How many SIG/M disks where there? - CP/M

This is a discussion on How many SIG/M disks where there? - CP/M ; Mr. Emmanuel Roche, France schrieb: > "CP/NET can be tunneled over TCP/IP, that's very different." ? I don't > understand. What I have done is encapsulating the CP/NET protocol into TCP/IP packets, which is layer 4 (OSI). That is what ...

+ Reply to Thread
Page 2 of 6 FirstFirst 1 2 3 4 ... LastLast
Results 21 to 40 of 118

Thread: How many SIG/M disks where there?

  1. Re: How many SIG/M disks where there?

    Mr. Emmanuel Roche, France schrieb:

    > "CP/NET can be tunneled over TCP/IP, that's very different." ? I don't
    > understand.


    What I have done is encapsulating the CP/NET protocol into TCP/IP
    packets, which is layer 4 (OSI). That is what you usually do if layers
    below won't transport your protocol and nowadays there is no link layer
    other than serial available anymore, that would do. Of course that
    implementation is inefficient as the CP/NET documentation explains in
    great detail, but well, it is easy to do and good enough for playing a
    little bit with this ancient stuff. I don't know if that helps you
    understanding what I have done, at least I tried to explain it as easy
    as possible.

    Udo Munk
    --
    The real fun is building it and then using it...

  2. Re: How many SIG/M disks where there?

    Hello, Udo!

    Ok. I am going to bed, and tomorrow, I will be away the whole day.

    So, this will be short.

    > For any protocol you need a physical layer and defined frames to
    > transport that protocol. E.g. for TCP/IP ethernet is pretty common. If
    > you read the ethernet specs you'll find that frames for TCP/IP, Decnet
    > and Novell IPX are defined, CP/NET is not. And so you can run TCP/IP,
    > Decnet and Novell on the same wire, but not CP/NET. And I'm looking
    > forward how you talk layer 1 vendors into implementing CP/NET for you.


    If I have well understood today's readings, most Web browsers work at the
    4th level, where "levels" are:

    4th: Application
    3th: TCP
    2nd: IP
    1st: Ethernet

    CP/NET was the first NOS for CP/M systems, and used the simplest possible
    protocol at the time (1982), down to a checksum similar to Intel HEX files.
    But I don't see why you could not change it to use a TCP/IP driver using an
    Ethernet card, such as those found on IBM PCs? (I found one article that
    claimed that TCP/IP could fit into 10 to 24KB. I have already disassembled a
    dozen programs bigger than this size.)

    > > (I have spent a good deal of time, today, investigating TCP/IP,

    searching
    > > how it could be implemented under CP/M.)

    >
    > Oh, it is already implemented, was done.


    ??? Under CP/M? Where? When? The name? (Not found with Google.)

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  3. Re: How many SIG/M disks where there?

    Udo Munk wrote:
    (snip)

    > I have seldom read rubbish worse than this. Obviously you don't know
    > that UNIXs on the same platform keept binary compatibility since
    > decades, longer than CP/M was used actively. On a 21th century x86 UNIX
    > system you could right now run iBCS compatible COFF binaries programmed
    > in the 80th.


    Maybe, but I wouldn't be surprised if it didn't. People like
    to change things too much.

    IBM has done pretty well, though. OS/360 programs from the 1960's
    will run under z/OS on current 64 bit z/Architecture machines.
    That is without recompilation, reassembly, or relinking.

    > The current ELF 64bit binary format used on all UNIX
    > platforms was developed 1993.


    It has to be also the same calling convention (if you want
    to link with new code) and same system calls to the OS.
    Those don't always stay constant.

    -- glen


  4. Re: How many SIG/M disks where there?

    On Mon, 21 Jul 2008 20:48:48 +0200, Udo Munk
    wrote:

    >Mr. Emmanuel Roche, France schrieb:
    >> Hello, Udo!
    >>
    >>> There is not much CP/M discussion here

    >>
    >> ??? Strange, I was thinking that this is the comp.os.CPM Newsgroup!...

    >
    >Yeah, and today I stamped every thing I saw an X because I though it was
    >an X.
    >
    >>> If you want to use CP/M to access the Internet then you need to
    >>> implement appropriate software for this

    >>
    >> This is, indeed, one of my programming projects since, years after years, if
    >> I don't do something, nobody else do anything.

    >
    >Year after year you are working on this? Oh my. In the 80th we used uucp
    >to exchange email and newsgroups via dialup lines. I used UNIX systems
    >for that, but on the Walnut Creek CP/M CD-ROM I see uucp,
    >uudecode-/encode, rot13 and all that stuff necessary. Looks like someone
    >was much faster than you and did all that already in the 80th. And I see
    >various TCP/IP implementations for Z80 systems, tiny Z80 systems acting
    >as a web server just for fun. If you do some careful research you even
    >will find web admins wondering, that someone with a CP/M web browser
    >seems to get contents from their server.
    >Has all been done already, you're too late. If you want that stuff then
    >get it and use it.
    >
    >>
    >>> Excuse me, but at work I use real big servers all running Unix

    >>
    >> So what? Don't you know that I was a COBOL programmer on IBM Mainframes? The

    >
    >No, how would I know?
    >
    >> computing world is divided in several "worlds". There is/was the IBM
    >> Mainframes, the DEC VAXes, then the IBM PC. Unix was only used inside
    >> American Universities. One of my friend was a total FORTRAN fanatic. Have
    >> you any idea of how many Mega/Giga/Tera Bytes of FORTRAN code exist? (and
    >> COBOL?) You would be astonished. And they are not half-cooked code made by
    > > students in American Universities...

