Conference hydra::amiga_v1

    Hmm... I wonder if you really have to have an ethernet card.  Shouldn't
    there be some way of playing X on some sort of point-to-point link?
    It would have to be at least 9600 baud, according to the X-windows
    documentation I've seen, but that's not as impossible as it used
    to be.  
    
    Does X-windows have it's own comm. protocal?  I thought DECwindows
    used DECnet.  Come to think of it, there's no reason why it would
    have to.  Other protocols can and do co-exist with DECnet on the
    ethernet (LAT, TCP/IP, ...)  But if it ran it's own communication
    protocol, it seems logical that it be designed to run through some
    other communications port (i.e. serial port) or through an ethernet 
    card.  
    
    Is there some design restriction that I'm not aware of?  Somebody
    with some technical muscle in this area care to comment?
    
    
    				John

X does not specify what the low level "wire protocol" is. It gives a set of
events that pass between a client and a server which may or may not be seperated
by a "network" (i.e. may both be on a local machine)

DECwindows can use various low level protocols (DECNET, TCP/IP, LOCAL). The 
default is to use a LOCAL transport for client/server on the same machine and 
use DECNET otherwise. Several people are using TCP/IP on both VMS and ULTRIX.

If you wanted to write an X Transport for RS232, it would work there as well.

John (who may be way off!)

    Since Xwindows and TCP/IP both run on the Amiga and most TCP/IP
    implementations (not sure about all) support RS232, an Amiga should
    work fopr this.  DECnet is being developed by some company for the
    Amiga, however I don't know if the people who ported Xwindows to
    the Amiga have any plans on integrating it with the DECnet release.

Re: .1

I read someone on Usenet that quoted Dale Luck as saying that X windows
would run over the Amiga's serial port.  I guess that Dale is hoping that
Amiga 500s will become low cost X terminals (and that everyone will buy
his software).

!!!  You mean I can log into my workstation (WJG::) from home and become
an X server?  

Even at 2400 baud, this might be nice!

    Hmm... Somebody help me here.  I haven't been able to play with
    a VAXstation since I moved to the field about a year ago.  If I
    get X up and running on my Amiga and get some communications protocol,
    say TCP/IP, going between my Amiga and a VAX in the office, will
    I be able to run the VMS mail application on my Amiga?  And NOTES
    and those other goodies?  ... I thought that was how X was supposed
    to work, but I haven't played with it since late 1986, and DECWindows
    was still in the early development stages then.  
    
    Correct me if I'm wrong, please... With X, you're supposed to be
    able to be able to run the application on another machine, say the
    DECwindows Mail Application on a remote VAX, and just get the display
    'mapped' to your X-windows terminal, even if that terminal/computer
    can't run the VAX code itself.  So with X, we wouldn't have to port
    VAX graphics applications to the amiga before we ran them.  We could
    just run them on the VAX.   Is this right, or have I taken an extended
    vacation from reality again.  (I HATE it when that happens.  ;-)
    
    
    					John

   This is far from the definitive answer, but I'll give you some info.
   DECwindows output can be displayed on a system while the processing is
   remote. Right now I am displayiing DECwindows notes on a workstation
   while the bulk of the processing, notes access etc. is happening on
   a batch queue on a cluster. I use the DCL SET DISPLAY command
   to point to the VAXstation. So the answer to part of your question
   is yes. You can run remotely and display locally. However, I have
   the full DECwindows kit installed on my workstation, and do not know how
   much of it is being used for the display task. I suspect that an Amiga with
   X-windows installed may not understand a DCL command called SET DISPLAY. ;^}

   Dave

To make it (.7) a little more clear, the SET DISPLAY is set from  the cluster, 
but checks with the display "server" at some point. The terminology gets tricky.
Compute server = VAX, and Display server = Amiga?  There will be no more 
computers, only servers ;^}.

So you need "only" the display software on the Amiga. The PCDECWINDOWS 
conference on MOSAIC:: would be a good source for more info since DEC is
implementing this on PCs.
Dave

    RE. 8
    
    Using the term "servers" can be very confusing.  A more precise
    terms are; client and server.  The client being where the application
    is executing, while the server is where the output is displayed.
    As indicated the client and the server can either be the same machine
    or different.  The client and server communicates with each other
    by TCP/IP or DECNET.  TCP/IP and DECNET are both supported in the
    X11 protocol.
    

