Subject: Linux-Development Digest #736
From: Digestifier <Linux-Development-Request@senator-bedfellow.MIT.EDU>
To: Linux-Development@senator-bedfellow.MIT.EDU
Reply-To: Linux-Development@senator-bedfellow.MIT.EDU
Date:     Thu, 19 May 94 23:13:09 EDT

Linux-Development Digest #736, Volume #1         Thu, 19 May 94 23:13:09 EDT

Contents:
  Help with boot floppy containing bootp/tftp (Mike Bone)
  ET++ Application Framework (John Halar)
  32-bit Novell desktop OS combines Unix, DOS 7 (Eric Kimminau)
  Kernel 1.0.x: NET-2Debugged or NET-2E or ... ? (Bernd Anhaeupl)
  Re: TCP lockups through Cisco box (Bradley E. Smith)
  Re: Appletalk support? (Erik Troan)
  Kernel profiling (Hildo Biersma)
  Re: SIGHUP - Deep Kernal Guts question! (Alan Cox)
  Re: 32-bit Novell desktop OS combines Unix, DOS 7 (Alan Cox)
  Re: SIGHUP - Deep Kernal Guts question! (David Monro)
  Re: Anybody working on BSD dump porting? (Rob Janssen)
  Re: 32-bit Novell desktop OS combines Unix, DOS 7 (Matt Welsh)
  Re: X Programming Example Wanted ... (Matt Welsh)

----------------------------------------------------------------------------

From: bone@uncledad.nwscc.sea06.navy.mil (Mike Bone)
Crossposted-To: comp.os.linux.help
Subject: Help with boot floppy containing bootp/tftp
Date: 19 May 1994 10:33:53 -0500

I'm looking for a way to create a Linux boot floppy which runs bootp
and tftp. I want to use this with the Diskless Linux Xterminal package
instead of the boot floppy image that comes with it. The problem with
the current boot floppy is that 1) you need to hard code your ethernet
card type, interrupt, and other settings in the image and 2) you need
Turbo C to compile the boot image for your various ethernet cards. With
a Linux boot floppy, the kernel could probe for the network card so one
floppy would work for several different cards and settings. Also, it
could be compiled using only free software.

If anyone can offer any help or pointers to information on creating
boot floppies and the availablility of bootp and tftp source to run
under Linux, I would really appreciate it.

Thanks in advance,

Mike Bone       (bone@uncledad.nwscc.sea06.navy.mil)

------------------------------

From: j.halar@uow.edu.au (John Halar)
Crossposted-To: comp.os.linux.misc
Subject: ET++ Application Framework
Date: Mon, 16 May 1994 15:24:54

I seem to remember a couple of months ago when I read an news article about 
somebody trying to adapt the ObjectWorks ET++ Application Framework for C++ 
to run on Linux under X11/R5.

A couple of questions:

1) Is this software commercial or shareware/public domain??
2) If it is, has anybody successfully compiled it and had it working 
   for linux??
3) Is there an ftp site from which I could get it from?

Any answer to these questions will be appreciated.

Cheers

John Halar
E-mail: j.halar@uow.edu.au



------------------------------

From: ekimmina@pms709.pms.ford.com (Eric Kimminau)
Subject: 32-bit Novell desktop OS combines Unix, DOS 7
Date: 19 May 1994 15:15:17 GMT

Reprinted verbatim from an article by Anne Knowles in the May 16th,
1994 issue of PCWeek Magazine without permission of Ziff publishing
Corp. The article appeared on page 1, with a continuation on page 14.

32-bit Novell desktop OS combines Unix, DOS 7
by Anne Knowles

Novell Inc. is targeting July for the launch of its 32bit multitasking
desktop operating system called Corsair. Tentatively priced at $99.00,
the operating system combines Novell DOS7 with a kerrnel based on
Linux 1.0, a Unix clone for Intel Corp. PCs that is freely available
over the Internet, according to a Novell internal document examined by
PCWeek.

The document describes Corsair as an "enterprise networking client
fully integrated into NetWare and Unix networks."

Code named Expose', the software will run on systems based on Intel
and PowerPC processors and "other leading RISC platforms," according
to the document. Users will be able to run Windows, DOS, and Intel
Application Binary Interface compliant Unix applications
simultaneously.

A "virtual world" 3-D graphical interface will emulate real-world job
functions, thereby reducing training costs and increasing
productivity, the document states. Corsair will also feature a
World-Wide-Web browser that may be based upon, or similar to, Mosaic,
a popular interface that is also freely available on the Internet.

