Subject: Linux-Development Digest #299
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:     Wed, 8 Dec 93 11:13:10 EST

Linux-Development Digest #299, Volume #1          Wed, 8 Dec 93 11:13:10 EST

Contents:
  SCO binaries ??? (Andreas Nolte)
  Re: Notif - The Free MOTIF Clone Project (Matt Bonner)
  Let's nominate Marc&Co. for Genius award  (Bruce White)
  Re: Merry $*!@ing Christmas! (Benjamin Z. Goldsteen)
  (none) (Theodore Ts'o)
  Re: Lots more stats in /proc aka /proc/loadavg (Donald J. Becker)
  Re: Always -- IN-2000 (Mark Buckaway)
  How about C.O.L.kernel ? (Michael A. Irons)
  Re: How about C.O.L.kernel ? (Kai Petzke)
  Re: Merry $*!@ing Christmas! (Dave Sill)
  Re: EXT2 filesystem unrecoverable error - HELP!!!!! (Holm Sieber)
  [Q] ISODE port for Linux ?? (Ulf Karbstein)
  Re: Appletalk and LINUX Update (Neal Becker)

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

From: i17i@zfn.uni-bremen.de (Andreas Nolte)
Subject: SCO binaries ???
Date: 7 Dec 1993 09:21:36 GMT

        Hi there!

        I've heard rumors about SCO binaries (or other commercial Unix bimaries)
        running under Linux. Could please someone tell me, if this is true, and
        if it's any stable, since I'd like to work with WordPerfect for Unix 
        under Linux. Right now wordprossesing is the only reason for me working
        with hated crashing-all-the-time slow Windows, and I hate it.

        Please someone comment on that,

                Thanks,

                        A.N.


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

From: mattb@sdd.hp.com (Matt Bonner)
Crossposted-To: comp.windows.x.motif
Subject: Re: Notif - The Free MOTIF Clone Project
Date: 7 Dec 1993 22:16:02 GMT

rick@digibd.com (Rick Richardson) writes:

>> pfox@Ingres.COM (Paul Fox) writes:
>> 
>> >This is going to be a Motif clone hopefully eventually conforming to
>> >Motif 1.2 PD specification.
>> 
>> You should really look into using wbuild 2.0 to create the widgets
>> for NOTIF.  The widgets come out of it self-documenting, and it
>> creates all the unreadable boilerplate C code for you.  It can
>> be gotten from the FWF at a.cs.uiuc.edu:/pub/FWF.

This sounds pretty cool, as long as there's a way to edit a widget
later on.  ie, can I create a widget with wbuild, realize I messed
something up then go back and fix it?

Has OSF issued any sort of statement about people working on this
project who have seen the Motif source?  If not, could they?

I also wonder if there's any room to improve on Motif without breaking
existing apps (assuming they don't illegally delve into the private
data of widgets).  One of the complaints I have heard about Motif is 
the somewhat messy class hierarchy, making subclassing from Motif widgets
difficult.  Just wondering if there's any room to clean that up in
the design phase...

Thanks,
Matt

Matt Bonner                  |  Email: mattb@sdd.hp.com
Hewlett-Packard Company      |  Snail: 16399 West Bernardo Drive
Ink-jet Components Division  |         San Diego, CA  92127-1899

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

Crossposted-To: comp.infosystems.www,comp.windows.x,comp.windows.x.i386unix,comp.windows.x.motif,gnu.misc.discuss,comp.sources.d
From: bw@TorreyPinesCA.ncr.com (Bruce White)
Subject: Let's nominate Marc&Co. for Genius award 
Date: Wed, 8 Dec 93 01:57:35 GMT

>Perhaps we need some positive feedback to Marc, Rob and Eric to balance the
>negative crap they've been getting recently.  Thanks to all of them for
>providing some excellent software, and for dragging the HTTP and HTML
>protocol suite into a damn fine system.  I make no contribution to their
>financing whatsoever, being a UK person, and still they let me use the fruits
>of their time and effort at no cost, and provide a friendly and useful 
>support function whenever things do go wrong.

Amen. I'd like to nominate the whole staff for a Macarthur (sp?) award.
They've created a superb product.