    I did a little digging and what I've found seems to bode well for
    an Amiga X workstation.  I talked with one of the specialists who's
    been gearing toward DECwindows and he thinks it should do most if
    not all of what I want.  He showed me some DECwindows architecture
    stuff which seems to back up his claim.  I thought I'd pass along
    what I've picked up so far, in case anybody else is interested.
    
    There's an interesting architectural diagram that's pertinant:
    
         -------------------------------------------------------
         |                  Application                        |
         |         ---------------------------------------------
         |         |              |
         |      ----              |
         |      |                 |
         |   ----   --------------------------------------------
         |   |      | Industry Standard Libsraries | Extension |
         -----   ----------------------------------| Libraries |
                 |          DECtoolkit             |           | <<<<<<<<<<
              -------------------------------------| -PEX      | < CLIENT <
              |            Xtk (Intrinsics)        | -PostScrpt| <<<<<<<<<<
         ------------------------------------------| -Imaging  |
         |                     Xlib                |           |
         -------------------------------------------------------
    	                          |
         -------------------------------------------------------
         |                 Transport Mechanism                 |
         -------------------------------------------------------

    ........................... X11 Wire Protocol ............................
    
         -------------------------------------------------------
         |                 Transport Mechanism                 |
         -------------------------------------------------------
                                  |
         -------------------------------------------------------
         |                                        | Extensions |
         |             X Server Kernel            |            | <<<<<<<<<<
         |                                        | -PEX       | < SERVER <
         |                                        | -Postscrpt | <<<<<<<<<<
         |                                        | -Imaging   |
         -------------------------------------------------------

    
    As this diagram illustrates, the VAX-specific application, the
    DECwindows toolkit, and whatever system service/RTL/machine specific
    routines are required for that application are all contained on
    the CLIENT VAX.  In effect, the CLIENT ships only graphics messages
    to the AMIGA SERVER... X messages telling the SERVER to make a new
    window, print text at the current cursor position, move the cursor,
    etc.
    
    So in a sense, your X SERVER is becoming the workstation equivalent
    of a terminal emulator.  The application sends output information
    to any X server (theoretically), and that X server then translates
    this information into a display using whatever resources it has
    at it's disposal.  Thus hardware type is no longer as much of a
    concern.
    
    Anyway, that's how it's explained to me.  This would seem to indicate
    that reply -.2 or -.3's DCL "SET DISPLAY" command would get executed
    on the VAX CLIENT (if it does what I think it does), or get translated
    to the X equivalent and sent to the SERVER where it is then translated
    to the machine (AMIGA) specific equivalent.  Either way, having
    an X-windows SERVER instead of a DECwindows server should make no
    difference.  
    
    Lets hope things work as theorized.  It'd be great to be able to
    turn your VAX into a color workstation by using your Amiga.  
    
    
    					John

    There was blurb in Digital News last week about an X terminal which
    DEC is supposedly working on.  Sounded like a pretty good idea, but
    I'll bet it costs a lot more than a A500 + X11.  Of course so does a
    VT125...
    
    - steve

John,

You've got it figured just right.  The VMS SET DISLPAY command not only points
the X output to a remote workstation (the server), but also indicates the
protocol (DECnet or TCP/IP) to be used.  I don't remember how right now, its
back at the office.

Wasn't Dale Luck porting the X server to the Amiga?  It would sure be neat
to be able to do it.  Now all we need is a way to hook up a color 19-inch
monitor!!!

...richard

    CBM-AMIGA have just announced their 2084, which is a 19" b+W monitor,
    so they're getting close.
    
    Martin

    Now that Amiga has shipped its millionth machine, I would expect
    larger monitors to start showing up.  The operating system can handle
    a larger display (1024 X 1008 max. is what the chips were designed
    for, right?) so all I would think they'd have to come up with is
    a card to drive the larger monitor.  Sorry you guys with flickerfixer,
    it'd probably have to plug into the video slot... but then you guys
    like Ed A. and John S. would probably just buy a second A2000 to
    plug this video card into, eh?
    
    Re: -1
    	But one of the nice things about X-windows is that at least
    it'll WORK with your normal monitor.  Your X server software will
    take care of mapping windows to the space you have.  (Don't know
    what would happen if you tried to create a window that was bigger
    than your physical screen size.  X would probably just trim it back
    and report its actual dimensions to the CLIENT.  The more I look
    into X-windows, the more I like it!)  Of course, I agree that a
    larger display is an incredibly wonderful thing... (Oh no... I'm
    sounding more and more like a hardware junkie every day!)
    