    >
    >Rubish, that are all just binary computers switching bits forth and
    >back. UNIX systems were worldwide available outside universities in the
    >80th, this OS was written 1968. Even Z80 UNIX systems existed and were
    >sold to anyone who wanted to buy one. As an example I used Cromenco
    >systems, outside of universities.
    >Yes I have an idea how many lines of FORTRAN and PL/1 was written,
    >because I wrote quite some of that stuff. I never touched COBOL tho.
    >
    >> Also, there is the problem that there is several kinds of OSes. As my
    >> experience demonstrates, you can be perfectly happy with a 4-MHz Z-80
    >> running a single-user OS like CP/M Plus during 15 years. Multi-tasking OSes
    >> are, inherently, more difficult to use. No wonder that the market uses, by a
    >> huge percentage, Windows rather than Linux. There are also real-time OSes,
    >> but they are needed only in special cases. That's why they are so little
    >> known.

    >
    >Because that isn't so companies like DRI are long gone, Microsoft threw
    >DOS out of the Windows and Caldera gave up on DR-DOS it seems.
    >It was proved 1981 by DRI with MP/M and before with other multiuser
    >systems, that users won't notice the difference. They switch on their
    >terminal and start to use the machine. Windows is a multi-user and
    >multitasking system, if not it would have been gone quite some time.
    >
    >> How many Linux work on a 4-MHz Z-80 with 64KB of RAM?

    >
    >Linux never was ported to 8bit CPU's, 16bit micro controllers is the
    >smallest unit that makes sense nowadays. And unlike Windows several Z80
    >UNIX systems existed, even ports of the original AT&T system.


    Uzi unix was on Z80 and some of the small linux like uClinux cols
    possibly fit if there was memory management (banking).

    >
    >>> Reading my email on a CP/M system isn't in that category.

    >>
    >> Too bad for us.

    >
    >You better speak for your self only, I doubt that there is much demand
    >for reading mail on a CP/M system, and if so then just do it.


    I did it way back there isn't much special to doing it if you have
    access and a protocal to the mailserver. It was a trivial thing
    wirtten in BDS C to parse the mail, present it, and mark the
    read or deleted stuff. The real work was was getting the
    modems of the day to behave and some semblance of
    a protocal. thank the diety mail back them was text only.

    >>> (...) If you want
    >>> me to build your stuff then you are supposed to pay me for that.

    >>
    >> Hahaha! Me, I have (according to Google) made more than 20 BASIC programs
    >> that made something useful under CP/M, during the last 10 years. All that
    >> for free, with a text explaining in depth why the program was made. I am a
    >> programmer, not a carpet dealer.

    >
    >20 BASIC programs in 10 years and all free, I'm impressed, NOT.
    >
    >>> CP/NET is a very simple point to point protocol for remote procedure
    >>> calls to share some resources like disks and printers. With that you
    >>> won't go anywhere far. Internet uses TCP/IP protocol, something very
    >>> different. Instead of trying to explaining the Internet to me get your
    >>> self busy understanding the protocols. Then you can write software using

    >> it.
    >>
    >> Don't worry, I will do it, in times. Haven't you noticed that I do things
    >> step-by-step?

    >
    >No, I'm used to humans doing many things at once.
    >
    >> That's why I mentioned CP/NET version 1.2. 1) I hoped that its mention would
    >> revive its use for CP/M 2.2 systems, 2) CP/NET is not tied to any particular
    >> protocol: it could run just as well with TCP/IP, 3) I was thinking, of
    >> course, that it would be a good first step to do before going farther.

    >
    >CP/NET was thrown away early in the 80th by DRI, because they figured it
    >is useless compared to other protocols. Since they obviously had DEC
    >equipment at their facility, they knew about Decnet e.g.
    >CP/NET won't run together with TCP/IP, read the specs, it can be
    >tunneled over TCP/IP, very different.


    What CPnet does plain CP/M can do there is little special there and
    it's mostly a bios issue.

    >
    >> Meanwhile, since I am a fan of MP/M-II (I deeply regret not to have a
    >> MP/M-II system. I am really sorry that Bruce Jones disappeared.), I hope to
    >> be able to play again on your MP/M-II system. In particular, I would like to
    >> try CP/NET, to see how you implemented it. Hector Peraza wrote a CP/NET
    >> server for Linux. Here is what he wrote:

    >
    >I did not implement CP/NET this was implemented by DRI, remember? Same
    >as with CP/M they split the system dependent parts, so that it could
    >easily been adapted to any kind of hardware. Adapting it to some sort of
    > serial UART just is changing address of I/O ports and bitmasks for
    >status bits, can be done in half an hour, plus some time for debugging
    >and testing.


    Udo, your chasing an untaimed ornithoid and whatsit isn't listening.

    Later,

    Allison

    >
    >Udo Munk



  5. Re: How many SIG/M disks where there?

    On Mon, 21 Jul 2008 22:12:50 +0200, Udo Munk
    wrote:

    >Mr. Emmanuel Roche, France schrieb:
    >...
    >> I can confirm that Digital Research used DEC VAXes. The CP/M Plus BIOS
    >> provided in the System Guide was connected to a VAX, among other stuff. Now,

    >
    >Oh, well, I can confirm that DRI used a VAX as filestore for all their
    >sources, because their build scripts for CP/M 3 and MP/M fetch the
    >source files from a VAX, not via CP/NET by the way. But that is not
    >really new, the sources are available quite a while.


