11:51 pm, Bill wrote:
>
>
>
> > On Fri, 04 Apr 2008 23:32:01 GMT, no.s...@no.uce.bellatlantic.net
> > wrote:
>
> > > on any given day there are more people here than either ran, run
> > >or have built systems around CP/M then in 1979.
>
> > There are many times more people running Symbian than the
> > total who ever ran CP/M. Nokia gives away really nice SDK's
> > and encourages people to program applications. Sony and
> > others also build Symbian devices.
>
> > There's music players, video recorders, cameras, GPS, plus
> > the fact that wireless connection and bluetooth is common,
> > and most have USB. Word processing with a telephone
> > pad turned out to be a nuisance so most later generations
> > either have a keyboard somewhere (flip, twist, slide) or allow
> > using a USB or Bluetooth one. Some people are putting movies
> > in theirs. Basically, pocket computers WAY more powerful
> > than anything we had running CP/M 'in the day'.
>
> > Until a few weeks ago, I knew NOTHING about it.
>
> > Then, I bought a Nokia 6620 in an auction. For a BUCK!
>
> > You want something to hack on? Hack Symbian! You can
> > carry the results around in your pocket!
>
> > Bill
>
> This is comparing apples and oranges though. Getting a CP/M system up
> and running would be akin to completly rebuilding a Ford Model T from
> boxes of parts versus Tricking out a 2008 Ford Focus (ie getting
> Symbian runnning).
>
> Allison's thoughts on picking a specific known system to start out
> with is a good one. I could picture a web page that says:
>
> "So you want to make a CP/M system?"
>
> And then it goes on to list some well known, well behaved systems and
> the components you should look for, almost a check list.
Ok, I want a CP/M system with a Z80, 64K RAM, 1M serial flash RAM as
boot disk, and four DB9-M connectors wired to accept little dongles
that are accessed on a SPI interface. Dongles for USB, an SD socket,
an RS232C UART, a math co-processor, and an A/V video display (OK,
two ... NTSC and PAL). And for the front two of the DB-9 plugs, if you
plug in an old-fashioned one button joystick and push the button, the
system can read that a joystick is attached instead and read it like a
joystick.
That would make a great little computer laboratory to play around
with.
I'm not 100% sure I want to make it, but if I can't make that, I
likely can't make anything.
-
Re: What are your obstacles to running CP/M?
no.spam@no.uce.bellatlantic.net schrieb:
> I've always maintaind that for getting to bare metal CP/M is it.
If you want to get at the bare metal of a Sun Microsystems Sunfire
mainframe CP/M is it not. If you have learned Forth before, maybe on a
CP/M system, that would help, because the firmware to get at the bare
metal of this mainframes is written in Forth. One example of many.
> Also while Linux both popular and useful it's also big wher CP/M BDOS
> source is around 140K source and assembles to around 5.5k the whole
> OS (CCP, BDOS, standard style BIOS) binary is around 8K total and the
> source is maybe 250-300K. It's one of the few OSs out there that
> every one that wants to can wrap their minds around the entire source
> line by line. You can see the hardware and software interaction and
Read newsgroups for Linux, BSD, Minix systems. There are nerds who can
read the whole kernel sources to you. Of course it is much easier to
handle the few lines of a CP/M kernel. That is good for total beginners
who would like to learn about OS's. After a few weeks you move on to
bigger and maybe better things.
> also tradeoffs made in to make CP/M more portable. This is very
> useful and instructive as everthing needed to understand OSs are there
> (even if elementary level). There is useful software for it
> including all of the basic tools one would expect and an assortment of
> languages and user applications as well. It's old but it still can do
> useful work.
Yep, there is tons of software for this old systems and if you look into
it, you'll find that not that much has changed on modern systems. This
old software is a good base to start learning programming, because it's
easier to understand than newer, more complex systems.
Usefull work? OK, I would like to write a simple letter, just a few
lines of text, nothing fancy. So I start some ancient CP/M machine and
type the few lines with ed and format that manually, no problem. Now I
want to print that letter. No printer with serial or centronics ports
available since ages, they all have USB or TCP/IP ports and require a
150MB Windows or UNIX driver before that stuff can print anything. OK,
no problem, so I connect my ancient CP/M system via serial to a more
modern system, transfer my file and print it there. Now, why didn't I
write it in the first place with UNIX ed on the more modern system, when
I have to use the machine anyway? That is just a very basic and simple
example for usefull work not done with that systems anymore.