    
    				Glazy-eyed and detached looking,  8'!
    
    				John

    The only "gotcha" in this whole beautiful new world of DECwindows
    is that _officially_ DEC does not support using TCP/IP as the
    transport protocol for _VMS_DECwindows.  Unofficially, there are ways
    to make it work with the VMS/ULTRIX Connection product, but for now
    unless you've got DECnet on the Amiga (tested to work with your
    Amiga X-Server) your VMS DECwindows applications won't work (unless
    you happen to be a X11/TCP/VMS guru, a rare bird).
    
    Next release of DECwindows may fix this with Wollongong's TCP. 
    But until then, you may want to look into ULTRIX for client
    applications, like DECwindows Mail; ULTRIX supports TCP for DECwindows
    just fine.
    
    <<< Tim >>>

RE: .8 and .9
It may be a nit, and it is a tangent, but using the terminology server for
the display and client for the "other system" is confusing. I know that 
explanation is technically correct. Since many people think of the "big"
computer as the server and the "small" computer as the client, this X-windows
terminology has confused quite a few. Take a look in the DECwindows conference.
What will we do when we have font servers, print servers, display servers and
applications running on multiple CPUs or systems? The important part of the 
description is the action (display, compute, print etc.), not who is client to 
whom or server for what . It all loses meaning with distributed computing. 

    Confusing?  You should see my customers when this discussion comes
    up!!!  They throw their hands up and howl!!!!
    
    But we didn't invent the terminology.  Officially, it was a joint
    development of IBM, MIT and DEC.  Unofficially, it was IBM's money
    with MIT's ambition and DEC's engineering talent.  So the blame
    can be pretty-well spread around, and anybody who supports the X
    Window System (its official name) will be in the same terminology
    pickle.
    
    Oddly enough, the more you work with it, the more the client-server
    business makes sense.  When you run a distributed database program,
    it will look for the database server where the files are kept, because
    it "needs" the files in order to function.  If you don't have access
    through the server's security, the server won't let your database 
    program access it.  Same is true for the X Window System.  When you 
    run a program with an X-based distributed graphical user interface, 
    it "needs" a place to display it and get user input from (mouse and 
    keyboard).  It "needs" a display server.  And if the X client program 
    doesn't have access through the server's security, the display server 
    won't let the client program have access to it.
    
    I think what most people don't like is the implication that since
    they are sitting at the server, they are at the beck and call of
    some piece of software.  Sort of like the computer is running the
    people.  But, then again, I think we're being programmed by our
    computers all the time, anyways...8^).
    
    <<< Tim >>>

    re: .14---Actually, I don't have a flickerfixer, or a multisync
    monitor.  I'm waiting for the new chip set, which will fix flicker the
    real way, by doubling the pixel rate.  Maybe Christmas 1990.
        John Sauter

    re: .19---I doubt it.  I don't think the A1000 has enough address
    lines, whereas the A2000 and A500 were designed to accept the new
    chips, someday.
        John Sauter

    
    I'll bet anyone a donut that some hack will figure out a way to smash an
    ECS board into an A1000.  Even if only 5% of 150,000 A1000 owners paid
    $100 for such a board, it would make some decent money for someone.  In
    fact. I think someone quoted Jay Miner as saying sure it was possible,
    but very messy.
    
    Ed.

    I spoke to CBM reps at a recent show about this. I had heard rumours
    that it wouldn't be available for A1000's and early A2000's, and
    was not impressed (Having spent a lot of money on an A2000 - which
    is supposed to be the 'latest' Amiga).
    
    The CBM rep told me that only the A500 and,so called, B2000 would
    be supported by ECS, and that it was not possible to make it work
    with the A1000 or the early A2000.
    
    Martin

Re: .22

The CBM rep was right: the new chips only fit the sockets in the A500
and B2000.  The original Amiga 2000, which was sold in Europe, used
the same chips as the Amiga 1000.  The original Amiga 1000 Agnus chip
had a different number of pins than the new "fat" Agnus in the Amiga
500 and the Amiga 2000 "model B".

There is a quick way to tell if you have a B2000.  If your 2000 has
three RCA jacks (left audio, right audio, and monochrome video) on the
back, then you have a B2000.  If your 2000 only has two RCA jacks
(left audio and right audio), you have an A2000.