    Besides VMS back then only used DECnet as the networking protocal.
    If you had a local machine and no ehternet you could use a serial line
    running DDCMP as point to point protocal. However the small VAX
    (11/750) they had was likely a serial line running kermit or other
    similar protocal over a standard async line.

    >> what I don't understand is why you wrote: "CP/NET won't run together with
    >> TCP/IP, reads the specs"? I have retyped everything that I could find about
    >> CP/NET (and DR-Net for the IBM PC): nowhere I have seen any such limitation.

    >
    >For any protocol you need a physical layer and defined frames to
    >transport that protocol. E.g. for TCP/IP ethernet is pretty common. If
    >you read the ethernet specs you'll find that frames for TCP/IP, Decnet
    >and Novell IPX are defined, CP/NET is not. And so you can run TCP/IP,
    >Decnet and Novell on the same wire, but not CP/NET. And I'm looking
    >forward how you talk layer 1 vendors into implementing CP/NET for you.


    CP/net is network aware but protocal and hardware blind. It's a NIOS
    responsability plus and undelying hardware to do IP, DECnet
    or IPX (or any others).

    >> "CP/NET can be tunneled over TCP/IP, that's very different." ? I don't
    >> understand.

    >
    >Yep, you need to learn about networking before.
    >
    >> (I have spent a good deal of time, today, investigating TCP/IP, searching
    >> how it could be implemented under CP/M.)

    >
    >Oh, it is already implemented, was done.


    Yes, the KA9Q suite did IP and UDP. it also contained ICMP, SMTP,
    SLIP, TELNET, FTP and IProute. It's quite old 1985 and was implemeted
    using CP/M 2.2 plus X.25 for the bit level protocal for amateur
    packet ratio (store and forward messaging and email). Phil Karn
    {KA9Q} did some fine work.


    Allison
    >
    >Udo Munk



  6. Re: How many SIG/M disks where there?

    On Tue, 22 Jul 2008 01:38:54 GMT, no.spam@no.uce.bellatlantic.net
    wrote:

    >Yes, the KA9Q suite did IP and UDP. it also contained ICMP, SMTP,
    >SLIP, TELNET, FTP and IProute. It's quite old 1985 and was implemeted
    >using CP/M 2.2 plus X.25 for the bit level protocal for amateur
    >packet ratio (store and forward messaging and email). Phil Karn
    >{KA9Q} did some fine work.


    That reminded me ... TAPR (Tucson Amateur Packet Radio) guys
    were really into that too ... I went to a few of their meetings but
    decided it wasn't the kind of work (mostly, building repeaters to
    stick on top of high places like water towers etc) I wanted to do
    'just for the fun of it'. Anyway, try googling TAPR....

    They were really big on using surplus Zerox 820 single board
    (Z80) computers as the network switching devices ...1983/84?

    Bill

  7. Re: How many SIG/M disks where there?

    Mr. Emmanuel Roche, France wrote:
    (snip)

    > (I have spent a good deal of time, today, investigating TCP/IP, searching
    > how it could be implemented under CP/M.)


    UDP should be pretty easy, TCP somewhat harder.

    With UDP you could use TFTP for file transfer.

    -- glen


  8. Re: How many SIG/M disks where there?

    glen herrmannsfeldt schrieb:
    > Udo Munk wrote:
    > (snip)
    >
    >> I have seldom read rubbish worse than this. Obviously you don't know
    >> that UNIXs on the same platform keept binary compatibility since
    >> decades, longer than CP/M was used actively. On a 21th century x86
    >> UNIX system you could right now run iBCS compatible COFF binaries
    >> programmed in the 80th.

    >
    > Maybe, but I wouldn't be surprised if it didn't. People like
    > to change things too much.


    Still is available for Linux and former SCO OS's and there is no reason
    to modify it. It could be taken out sometime because no one uses any
    software in that binary format anymore.

    > IBM has done pretty well, though. OS/360 programs from the 1960's
    > will run under z/OS on current 64 bit z/Architecture machines.
    > That is without recompilation, reassembly, or relinking.


    That is another example, yes.

    >> The current ELF 64bit binary format used on all UNIX platforms was
    >> developed 1993.

    >
    > It has to be also the same calling convention (if you want
    > to link with new code) and same system calls to the OS.
    > Those don't always stay constant.
    >
    > -- glen


    Unlinke .com files, which are just binary images of executable programs
    linked to a fixed address, the ELF object format includes many of the
    necessary definitions. If the kernel system calls change one modifies a
    shared library and that's it, no applications need to be recompiled or
    relinked.

    Udo Munk
    --
    The real fun is building it and then using it...

  9. Re: How many SIG/M disks where there?

    > > > (I have spent a good deal of time investigating TCP/IP,
    > > > searching how it could be implemented under CP/M.)

    > >
    > > Oh, it is already implemented, was done.

    >
    > ??? Under CP/M? Where? When? The name? (Not found with Google.)