>
> Allison
Udo Munk
--
The real fun is building it and then using it...
-
Re: What are your obstacles to running CP/M?
On Sat, 05 Apr 2008 19:22:15 +0200, Udo Munk
wrote:
>no.spam@no.uce.bellatlantic.net schrieb:
>> I've always maintaind that for getting to bare metal CP/M is it.
>
>If you want to get at the bare metal of a Sun Microsystems Sunfire
>mainframe CP/M is it not. If you have learned Forth before, maybe on a
>CP/M system, that would help, because the firmware to get at the bare
>metal of this mainframes is written in Forth. One example of many.
>
>> Also while Linux both popular and useful it's also big wher CP/M BDOS
>> source is around 140K source and assembles to around 5.5k the whole
>> OS (CCP, BDOS, standard style BIOS) binary is around 8K total and the
>> source is maybe 250-300K. It's one of the few OSs out there that
>> every one that wants to can wrap their minds around the entire source
>> line by line. You can see the hardware and software interaction and
>
>Read newsgroups for Linux, BSD, Minix systems. There are nerds who can
>read the whole kernel sources to you.
I do and I have BSD unix on my PDP-11 and Minix on a 486 crate too.
I just finished putting Puppy Linux up on a miniITX system. Differnt
challenges and differnt fun.
>Of course it is much easier to
>handle the few lines of a CP/M kernel. That is good for total beginners
>who would like to learn about OS's. After a few weeks you move on to
>bigger and maybe better things.
I do a lot fo things from Blackfin and friends to PICs. Sometimes I
want to have fun and not need an engineering suite to do it.
>> also tradeoffs made in to make CP/M more portable. This is very
>> useful and instructive as everthing needed to understand OSs are there
>> (even if elementary level). There is useful software for it
>> including all of the basic tools one would expect and an assortment of
>> languages and user applications as well. It's old but it still can do
>> useful work.
>
>Yep, there is tons of software for this old systems and if you look into
>it, you'll find that not that much has changed on modern systems. This
>old software is a good base to start learning programming, because it's
>easier to understand than newer, more complex systems.
Indeed! I also get a bit of fun starting with CP/M and trying to add
things like multitasking (for V2.2) or trying to make the CP/M FS
do heirachal directories. No real sense in doing that but challenging
all the same.
>
>Usefull work? OK, I would like to write a simple letter, just a few
>lines of text, nothing fancy. So I start some ancient CP/M machine and
>type the few lines with ed and format that manually, no problem. Now I
>want to print that letter. No printer with serial or centronics ports
>available since ages, they all have USB or TCP/IP ports and require a
>150MB Windows or UNIX driver before that stuff can print anything. OK,
>no problem, so I connect my ancient CP/M system via serial to a more
>modern system, transfer my file and print it there. Now, why didn't I
>write it in the first place with UNIX ed on the more modern system, when
>I have to use the machine anyway? That is just a very basic and simple
>example for usefull work not done with that systems anymore.
Lets see, I use Vedit to write the letter after the less tha 15 second
boot from cold iron and print it on either my Epson LQ5000 or HP4L
or even an HP4200. The 4L is a bit old and 4ppm is not fast but the
12ppm 4200 seems to like the output of my s100 crate and even my
kaypro 4/84. But then again you can still buy new LQ5000s and the
wide platen version as well. The only printers I find that don't work
are the brainless ones that require GDI resources in winders and they
tend to be pretty cranky. The only headache for printing from old
systems is many have the wrong (non PC) connector and thats easy
to sort out. Of course If I really expeced to do that alot (I do) I
have a RS232 to Ethernet bridge to send it to the household print
server.
I guess it wouldn't be that hard to create a S100 to USB adaptor
or a Serial to USB adaptor using a Rabbit or other kit. That would
certainly be a project and get around the lack of new printers with
old interfaces.
Can the PC be faster... sure load up Xp or Ubuntu, wait what 2-5minuts
for boot and even then with a dual core and 4gb of ram it still takes
a bit to get Abiword cranked up. Makes sense if the document is
going to be emailed anyway. If it's a print and stuff an envelope
maybe it doesn't mater.
For small tasks a small systems does well enough, sometimes better
and somtimes not at all. All depends on the need.
Allison
-
Re: What are your obstacles to running CP/M?
no.spam@no.uce.bellatlantic.net schrieb:
> On Sat, 05 Apr 2008 19:22:15 +0200, Udo Munk
>> Read newsgroups for Linux, BSD, Minix systems. There are nerds who can
>> read the whole kernel sources to you.