"All elements of the (See Corsair, Page 14) Corsair user interface are
remotely manageable, further reducing support and administration
overhead," the document states. However, there were no details
regarding which systems-management protocols, platforms, or
applications are supported.

A Novell spokeswoman cautioned that there is no set release date for
Corsair. And since corporate buyers must already sift through several
commercial operating systems, some users questioned where the new
software would find a home.

"It would be something we'd look at, but I don't really see a
compelling reason to change to it," said Mark Butler, systems
administrator at the state of Washington's Department of
Transportation, in Seattle. "We are stabilized on Windows 3.1," said
Butler, who supports about 1000 NetWare and Windows 3.1 users.

"We'd have to be able to support [Corsair] without any problems," he
added. "We'd accept some performance degradation [on Windows
applications' performance], but we want plug and play. If it requires
a lot of complexity, we're not interested.

Another NetWare user had a siimilar reaction.

"I'd be interested in implementing it as an option on a subset of
machines to provide an interface to the outside world," said Peter
Brantley, MISmanager with the AIDS office of the Department of Public
Health for the city and county of San Francisco. "But I couldn't adopt
it as a desktop because of lack of funding at the department"

"Sounds great, neato, but I couldn't see it being more than a couple
of machines," Brantley added. "Anything more than that seems
foolhardy."

According to sources, Novell Chairman Ray Noorda picked the Corsair
name for what has been his pet project. At one point, the Provo, Utah,
company decided not to continue developing it, but Noorda put forth
his own money to keep work going.

Corsair will be available through all of Novell's sales channels,
according to the document.

Additional reporting by Mary Jo Foley and Eric Smalley.
=========================end of article==========================

Heres my .02. This is a huge HUGE bonus for the Linux community.
Granted Novell's GUI and DOS 7.0 are their property, and will not be
in the public domain, but more than likely they will make the patches
available on the net, and for some modest fee probably distribute the
binaries encryted on the net, somehow. This is my guess completely.
Look at the price. Novell DOS 7 is selling for like $59 or $69.00
right now and includes a whole boatload of stuff INCLUDING the newest
version of Stacker. Lets take a really wild stab at something and hope
that Novell has Linux/Corsair running on a stacked partition. Sounds
like something really cool to me!

Lets also look back to the first article about Expose' where they said
that Expose was already running the entire suite of MSOffice apps
without MSWindows installed. Now look back at this latest areticle
where they suggest that Corsair will run DOS, Windows and UNIX
binaries all at the same time. Guys, THIS is want I want my PC to do!
I don't know about the rest of you. And for $99.00, if that includes
GOOD online documentation and a couple printed reference manuals/user
guides, you better believe Im going to buy a copy the day its
released.

As a side note, on page 55 of the same May 16th PCWeek, there is an
announcement of a "Corsair SDK"  in an article on "Windows-on-Unix
emulation options grow with upgrades" by a company called Multiport,
Inc. in Palo Alto, CA. (415)494-2782. This SDK is supposed to allow
running and porting Windows applications to Unix.

Tell me the 2 are not related! Novell is about to give us all a 32bit 
UNIX/DOS/Windows box for $99.00. Can anyone else remember how much
MSWindows sold for when it was released? Then lets talk about the
other DOS and windows emulators for sale on othe Unix platforms and
how much JUSTthe OS costs, not to mention the 300-500.00 for the
emulator, PLUS windows, plus the compiler, plus all the other stuff
that we should STILL be able to get and run for free from the
internet. Does anyone else besides me want to be able to run all the
Unix things ported to linux at the same time as all the things that I
used to be able to run in DOS/Windows? I havent booted DOS on my PS at
home in almost 4 months, and I miss Excel. I want my Andrew Tobias'
Managing your Money, back. I want to play DOOM again! 

Think about the possibilities. Linux, DOS and Windows all on the same
desktop, in their own windows. Ill be waiting over here in front of
the window marked "Corsair, coming soon".