    > Yes, the KA9Q suite did IP and UDP. it also contained ICMP, SMTP,
    > SLIP, TELNET, FTP and IProute. It's quite old 1985 and was implemeted
    > using CP/M 2.2 plus X.25 for the bit level protocal for amateur
    > packet ratio (store and forward messaging and email). Phil Karn
    > {KA9Q} did some fine work.


    Ok. I am back. Today, I spent 2 hours, trying to find KA9Q for CP/M, without
    finding it. So, if someone has this rarity, here is the occasion to save it.
    (So, I was right: TCP/IP for CP/M does not exist, since even myself cannot
    find it.)

    I have also found a 2002 Web page by Phil Karn saying that he was getting a
    lot of questions about his package. Seems it is not well documented. (In
    this light, it was fun to discover what a Swedish computer scientist was
    using (Adam Dunkels, creator of Contiki and uIP): "The Contiki source code
    is documented using a tool called Doxygen. Doxygen reads specially tagged
    comments in the source code and produces documentation in both HTML and PDF
    format." That is to say; exactly in essence, what my LitProg.BAS does, with
    WS4 and a Programming Language.) (I have also a separate WS4-to-HTML File
    Converter.)

    On his Web page, Phil Karn mentions 3 useful books, to understand TCP/IP. I
    found exactly those books in the USA, and am waiting for them, now. Looks
    like I could start a "TCP/M Project", since nobody else wants to do it under
    CP/M...

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  10. Re: How many SIG/M disks where there?

    Mr. Emmanuel Roche, France schrieb:
    > Hello, Udo!
    >
    > Ok. I am going to bed, and tomorrow, I will be away the whole day.
    >
    > So, this will be short.
    >
    >> For any protocol you need a physical layer and defined frames to
    >> transport that protocol. E.g. for TCP/IP ethernet is pretty common. If
    >> you read the ethernet specs you'll find that frames for TCP/IP, Decnet
    >> and Novell IPX are defined, CP/NET is not. And so you can run TCP/IP,
    >> Decnet and Novell on the same wire, but not CP/NET. And I'm looking
    >> forward how you talk layer 1 vendors into implementing CP/NET for you.

    >
    > If I have well understood today's readings, most Web browsers work at the
    > 4th level, where "levels" are:
    >
    > 4th: Application
    > 3th: TCP
    > 2nd: IP
    > 1st: Ethernet


    This is not the OSI layer model, but nevermind, this model will do too.
    Let's call layer 1st transport layer, TCP/IP doesn't work on ethernet
    only, best example is your nowadays DSL connection to the Internet.

    > CP/NET was the first NOS for CP/M systems, and used the simplest possible
    > protocol at the time (1982), down to a checksum similar to Intel HEX files.


    The DRI guys were smart. The checksum is optional, only needed with a
    transport layer, that doesn't do checksum and retransmission already.

    > But I don't see why you could not change it to use a TCP/IP driver using an
    > Ethernet card, such as those found on IBM PCs? (I found one article that


    That is what I have done. It is called tunneling protocol xxx over
    TCP/IP. Another term used is encapsulating protocol xxx in TCP/IP.

    > claimed that TCP/IP could fit into 10 to 24KB. I have already disassembled a
    > dozen programs bigger than this size.)


    Such claims are pretty useless. There is no need to disassemble, the
    TCP/IP stacks for Z80 systems are available as source code.

    >>> (I have spent a good deal of time, today, investigating TCP/IP,

    > searching
    >>> how it could be implemented under CP/M.)

    >> Oh, it is already implemented, was done.

    >
    > ??? Under CP/M? Where? When? The name? (Not found with Google.)
    >
    > Yours Sincerely,
    > Mr. Emmanuel Roche, France


    Before I get into that 21th century google mystery, not running on CP/M
    or Windows, I'll tell you a top secret. The company Zilog, the guys who
    invented the Z80 processor, still produce a Z80 CPU right now. The thing
    is called eZ80. This little thingy is a Z80 CPU (with some extensions)
    and a complete TCP/IP protocol stack put into the hardware. There is
    much need for such a controller, so that you can get your coffee machine
    and your fridge connected to your LAN at home. Now that wouldn't help
    you much, you don't want to read your email from the display of a
    fridge, you want CP/M right? Well, look here: http://www.ez80sbc.com/
    A complete system with that eZ80 already running CP/M 3. From the
    pictures it looks like ethernet is available via the usual CAT5
    connector. All that is left to read email with this system is writing a
    BSD compatible socket library to interface with the TCP/IP protocol
    stack already available in the hardware, and then compiling the BSD mail
    and news programs (no licence restrictions, use other if you like) with
    a C compiler. 2-3 evenings of work maybe?

    Google... when I enter "Z80 TCP/IP" as search string I get hundreds of
    hits, good matches on the first page already, describing ancient CP/M
    systems enhanced with a TCP/IP stack to run web servers, ftp clients,
    mail clients and all that stuff, ready to download and use, with all
    sources. The search string is pretty easy, you tried that of course, but
    obviously you didn't get the same result as I got. The top secret (if
    you tell that anyone I've to shot you, or so) is the national language
    configured in your web browser. I use different preferences than you, so
    I get different results, because google uses the language preference you
    configured in your web browser to compute the rank. OK, now all that
    learned, my language preference is de, en_us, en. I will not post the
    links I found, because obviously you need to learn to google and it
    won't help if I do it for you.

    Udo Munk
    --
    The real fun is building it and then using it...