>
> I do and I have BSD unix on my PDP-11 and Minix on a 486 crate too.
> I just finished putting Puppy Linux up on a miniITX system. Differnt
> challenges and differnt fun.
There you go. You probably know then that the kernel sources for early
PDP-11 UNIX's and Minix can be fully understood by a single person in a
couple of month.
Puppy Linux I have installed on a USB memory stick, usefull to boot
systems that can boot from USB, if the local installed OS got damaged
and won't boot anymore.
>> Of course it is much easier to
>> handle the few lines of a CP/M kernel. That is good for total beginners
>> who would like to learn about OS's. After a few weeks you move on to
>> bigger and maybe better things.
>
> I do a lot fo things from Blackfin and friends to PICs. Sometimes I
> want to have fun and not need an engineering suite to do it.
Yes of course, guess why I still mess with something like CP/M.
>> Usefull work? OK, I would like to write a simple letter, just a few
>> lines of text, nothing fancy. So I start some ancient CP/M machine and
>> type the few lines with ed and format that manually, no problem. Now I
>> want to print that letter. No printer with serial or centronics ports
>> available since ages, they all have USB or TCP/IP ports and require a
>> 150MB Windows or UNIX driver before that stuff can print anything. OK,
>> no problem, so I connect my ancient CP/M system via serial to a more
>> modern system, transfer my file and print it there. Now, why didn't I
>> write it in the first place with UNIX ed on the more modern system, when
>> I have to use the machine anyway? That is just a very basic and simple
>> example for usefull work not done with that systems anymore.
>
> Lets see, I use Vedit to write the letter after the less tha 15 second
> boot from cold iron and print it on either my Epson LQ5000 or HP4L
> or even an HP4200. The 4L is a bit old and 4ppm is not fast but the
> 12ppm 4200 seems to like the output of my s100 crate and even my
> kaypro 4/84. But then again you can still buy new LQ5000s and the
> wide platen version as well. The only printers I find that don't work
> are the brainless ones that require GDI resources in winders and they
> tend to be pretty cranky. The only headache for printing from old
> systems is many have the wrong (non PC) connector and thats easy
> to sort out. Of course If I really expeced to do that alot (I do) I
> have a RS232 to Ethernet bridge to send it to the household print
> server.
OK, so you also keep ancient printers, terminals and such alive, but for
how long? A running joke at work is, that sometime we will have to break
into a computer museum at night to get spare parts ;-) And that are not
CP/M machines or alike, that are 15 years old UNIX servers with 15 years
old UNIX version with 15 years old Oracle database. No hardware or
software support anymore for anything and the applications on that
systems can't be replaced easily.
Of course you can use a serial to ethernet bridge, but that was my
point. Why write that letter on the CP/M system, if some sort of modern
server system is up and running all the time anyway, ready to login and
get that simple work done there? You do that because you can, not
because it would make any sense at all ;-)
> I guess it wouldn't be that hard to create a S100 to USB adaptor
> or a Serial to USB adaptor using a Rabbit or other kit. That would
> certainly be a project and get around the lack of new printers with
> old interfaces.
Serial to USB adapters are available, we use them to connect modern
notebooks to ancient industrial machine controllers supporting a serial
console only.
> Can the PC be faster... sure load up Xp or Ubuntu, wait what 2-5minuts
> for boot and even then with a dual core and 4gb of ram it still takes
> a bit to get Abiword cranked up. Makes sense if the document is
> going to be emailed anyway. If it's a print and stuff an envelope
> maybe it doesn't mater.
That's why sleep and suspend modes have been invented, the machines are
up in 5 secounds, faster than it takes to boot CP/M from some ancient
floppy disk drive, with head load times of 500ms and such ;-)
> For small tasks a small systems does well enough, sometimes better
> and somtimes not at all. All depends on the need.
>
>
> Allison
>
Agreed. But the definition what is small has changed a bit over the years.
Udo Munk
--
The real fun is building it and then using it...
-
Re: What are your obstacles to running CP/M?
On Sun, 06 Apr 2008 17:00:07 +0200, Udo Munk
wrote:
>no.spam@no.uce.bellatlantic.net schrieb:
>> On Sat, 05 Apr 2008 19:22:15 +0200, Udo Munk
>>> Read newsgroups for Linux, BSD, Minix systems. There are nerds who can
>>> read the whole kernel sources to you.
>>
>> I do and I have BSD unix on my PDP-11 and Minix on a 486 crate too.