Bruce



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

Crossposted-To: comp.infosystems.www,comp.windows.x,comp.windows.x.i386unix,comp.windows.x.motif,gnu.misc.discuss,comp.sources.d
From: ben@rex.uokhsc.edu (Benjamin Z. Goldsteen)
Subject: Re: Merry $*!@ing Christmas!
Date: Wed, 8 Dec 1993 03:37:32 GMT
Reply-To: benjamin-goldsteen@uokhsc.edu

hasty@netcom.com (Amancio Hasty Jr) writes:

>>|> own thing then start your own company with your own capital. In the meantime
>>|> you work for us. 
>>|> 
>>
>>Wrongo!  Nice try though.  I'm funded by research grants from commercial
>>companies not your tax money.  So no, I don't work for you.

>I am curious is your Motif license also funded by grants from
>commercial companies or are you using the University's tax payer
>paid Motif license?

     My bet the Motif license they are using is the UofI site license
(which I think is a source license...aren't I jealous!).  The building
they use was probably funded with at least some state/federal money
(even if they work in Beckman, I wouldn't believe no federal money was
used).  The Internet connection probably involves tax payer money.  So
what?  I would really like to fly an F16, but the military says I
can't.  NASA produces software, but you have to buy it from COSMIC. 
Wait, I paid for them to develope something and now I have to pay to
get it?  Yup!

     This whole thing is pretty darn silly.  The only thing that upsets
me is that Mr. Bina spent 10 hours converting XMosaic to Athena; a
total sinkhole in terms of time spent.  Somehow, I feel it is my fault
for not defending the choice of Motif (in terms of posts, it looked
like everybody was against NCSA).  I showed somebody Mosaic once and he
was impressed; it is once of the most beautiful looking X applications
we had seen.

     I, therefore, put up $50 (sorry, I am a student now -- I don't have
much) in the hopes of getting matching funds.  By my calculations, Mr.
Bina's work on the conversion is worth about a $1000.  Any takers?

P.S.Is the Mosaic team going to "support" this Athena version?  Costs
could get out of control...
P.S.S.Sorry about any typos; testing out a new terminal server; appears
to be dropping characters...
-- 
Benjamin Z. Goldsteen

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

From: tytso@athena.mit.edu (Theodore Ts'o)
Subject: (none)
Date: 8 Dec 1993 00:05:35 -0500
Reply-To: tytso@athena.mit.edu (Theodore Ts'o)

It's been a while since I've released any patches to the serial driver;
well, I've finally taken the time to package them all out and release a
new patch set.  I'd appreciate it if people could try them out.  They
can be found on tsx-11.mit.edu, in /pub/linux/ALPHA/serial.  

(Yes, I've checked the permissions on the files before sending out this
announcement.  :-)

                                                - Ted

Here's the README file for this set of patches:
===============================================

This is a set of patches to the serial driver in 0.99pl14.  It
contains the following fixes /  improvements:

        * Set FIFO trigger level to 1 if baud rate is under 2400.
                This will improve response time on 16550A's when they
                are operating at slow speeds; specifically when you're 
                using them with a mouse. 

        * Overrun conditions are now passed up to tty_io.c, and an
                appropriate warning message is printed.  (Previously,
                tty_io.c would handle it as a randomly as either a
                frame error, a parity error, or a break.)

        * Use a separate flag for hardware flow control, so that
                the right thing happens when both XON/XOFF and RTS/CTS
                flow control are enabled.

        * Allow the callout devices to be a controlling tty.

        * Hangup handling fixed; some race conditions removed.  The
                device specific close now happens at the instant when
                the hangup is signaled, instead of waiting until the
                the process closed the file descriptor.  This should
                fix the problems with background processes that still
                have the tty opened when the modem hangs up.  Linux
                will now deal correctly with this case.

Please try these changes out, especially if you support dialup via
modems.  There are a bunch of tricky race conditions with the hangup
code, which I *think* that I've resolved.  I want to make sure that
they work, though, before submitting them to Linux for inclusion in
the mainline kernel.  Please try them out and let me know how they
work on your system.  