  11. Re: How many SIG/M disks where there?

    Mr. Emmanuel Roche, France schrieb:
    ....
    > On his Web page, Phil Karn mentions 3 useful books, to understand TCP/IP. I
    > found exactly those books in the USA, and am waiting for them, now. Looks
    > like I could start a "TCP/M Project", since nobody else wants to do it under
    > CP/M...
    >
    > Yours Sincerely,
    > Mr. Emmanuel Roche, France


    OK I give up, this guy is truly hopeless.

    Udo Munk
    --
    The real fun is building it and then using it...

  12. Re: How many SIG/M disks where there?

    On Wed, 23 Jul 2008 23:13:32 +0200, Udo Munk
    wrote:

    >Mr. Emmanuel Roche, France schrieb:
    >...
    >> On his Web page, Phil Karn mentions 3 useful books, to understand TCP/IP. I
    >> found exactly those books in the USA, and am waiting for them, now. Looks
    >> like I could start a "TCP/M Project", since nobody else wants to do it under
    >> CP/M...
    >>
    >> Yours Sincerely,
    >> Mr. Emmanuel Roche, France

    >
    >OK I give up, this guy is truly hopeless.


    Ah yes, affirmitive.

    Roche:

    FYI: when you find the IP stack you will also find it was already
    running under CP/M in 1985.

    Allison

    >
    >Udo Munk



  13. Re: How many SIG/M disks where there?

    Hello, Udo!

    Reading your message, I was laughing, since we are on such different
    wavelentghs!

    > This is not the OSI layer model, but nevermind, this model will do too.


    Anaway, as far as know the ISO OSI model is purely theorical, and has been
    overcomed by TCP/IP, by a "grass root effect" (as they say in
    English/American).

    > > claimed that TCP/IP could fit into 10 to 24KB. I have already

    disassembled a
    > > dozen programs bigger than this size.)

    >
    > Such claims are pretty useless. There is no need to disassemble, the
    > TCP/IP stacks for Z80 systems are available as source code.


    Yes, but probably in "C"... and, years after years, we have heard stories
    here, in the comp.os.cpm Newsgroup, that C is not portable, or too big, or
    too slow. In my opinion, Unix and C are viable only on a 16-bits+ computer.
    For CP/M, I think that assembly language is the only solution (I searched
    the Internet for "TCP/IP in Forth", but found nothing, to my surprise).

    > or Windows, I'll tell you a top secret. The company Zilog, the guys who
    > invented the Z80 processor, still produce a Z80 CPU right now. The thing
    > is called eZ80. This little thingy is a Z80 CPU (with some extensions)
    > and a complete TCP/IP protocol stack put into the hardware. There is


    Yes, but this is not a Z-80 running at 4-MHz under CP/M, so it does not
    count.

    On the Web, there is a video of a German surfing the Net on a C=64 (if I
    have well understood). This proves that my goal is doable. I was VERY
    surprised that no Z-80 port of uIP (or Contiki) were made. Maybe Adam Dunkel
    learned to program on a 6502, and don't knows that the Z-80 exists?

    > Google... when I enter "Z80 TCP/IP" as search string I get hundreds of
    > hits, good matches on the first page already, describing ancient CP/M


    Hahaha! Me, I had entered "TCP/IP for CP/M"...

    Ok. I will now wait for the TCP/IP books. I think that everything that I
    will find will be in "C", since Internet (and, hence, TCP/IP) is based on
    some Unix standards. I already know how to write a library of macros
    emulating the printf function of C. For years, I have wanted a "C-to-BASIC"
    File Converter... Maybe it is time to write it? (I searched "TCP/IP in
    BASIC": no answer! It would be funny to be the first man to program a TCP/IP
    in BASIC! This way, the source would be portable, and readable.)

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  14. Re: How many SIG/M disks where there?

    On Thu, 24 Jul 2008 00:41:10 +0200, "Mr. Emmanuel Roche, France"
    wrote:

    >Hello, Udo!
    >
    >Reading your message, I was laughing, since we are on such different
    >wavelentghs!
    >
    >> This is not the OSI layer model, but nevermind, this model will do too.

    >
    >Anaway, as far as know the ISO OSI model is purely theorical, and has been
    >overcomed by TCP/IP, by a "grass root effect" (as they say in
    >English/American).


    The OSI model is still valid even for TCP/IP.

    >
    >> > claimed that TCP/IP could fit into 10 to 24KB. I have already

    >disassembled a
    >> > dozen programs bigger than this size.)

    >>
    >> Such claims are pretty useless. There is no need to disassemble, the
    >> TCP/IP stacks for Z80 systems are available as source code.

    >
    >Yes, but probably in "C"... and, years after years, we have heard stories
    >here, in the comp.os.cpm Newsgroup, that C is not portable, or too big, or


    C can be portable and BASIC can be non portable it the coders
    responsability to insure portability.

    >too slow. In my opinion, Unix and C are viable only on a 16-bits+ computer.


    Sorry the embedded systems world use it for very small machines
    including PICs and Z80s.

    >For CP/M, I think that assembly language is the only solution (I searched
    >the Internet for "TCP/IP in Forth", but found nothing, to my surprise).


    Assembler in 1984 was the solution not for portability or size but
    speed. Things like interrupt handlers or other system specific
    events.

    In the case of the KA9Q-IP package it's .asm for the low level stuff
    like device drivers and .c for all the higher stuff.