-- 
Eric Kimminau                       Workstation Systems Department
313-322-3431                        Product & Manufacturing Systems
ekimmina@pms709.pms.ford.com        Ford Motor Co.
Planning and Implementation         "Not an official Ford Spokesperson"
TIP#111

------------------------------

From: anhaeupl@late.e-technik.uni-erlangen.de (Bernd Anhaeupl)
Subject: Kernel 1.0.x: NET-2Debugged or NET-2E or ... ?
Date: Mon, 16 May 1994 17:52:36 GMT


After upgrading from 0.99.14 to 1.0.7, I mentioned some problems with network 
connections. I thought, perhaps it would be better to install new versions of
the network daemons, got the sources from 
            nic.funet.fi:/...../PEOPLE/linus/Network/NETbsd 
(?; I do not remember the exact pathname), compliled them with gcc 2.5.7 and 
libc 2.5.8 (LST 1.7 Distribution), but the things got even worse. Therefore my 
questions:

      Are those sources compatible with the networking code in 1.0.7; where
      can I find the __sources__ for the networking __programs/daemons__
      for 1.0.x ?

      What are the differnces between the networking code in 1.0.7 and
      in NET-2E; which one is more stable? 
      
Bernd.

PS: Here are the symptoms:

        1.)   Using inetd, in.rlogind, rlogin ... from LST 1.7:

              a)

              rlogin localhost
              exit
              rlogin localhost

              The connection for the first rlogin is esteblished immediately,
              but the second one needs some time  to succeed. When I abort 
              the second rlogin with SIGINT (^C) a third rlogin succeds 
              immeditately again.
        
              b) 

              rsh localhost date
              rsh localhost date
              
              No problems, seemes to work (Except for some bugs in in.rshd:
              wrong or no HOME environment variable) .

        2.)   After installing inetd, in.rlogind, in.rsh, ... from 
              nic.funet.fi:.../PEOPLE/linus/Network/NETbsd(?)

              a) At first rlogin seemed to work. (Only about 1 s delay for 
                 the following rlogins).

              b) rsh now needs about 60s to complete. (Immediate output, but 
                 then .... nothing for 1 minute). If you enter RETURN within
                 this minute, then rsh completes immediately.

              c) rlogin localhost 
                 exit
                 (now wait about 10 or 20 seconds)
                 netstat
                 rlogin localhost

                 This time the rlogin also needs about 1 minute or even more
                 to succeed. The first netstat output reads (something) like

                 Proto  Recv-Q   Send-Q   Local Address   Foreign Address  (state)       

                 tcp     2        0       localhost:1023  localhost:shell   TIME_WAIT
                 tcp     1        0       localhost:shell localhost:1023    TIME_WAIT


                 If you call netstat from another virtual console you get something
                 like:
                 
                 Proto  Recv-Q   Send-Q   Local Address   Foreign Address  (state)       

                 tcp     0        2       localhost:1023  localhost:shell   SYN_SENT

                 which lasts for about 1 minute or more.




--
Bernd Anhaeupl                  Tel.:  +49 9131 857787
LATE - Uni Erlangen                     
Cauerstr. 7                     Email: anhaeupl@late.e-technik.uni-erlangen.de
91058 Erlangen

------------------------------

From: brad@bradley.bradley.edu (Bradley E. Smith)
Crossposted-To: comp.os.linux.help
Subject: Re: TCP lockups through Cisco box
Date: 18 May 1994 19:01:24 -0500

Not that it matters, but i just posted a problem with a cisco
terminal server.  linux tcp debugging messages show a problem with
sequence numbers.  hope someone can help.
-- 
Bradley Smith                    brad@bradley.edu ---  309-677-2337
Network & Technical Services @ Bradley University, Peoria, IL

"It's amazing how much scrap metal you get from 4 cans of beer"

------------------------------

From: ewt@merengue.unc.edu (Erik Troan)
Subject: Re: Appletalk support?
Date: 19 May 1994 19:42:38 GMT

In article <1994May18.141844.15810@uk.ac.swan.pyr>,
Alan Cox <iiitac@uk.ac.swan.pyr> wrote:
>In article <9413822.2009@mulga.cs.mu.OZ.AU> djh@cs.mu.OZ.AU (David Hornsby) writes:
>>Porting CAP in "Native EtherTalk" mode to Linux is not feasible unless
>>

I think this already has been ported. Take a look at /pub/Linux/system/Network
on sunsite - there are two files in there which begin with cap that ya'll 
may find interesting.

Erik

-- 
===========================================================================
"I'm not like that -- except when I am"   ewt@sunsite.unc.edu  = Erik Troan
                                          sasewt@unx.sas.com
    - Nora from "Pump up the Volume"

------------------------------

From: biersma@dutiws.twi.tudelft.nl (Hildo Biersma)
Subject: Kernel profiling
Date: Tue, 17 May 1994 22:08:47 GMT

Hi all,
This may be a FAQ; if so, please direct me to the proper documentation.