>> I just finished putting Puppy Linux up on a miniITX system. Differnt
>> challenges and differnt fun.
>
>There you go. You probably know then that the kernel sources for early
>PDP-11 UNIX's and Minix can be fully understood by a single person in a
>couple of month.
Key word *months*. I have Lions comentary to help!
>Puppy Linux I have installed on a USB memory stick, usefull to boot
>systems that can boot from USB, if the local installed OS got damaged
>and won't boot anymore.
very useful tool.
>
>>> Of course it is much easier to
>>> handle the few lines of a CP/M kernel. That is good for total beginners
>>> who would like to learn about OS's. After a few weeks you move on to
>>> bigger and maybe better things.
>>
>> I do a lot fo things from Blackfin and friends to PICs. Sometimes I
>> want to have fun and not need an engineering suite to do it.
>
>Yes of course, guess why I still mess with something like CP/M.
>
>>> Usefull work? OK, I would like to write a simple letter, just a few
>>> lines of text, nothing fancy. So I start some ancient CP/M machine and
>>> type the few lines with ed and format that manually, no problem. Now I
>>> want to print that letter. No printer with serial or centronics ports
>>> available since ages, they all have USB or TCP/IP ports and require a
>>> 150MB Windows or UNIX driver before that stuff can print anything. OK,
>>> no problem, so I connect my ancient CP/M system via serial to a more
>>> modern system, transfer my file and print it there. Now, why didn't I
>>> write it in the first place with UNIX ed on the more modern system, when
>>> I have to use the machine anyway? That is just a very basic and simple
>>> example for usefull work not done with that systems anymore.
>>
>> Lets see, I use Vedit to write the letter after the less tha 15 second
>> boot from cold iron and print it on either my Epson LQ5000 or HP4L
>> or even an HP4200. The 4L is a bit old and 4ppm is not fast but the
>> 12ppm 4200 seems to like the output of my s100 crate and even my
>> kaypro 4/84. But then again you can still buy new LQ5000s and the
>> wide platen version as well. The only printers I find that don't work
>> are the brainless ones that require GDI resources in winders and they
>> tend to be pretty cranky. The only headache for printing from old
>> systems is many have the wrong (non PC) connector and thats easy
>> to sort out. Of course If I really expeced to do that alot (I do) I
>> have a RS232 to Ethernet bridge to send it to the household print
>> server.
>
>OK, so you also keep ancient printers, terminals and such alive, but for
>how long? A running joke at work is, that sometime we will have to break
>into a computer museum at night to get spare parts ;-) And that are not
>CP/M machines or alike, that are 15 years old UNIX servers with 15 years
>old UNIX version with 15 years old Oracle database. No hardware or
>software support anymore for anything and the applications on that
>systems can't be replaced easily.
I keep them till they stop working. If buy a new inkjet it's
already obsolete and the last three didn't last a year befor
developing mechanical and other problems.
>Of course you can use a serial to ethernet bridge, but that was my
>point. Why write that letter on the CP/M system, if some sort of modern
>server system is up and running all the time anyway, ready to login and
>get that simple work done there? You do that because you can, not
>because it would make any sense at all ;-)
That and It's a good reason in it self to keep the hardware going
rather than piled in a shed.
>> I guess it wouldn't be that hard to create a S100 to USB adaptor
>> or a Serial to USB adaptor using a Rabbit or other kit. That would
>> certainly be a project and get around the lack of new printers with
>> old interfaces.
>
>Serial to USB adapters are available, we use them to connect modern
>notebooks to ancient industrial machine controllers supporting a serial
>console only.
Yes I work with them. Most are host to client/slave rather than
reverse.
>> Can the PC be faster... sure load up Xp or Ubuntu, wait what 2-5minuts
>> for boot and even then with a dual core and 4gb of ram it still takes
>> a bit to get Abiword cranked up. Makes sense if the document is
>> going to be emailed anyway. If it's a print and stuff an envelope
>> maybe it doesn't mater.
>
>That's why sleep and suspend modes have been invented, the machines are
>up in 5 secounds, faster than it takes to boot CP/M from some ancient
>floppy disk drive, with head load times of 500ms and such ;-)
Well the new Dell dual core with 4gb ram and 160mb of raid between the
LCD display and all still needs 15-20sec to "wake up" and be useful.
>> For small tasks a small systems does well enough, sometimes better
>> and somtimes not at all. All depends on the need.