    >> or Windows, I'll tell you a top secret. The company Zilog, the guys who
    >> invented the Z80 processor, still produce a Z80 CPU right now. The thing
    >> is called eZ80. This little thingy is a Z80 CPU (with some extensions)
    >> and a complete TCP/IP protocol stack put into the hardware. There is

    >
    >Yes, but this is not a Z-80 running at 4-MHz under CP/M, so it does not
    >count.


    Why not? Z80s were/are available to 10mhz and the eZ80 can and does
    run CP/M.

    >On the Web, there is a video of a German surfing the Net on a C=64 (if I
    >have well understood). This proves that my goal is doable. I was VERY
    >surprised that no Z-80 port of uIP (or Contiki) were made. Maybe Adam Dunkel
    >learned to program on a 6502, and don't knows that the Z-80 exists?


    He knows it exists but that not his favorite apparently.

    >> Google... when I enter "Z80 TCP/IP" as search string I get hundreds of
    >> hits, good matches on the first page already, describing ancient CP/M

    >
    >Hahaha! Me, I had entered "TCP/IP for CP/M"...


    I just searched the copy of the WC CP/M cdrom I keep on hard disk.
    160 Gigabytes of storage are cheap in this day of Terabyte sized
    drives.

    >Ok. I will now wait for the TCP/IP books. I think that everything that I
    >will find will be in "C", since Internet (and, hence, TCP/IP) is based on
    >some Unix standards. I already know how to write a library of macros
    >emulating the printf function of C. For years, I have wanted a "C-to-BASIC"
    >File Converter... Maybe it is time to write it? (I searched "TCP/IP in
    >BASIC": no answer! It would be funny to be the first man to program a TCP/IP
    >in BASIC! This way, the source would be portable, and readable.)


    Your reaching a bit. There is no written in C assumption and there
    were many implementations on machines with OSs where C was rarely
    used. C compilers for everything and anything is far more recent
    than IP. Also the assumption that IP and Unix are linked closely
    fails for the same reason.

    In BASIC (for any Z80 based system) variables are all global,
    something that is not true in C. Most Basics do not support
    recursion. Learn the language before you get hurt. While your at it
    the C language does not have any native IO, Printf() is a library
    function and could be coded in assembler, or any other competent
    compiled language.

    If your going to be critical of C first read "Computing Science
    Technical Report #31" or as it's was known in other spaces
    "the C programming Language By Kerninghan, Richie, and Lesk"
    written in 1975. Then read the ANSI specification for C.

    IP in BASIC is possible. It would be very big, very slow and far from
    portable as BASIC is not portable. Why, since speed would dictate a
    compiler and the most likely compiler is the MSbasic (BASCOM) for
    CP/M has a 24K runtime package thats the minimum size of the code
    then. Interpreted BASIC was hard pressed to do IO fast e4nough to
    keep a TTY busy (110 baud).


    Allison




    >
    >Yours Sincerely,
    >Mr. Emmanuel Roche, France
    >
    >



  15. Re: How many SIG/M disks where there?

    On 2008-07-24, no.spam@no.uce.bellatlantic.net wrote:
    > Sorry the embedded systems world use it for very small machines
    > including PICs and Z80s.


    Indeed. C on 68HC11 is *very* nice.
    --
    roger ivie
    rivie@ridgenet.net

  16. Re: How many SIG/M disks where there?


    "Mr. Emmanuel Roche, France" wrote:

    > Anaway, as far as know the ISO OSI model is purely theorical, and has been
    > overcomed by TCP/IP, by a "grass root effect" (as they say in
    > English/American).


    Must have been tired. According to my English dictionaries, the correct
    expression is "grass root MOVEMENT" (not "effect").

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  17. Re: How many SIG/M disks where there?

    Udo Munk wrote:

    > Before I get into that 21th century google mystery, not running on CP/M
    > or Windows, I'll tell you a top secret. The company Zilog, the guys who
    > invented the Z80 processor, still produce a Z80 CPU right now. The thing
    > is called eZ80. This little thingy is a Z80 CPU (with some extensions)
    > and a complete TCP/IP protocol stack put into the hardware. There is
    > much need for such a controller, so that you can get your coffee machine
    > and your fridge connected to your LAN at home. Now that wouldn't help
    > you much, you don't want to read your email from the display of a
    > fridge, you want CP/M right? Well, look here: http://www.ez80sbc.com/
    > A complete system with that eZ80 already running CP/M 3. From the
    > pictures it looks like ethernet is available via the usual CAT5
    > connector. All that is left to read email with this system is writing a
    > BSD compatible socket library to interface with the TCP/IP protocol
    > stack already available in the hardware, and then compiling the BSD mail
    > and news programs (no licence restrictions, use other if you like) with
    > a C compiler. 2-3 evenings of work maybe?


    I've tripped over the eZ80 system several times in my travels on the net. The
    price is unfortunate, though. It's just outside of what I would consider
    spending.

    If the complete system (ethernet + download cables) was $200 or under, I'd go
    for it in a heartbeat. Unassembled is a non-starter, due to the heavy
    (exclusive?) use of SMD technology.

    At $350, it will remain on my wish-list. Too bad...

    Steve

  18. Re: How many SIG/M disks where there?

    Mr. Emmanuel Roche, France schrieb:
    > Hello, Udo!
    >
    > Reading your message, I was laughing, since we are on such different
    > wavelentghs!