My question concerns kernel profiling.  When reading through the
kernel sources, I noted that a buffer of longwords is created
that counts the number of times that each 4-byte area in the kernel
is 'current' code.  This buffer can be mapped using the kernel symbol table.

My question is: are there any existing tools to extract/display this
data, preferably sorted by use-count?  If not, I'll write one myself;
but, being lazy, using existing tools seems a Good Idea.

Thanks for any help,
Hildo Biersma
-- 
Hildo Biersma

------------------------------

From: iiitac@uk.ac.swan.pyr (Alan Cox)
Subject: Re: SIGHUP - Deep Kernal Guts question!
Date: Thu, 19 May 1994 19:15:37 GMT

In article <1994May19.014201.14298@loreli.ftl.fl.us> nate@loreli.ftl.fl.us (Sean Puckett) writes:
>Why is it in the specs?
>
>Some lame reason like "because it has always been done that way" is not
>acceptable.

NOPE: It's in the specs for good reasons... kernels should not set policy.
If you want people to get the boot then the shells you choose to issue them
with should follow the policy. In addition shells can do cleverer things
such as killing other process groups they have created. Fix your shell.

Alan


------------------------------

From: iiitac@uk.ac.swan.pyr (Alan Cox)
Subject: Re: 32-bit Novell desktop OS combines Unix, DOS 7
Date: Thu, 19 May 1994 19:22:01 GMT

In article <2rfvq5INNrlp@ope001.iao.ford.com> ekimmina@pms709.pms.ford.com (Eric Kimminau) writes:
>version of Stacker. Lets take a really wild stab at something and hope
>that Novell has Linux/Corsair running on a stacked partition. Sounds
>like something really cool to me!

This is why I find the whole thing so improbable in some ways. There is no
way Novell could add Stacker support to the Linux kernel environment without
in effect releasing a GPL'd stacker equivalent. Given that stacker are
fairly protective of their code and ideas this is unlikely. Also the GPL
prevents patented items being GPL'd and the patent used to lock them down

In addition Novell would have to real how NCP works to write a Novell
client for Linux (ie NETX equivalent). This is I admit not really a secret
any more because people just figured out the packets. They couldn't however
include a client as part of the Linux kernel that did packet signing even
if they did this because of the RSA algorithms used for packet signing 
being patented.

Finally Novell have just got involved with AT&T to create a 'rival 8)' to
the internet running IPX for real commercial work using Novell stuff.
I have my doubts..

Alan

------------------------------

From: davem@extro.ucc.su.OZ.AU (David Monro)
Subject: Re: SIGHUP - Deep Kernal Guts question!
Date: Wed, 18 May 1994 16:02:02 GMT

[lots deleted..]
>do_exit sends SIGHUP to the group if (and only if)
>there are stopped processes in the group,

>it is only running tasks that i find detached from the terminal after
>the leader exits. but thats even if there are stopped jobs.
>(i dont know if the bit about stopped jobs in do_exit has anything
>much to do with this problem

Does this explain why if I kill an xterm/rxvt running less, the less
process goes beserk and eats all my CPU? If so, this is bad, and could
cause real havoc on a multiuser machine (user logs out, process goes
beserk, nobody can kill it till root comes along...)

        David

------------------------------

Crossposted-To: comp.os.linux.help
From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: Anybody working on BSD dump porting?
Reply-To: pe1chl@rabo.nl
Date: Wed, 18 May 1994 22:29:00 GMT

In <2rdcfn$lu4@csugrad.cs.vt.edu> mmead@csugrad.cs.vt.edu (matthew "pipeman" mead) writes:

>In article <CpyxL3.AzH@pe1chl.ampr.org>, Rob Janssen <pe1chl@rabo.nl> wrote:
>>Because of the many filesystems that Linux supports, trying to port
>>something like "dump" is not simple, and IMHO not a good idea.
>>
>>You are better off making backups with "tar" or "cpio".

>       What if you need to backup over the net?  There's rdump, but I assume
>that won't work for the same reasons that dump won't ... how would dumping to a
>disk mounted on a machine elsewhere on your local net be accomplished? :-)

GNU tar can backup over the network as well.
Of course, any backup program can be made to work over the net...