>>
>>
>> Allison
>>
>
>Agreed. But the definition what is small has changed a bit over the years.
The only constant is change. 
Sometimes trying to drag CP/M on Z80 into the 21st century is often a
challange worth effort and other times a firebreathing PC with GUI is
way too much challange. I try not to take any of them to seriously.
Besides a VAX or Alpha running VMS still seem to be worthy of effort
too. In the end some days I want to play with solder and other days
bits.
Allison
-
Re: What are your obstacles to running CP/M?
*Udo Munk* wrote on Sun, 08-04-06 17:00:
>Serial to USB adapters are available,
Not really. What you refer to is a USB slave. To use USB peripherals
from a CP/M machine you'd need a USB master with serial connection.
Quite possible to do using a microcontroller but no off the shelp part.
-
Re: What are your obstacles to running CP/M?
*Udo Munk* wrote on Sun, 08-04-06 17:00:
>But the definition what is small has changed a bit over the years.
Quite. The full complement of a whopping great 64 kB is a seriously big
machine -- for most users about 24 kB is adequate and nearly affordable.
-
Re: What are your obstacles to running CP/M?
On Apr 6, 12:54 pm, no.s...@no.uce.bellatlantic.net wrote:
> >Serial to USB adapters are available, we use them to connect modern
> >notebooks to ancient industrial machine controllers supporting a serial
> >console only.
> Yes I work with them. Most are host to client/slave rather than
> reverse.
And there are SPI USB host controller chips as well, which only
requires 4 GPIO lines. Oh, and the programming.
-
Re: What are your obstacles to running CP/M?
Axel Berger schrieb:
> *Udo Munk* wrote on Sun, 08-04-06 17:00:
>> Serial to USB adapters are available,
>
> Not really. What you refer to is a USB slave. To use USB peripherals
> from a CP/M machine you'd need a USB master with serial connection.
> Quite possible to do using a microcontroller but no off the shelp part.
I don't want to use it from a CP/M system or alike machine. I plugin
that adapter into a 21th century notebook, connect the RS232 to some
ancient machine, and use a serial communication program to talk to that
machine.
I don't have to upgrade an ancient machine to a 21th century machine if
those are readily available anyway. I also don't try to put an Airbus
engine into an ancient airplane, I know that this would be pretty
difficult, even not beeing an engineer for such stuff.
Udo Munk
--
The real fun is building it and then using it...
-
Re: What are your obstacles to running CP/M?
Axel Berger schrieb:
> *Udo Munk* wrote on Sun, 08-04-06 17:00:
>> But the definition what is small has changed a bit over the years.
>
> Quite. The full complement of a whopping great 64 kB is a seriously big
> machine -- for most users about 24 kB is adequate and nearly affordable.
One of the first computer systems I used had magnetical core memory, 4K
words of 12bits. That was good enough to run a Fortran compiler and an
Assembler on that machine. A system with 24KB is LARGE and unnecessary
luxus and a waste of resources ;-)
Udo Munk
--
The real fun is building it and then using it...
-
Re: What are your obstacles to running CP/M?
On Sun, 06 Apr 2008 23:31:26 +0200, Udo Munk
wrote:
>Axel Berger schrieb:
>> *Udo Munk* wrote on Sun, 08-04-06 17:00:
>>> But the definition what is small has changed a bit over the years.
>>
>> Quite. The full complement of a whopping great 64 kB is a seriously big
>> machine -- for most users about 24 kB is adequate and nearly affordable.
>
>One of the first computer systems I used had magnetical core memory, 4K
>words of 12bits. That was good enough to run a Fortran compiler and an
>Assembler on that machine. A system with 24KB is LARGE and unnecessary
>luxus and a waste of resources ;-)
>
>Udo Munk
As someone with a PDP-8 F and all I can appreciate and agree with
that. Bloatware is an ugly thing.
Allison
-
Re: What are your obstacles to running CP/M?
On Apr 6, 5:31 pm, Udo Munk wrote:
> One of the first computer systems I used had magnetical core memory, 4K
> words of 12bits. That was good enough to run a Fortran compiler and an
> Assembler on that machine. A system with 24KB is LARGE and unnecessary
> luxus and a waste of resources ;-)
Oh, I don't know about that, it also depends on how much K per user.
The first Forth systems were implemented on 16KB and 32KB
minicomputers, but since that was for multi-user multi-programming and
both data analysis and instrument control, I'm thinking that it might
be possible to justify that when its converted to a KB per concurrent
task basis.