    Glad I entertained you. No offense intended, but I do not wish to be on
    the same wavelength with you.

    >> This is not the OSI layer model, but nevermind, this model will do too.

    >
    > Anaway, as far as know the ISO OSI model is purely theorical, and has been
    > overcomed by TCP/IP, by a "grass root effect" (as they say in
    > English/American).


    This is not correct. All networking protocols in use build on the OSI
    layers since decades, and so does TCP/IP of course.

    >>> claimed that TCP/IP could fit into 10 to 24KB. I have already

    > disassembled a
    >>> dozen programs bigger than this size.)

    >> Such claims are pretty useless. There is no need to disassemble, the
    >> TCP/IP stacks for Z80 systems are available as source code.

    >
    > Yes, but probably in "C"... and, years after years, we have heard stories
    > here, in the comp.os.cpm Newsgroup, that C is not portable, or too big, or
    > too slow. In my opinion, Unix and C are viable only on a 16-bits+ computer.
    > For CP/M, I think that assembly language is the only solution (I searched
    > the Internet for "TCP/IP in Forth", but found nothing, to my surprise).


    I don't care much in which voodoo magic you believe. Fact is that C is
    the most portable programming language we have. Prove are sources older
    than 40 years that still compile on a 21th century system. Another prove
    is Linux and BSD, runs on almost every 21th century platform.

    DRI, the guys who designed CP/M obviously had a different opinion than
    you. Imagine, they used much PL/M instead of assembler to write CP/M,
    MP/M and lots of applications. Complete sources available, with scripts
    ready to compile the whole stuff and build bootable disks and use it.
    Want to know the URL, or does google work today for you too?

    Already in the 70th is was obviously figured that C is not too big. Tons
    of compilers available for real small machines like ancient CP/M
    systems. Most embedded controller programming is done in C since 20
    years or more, controllers smaller than a CP/M system.

    How many Z80 8bit UNIX's did you use? What are your experiences with
    this systems? Are they much slower than CP/M on that same computer, or
    maybe 2FAST4YOU? UNIX shell too complicated compared to the extremely
    minimalistic CCP?

    >> or Windows, I'll tell you a top secret. The company Zilog, the guys who
    >> invented the Z80 processor, still produce a Z80 CPU right now. The thing
    >> is called eZ80. This little thingy is a Z80 CPU (with some extensions)
    >> and a complete TCP/IP protocol stack put into the hardware. There is

    >
    > Yes, but this is not a Z-80 running at 4-MHz under CP/M, so it does not
    > count.


    I would suggest you read the eZ80 data sheet available from Zilog. This
    is a true Z80 processor, obviously.

    The processor can be clocked up to 50MHz, for slow guys like you I'm
    pretty sure that it also can be clocked with 4MHz.

    You deleted my link in your reply, look here: http://www.ez80sbc.com/
    The web page includes screen shots showing this eZ80 system running
    CP/M. Talk to Howard if he could offer you a machine running at 4MHz,
    I'd be surprised if you could manage your self to get it running at the
    desired speed.

    > Ok. I will now wait for the TCP/IP books. I think that everything that I
    > will find will be in "C", since Internet (and, hence, TCP/IP) is based on
    > some Unix standards. I already know how to write a library of macros


    This is totally wrong. The first TCP/IP reference implementation to
    prove that is works was written in C on a BSD UNIX system.
    The Internet and TCP/IP are based on international standards, the papers
    are called RFC (Request For Comment). This standards are free available
    to anyone and implementations exist in various languages and lots of non
    UNIX systems, even for CP/M, done more than 20 years ago.

    > emulating the printf function of C. For years, I have wanted a "C-to-BASIC"
    > File Converter... Maybe it is time to write it? (I searched "TCP/IP in
    > BASIC": no answer! It would be funny to be the first man to program a TCP/IP
    > in BASIC! This way, the source would be portable, and readable.)
    >
    > Yours Sincerely,
    > Mr. Emmanuel Roche, France


    Good luck, let us know when you're done. Would be great fun to finally
    have a 'truly portable' TCP/IP stack. Oh, and have a look at BASIC-E, it
    is written in PL/M, so chances are good that you get this one running
    everywhere then, NOT.

    Udo Munk
    --
    The real fun is building it and then using it...

  19. Re: How many SIG/M disks where there?

    Hello, Udo!

    > DRI, the guys who designed CP/M obviously had a different opinion than
    > you. Imagine, they used much PL/M instead of assembler to write CP/M,
    > MP/M and lots of applications. Complete sources available, with scripts
    > ready to compile the whole stuff and build bootable disks and use it.
    > Want to know the URL, or does google work today for you too?


    I am afraid that you are mistaken. I have disassembled PL/M code during more
    than 20 years, now. All application programs from Digital Research were
    written in PL/M. But all the serious stuff (like the BDOS, GSX, CP/NET and
    all the versions of CP/M) were written in the assembly language of the
    processor (the only exception seems to be CP/M-68K, which was written in C).
    (Have a look to what is now available.)

    > How many Z80 8bit UNIX's did you use? What are your experiences with
    > this systems? Are they much slower than CP/M on that same computer, or
    > maybe 2FAST4YOU? UNIX shell too complicated compared to the extremely
    > minimalistic CCP?