Rob
-- 
=========================================================================
| Rob Janssen                | AMPRnet:   rob@pe1chl.ampr.org           |
| e-mail: pe1chl@rabo.nl     | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU     |
=========================================================================

------------------------------

From: mdw@cs.cornell.edu (Matt Welsh)
Subject: Re: 32-bit Novell desktop OS combines Unix, DOS 7
Date: Thu, 19 May 1994 16:22:04 GMT

In article <2rfvq5INNrlp@ope001.iao.ford.com> ekimmina@pms709.pms.ford.com (Eric Kimminau) writes:
>Heres my .02. This is a huge HUGE bonus for the Linux community.
>Granted Novell's GUI and DOS 7.0 are their property, and will not be
>in the public domain, but more than likely they will make the patches
>available on the net, 

I find it highly unlikely that Novell's support for running MS-Windows
apps were integrated into the Linux kernel. Therefore, it's quite possible
that many of these "fantastic" features that we're hearing about won't
be freely distributable at all. Novell is only required to distribute 
code for free that is based on GPL code, e.g., any modifications to the
Linux kernel itself. This does not entail that the entire system will be
placed under the GPL.


------------------------------

From: mdw@cs.cornell.edu (Matt Welsh)
Subject: Re: X Programming Example Wanted ...
Date: Thu, 19 May 1994 16:28:36 GMT

In article <1994May19.153800.1@condor> kempsonc@logica.co.uk writes:
>I have just started some X development on my new                               
>Linux workstation and I have come across the following problem:     
>  
>   Many of the Xt Calls I am used to are no longer available       
>   eg. XtVaAppInitialize ...  Where are these?  I note that    
>   XtAppInitialize is available though.         

Interesting. Try the following short program, which displays your
colourmap (8 planes) in a box widget. It should compile directly under
Linux.

Compile with:
        gcc -o xfoo xfoo.c -lXaw -lXt -lXmu -lX11
        

This code is placed in the public domain. :)
---
#include <stdio.h>
#include <X11/Intrinsic.h>
#include <X11/StringDefs.h>
#include <X11/Xaw/Command.h>
#include <X11/Xaw/Box.h>
#include <X11/Xaw/Viewport.h>

extern void myrefresh();

GC theGC;
int myscreen; 
Display *mydisplay;
Colormap mycmap;

static XtActionsRec myactions[] = {
  {"myrefresh",myrefresh},
};

static char mytranslations_str[] ="#override \n\
    <Expose>: myrefresh()";

XtTranslations mytranslations;  

void myrefresh(Widget w, XEvent *event, String *s, Cardinal *np) {
        int i,j,k;
        i=0;

        for (j=0; j<16; j++) {
          for (k=0; k<16; k++, i++) {
            XSetForeground(mydisplay,theGC,i);
            XDrawLine(mydisplay,XtWindow(w),theGC,k*10,j*10,(k+1)*10,j*10);
          }
        }
}

int main(int argc, char **argv) {
  XtAppContext context;
  Widget foo, toplevel;

  toplevel = XtVaAppInitialize(&context,"XFoo",NULL,0,&argc,argv,NULL,NULL);
  foo=XtVaCreateManagedWidget("foo",coreWidgetClass,toplevel,NULL);
  XtVaSetValues(foo,XtNheight,160,XtNwidth,160,NULL);

  XtAppAddActions(context,myactions,XtNumber(myactions));
  mytranslations=XtParseTranslationTable(mytranslations_str);
  XtOverrideTranslations(foo,mytranslations);


  mydisplay=XtDisplay(toplevel); 
  myscreen=DefaultScreen(mydisplay);
  mycmap=DefaultColormap(mydisplay,myscreen);

  theGC=XCreateGC(mydisplay,DefaultRootWindow(mydisplay), 0,NULL);
  XCopyGC(mydisplay,DefaultGCOfScreen(XtScreen(foo)),
    GCForeground|GCBackground|GCFunction,theGC);
  XSetFunction(mydisplay,theGC,GXcopy);
  XSetLineAttributes(mydisplay,theGC,10,LineSolid,CapButt,JoinMiter);


  XtRealizeWidget(toplevel);
  XtAppMainLoop(context);
}

------------------------------


** FOR YOUR REFERENCE **

The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:

    Internet: Linux-Development-Request@NEWS-DIGESTS.MIT.EDU

You can send mail to the entire list (and comp.os.linux.development) via:

    Internet: Linux-Development@NEWS-DIGESTS.MIT.EDU

Linux may be obtained via one of these FTP sites:
    nic.funet.fi				pub/OS/Linux
    tsx-11.mit.edu				pub/linux
    sunsite.unc.edu				pub/Linux

End of Linux-Development Digest
******************************