    About 10 years ago, after using my Epson QX-10 during 15 years, I wondered
    if I should switch to some kind of Unix. I studied what was available, then.
    I felt that Minix was the best. I got the book (1st edition) and, after
    studying it, wrote to Andrew Tanenbaum, since he was in Europe. At the time,
    I was totally alone in France still using CP/M, and all the magazines
    talking about CP/M were dying one after the other. Finally, I decided to
    stay with CP/M Plus. Several years later, Rlee Peters told me that some
    source code from Digital Research were starting to appear. Now, we have
    probably everything that survived. "Refinements? My friend, they are up to
    you!" (Gary Kildall, BYTE magazine) So, since nobody else will do it, I will
    do a TCP/M.

    As I wrote in 2005:

    > Should I remind you that you can do only 6 things
    > with a computer, be it a Sinclair ZX-80 or a Connection Machine:
    > 1) word processing
    > 2) programming
    > 3) spread-sheet
    > 4) database
    > 5) communications
    > 6) graphics
    > As long as those 6 needs are satisfied, the computer is a
    > useful tool. And there are/were standards for those 6 needs
    > under CP/M: WordStar is obviously the standard for word
    > processors; BASIC is obviously the standard for programming,
    > MultiPlan is obviously the standard for spread-sheet, dBase II
    > is obviously the standard for database; XMODEM is obviously
    > the (lowest common denominator) standard for communications;
    > and finally GSX is obviously the standard graphics system for CP/M.


    Because of the de facto standard of the Internet for communications, XMODEM
    needs to be replaced by TCP/IP, if we want to continue to use CP/M yet
    another 15 years.

    > Good luck, let us know when you're done. Would be great fun to finally
    > have a 'truly portable' TCP/IP stack. Oh, and have a look at BASIC-E, it
    > is written in PL/M, so chances are good that you get this one running
    > everywhere then, NOT.


    Yes, Adam Dunkels (creator of Contiki and uIP) explains that, because TCP/IP
    comes from Unix, it is big, slow, and complicated. Hence the need for uIP,
    which seems to be quite successful, according to the number of Web sites
    dealing with it (on Linux servers!). He got a Ph.D. Thesis for this.

    As for BASIC-E, it is me who retyped the Thesis describing its inner
    working, not you...

    Yours Sincerely,
    Mr. Emmanuel Roche, France




  20. Re: How many SIG/M disks where there?

    Found this on the Internet...

    Is there a German out there, who could explain us what this is saying?

    CPMNET 1.4
    ---------------

    Mit den vielen Änderungen und Ergänzungen an Firmware, Treibern und
    Funktionen gab es ganz oben im CP/M-Programm natürlich dann auch jede Menge
    zu tun.

    CPMNET.COM liegt jetzt in der Version 1.4 vor und hat einen Stand erreicht,
    welcher eine universelle Nachnutzung auf beliebigen CP/M 2.2 kompatiblen
    Systemen mit Z80 CPU gestattet. Es werden alle Funktionalitäten des
    KCNET-Interface zugänglich gemacht, man kann mit diversen Parametern des
    WIZnet TCPIP-Stack herumspielen und für den Einsatz im Netzwerk bringt es
    gleich die absolut notwendigen Sachen mit, so dass man sofort loslegen kann.

    Eine komplizierte Angelegenheit war der Umbau der Quelltexte und ihre
    Modularisierung für die universelle Nachnutzung. Die KC85 spezifischen
    Sachen befinden sich nun alle in der Datei CPMNETKC.INC, welche man einfach
    durch die Variante CPMNET.INC ersetzen lassen kann.

    Was genau gemacht werden muss, werde ich in einem getrennten Artikel
    schreiben. Dann sollte das Programm nach minimalen Anpassungen und erneuter
    Übersetzung sofort mit anderen CP/M 2.x kompatiblen Systemen funktionieren,
    welche direkten Zugriff auf die Z80-PIO und damit das KCNET-Interface haben.

    Der TCPIP-Treiber mit SOCKET-Schnittstelle ist ebenfalls in einer eigenen
    Datei CPMNETW2.INC und vollkommen hardwareunabhängig bezüglich des KCNET
    Interface-Treibers.

    Das eigentliche Programm wurde nun auch mit diversen Abfrage- und
    Testroutinen versehen, so dass notwendige Systembedingungen überprüft werden
    und bei Nichtvorhandensein eine Meldung erfolgt und der Programmstart
    abgebrochen wird. Auf dem KC85 System kann man stressfrei sowohl mit
    KCNET.KOP als auch KCNET.DRV arbeiten, die DRV-Variante wird bevorzugt, wenn
    beide Treiber geladen sind.

    Im Zuge der Überarbeitung zur Version 1.4 bekam CPMNET.COM ein weiteres Menü
    spendiert, so dass man jetzt passend zur Hard- und Softwarelogik auf 3
    Ebenen Funktionen oder ganze Programme aufrufen kann, ich fange mal von
    unten an.

    (There are also several long messages written by one "Ralf Kästner", which
    look interesting...)

    (You may remember that I proposed, several years ago, on the comp.os.cpm
    Newsgroup, that, since the Internet is based on the client/server idea, we
    should make such a "CP/M client" program, which would use an 'IBM Clown" as
    its server, to surf the Net...)

    Yours Sincerely,
    Mr. Emmanuel Roche, France




+ Reply to Thread
Page 2 of 6 FirstFirst 1 2 3 4 ... LastLast