                                        Ted Ts'o
                                        tytso@mit.edu
                                        12/8/93

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

From: becker@super.org (Donald J. Becker)
Subject: Re: Lots more stats in /proc aka /proc/loadavg
Date: Tue, 7 Dec 1993 16:10:03 GMT

In article <1993Dec6.202232.6597@fix.kmk.rhein-main.de>,
Kai Kretschmann <kai@fix.kmk.rhein-main.de> wrote:
>I think it is usefull, better would be your idea with the seperate
>subdir /proc/stats. The namespace of /proc won't be blown up.
>
>Some more statistics will come for shure like average logged in
>user count, network packet flow, and so on. Just take the first
                ^^^^^^^^^^^^^^
>step and get it included in the next kernel.

Check out /proc/net/dev, which contains the netcard device-level statistics.


Actually, I'm very open to comments about the format of /proc/net/dev.
 I put /proc/net/dev in as a low-cost way to get hardware-related netstats,
but it's current implementation is lacking a few features.  Things I would like
to change are:

        Some of the error stats are summaries, with no way to get the
        broken-out values.

        Perhaps the stats should include the device type (e.g. "NE1000"),
        network type ("10Mbs Ethernet") and the open/closed status.

        Other useful stats might be total transferred octet count,
        data-bytes-only count, and time since the last network operation.

Things I don't want to change are:

        It should remain human-readable, so another special tool isn't
        required. (I especially don't like /proc/net/arp, which was clearly
        designed by someone that prefers ioctl()s.)

        Device name should be easily accessible.  I use the the following
        in my /etc/rc.local:
                grep -s eth0 /proc/net/dev  &&  /etc/rc.net
        This avoids the error message if there isn't a netcard installed.
        On another machine that sometimes has an AT-Lan-Tec pocket adaptor
        installed (BTW, my driver for this $130-$150 BNC+RJ45 adaptor will
        be released RSN) I use something like the following:
                grep -s eth0 /proc/net/dev  &&  NETDEV=eth0
                grep -s atp0 /proc/net/dev  &&  NETDEV=atp0
                ...
                ifconfig $NETDEV ...

        It would be nice to keep it to 80 columns (that's the reason I don't
        currently print all of the stats.)

-- 

Donald Becker                                          becker@super.org
IDA Supercomputing Research Center
17100 Science Drive, Bowie MD 20715                        301-805-7482

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

From: mark@datasoft.com (Mark Buckaway)
Subject: Re: Always -- IN-2000
Date: Tue, 7 Dec 1993 18:47:20 GMT

T. Dixon Hutchinson (hutch@teal.csn.org) wrote:
: I'm just starting this process...

: I have all of my drives hanging of an IN-2000 scsi controller.
: I have made all of the disketts for the SLS release, but I don't
: know how to get an IN-2000 literate kernel onto the A1 diskette.

: Please help this poor lost sole.

Take it from me (at least don't get offended) an owner of a Always IN2000
and a Adaptec 1542b. Get another SCSI card. Seriously. IF you get the
IN2000 driver going (I never could), there are numerous problems reports
with it.

Note, the problem is not with the Always card, it's the broken driver in
Linux.

Mark
--
==============================================================================
Mark Buckaway           | root@datasoft.com         |  DataSoft Communications
DataSoft Communications | uunorth!datasoft!root     |  62 Rock Fernway
System Administrator    | Voice: +1 416 756 4497    |  Willowdale, ON M2J 4N5
==============================================================================
    "UNIX and OS/2 are operating systems. Windows is a shell, and
                         DOS is an boot virus"
==============================================================================

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

From: mirons@icarus.ci.net (Michael A. Irons)
Subject: How about C.O.L.kernel ?
Date: Tue, 7 Dec 1993 00:57:19 GMT


        My understanding of C.O.L.Development was that is was for
kernel related work.  As this newsgroup seems to be drifting from that
towards applications how about a sibling newsgroup with a more 'in
your face' obvious name like C.O.L.kernel?


-- 

                                Mike Irons

                        mirons@Icarus.CI.NET

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

From: wpp@lise.physik.tu-berlin.de (Kai Petzke)
Subject: Re: How about C.O.L.kernel ?
Date: 8 Dec 1993 10:18:17 GMT

In <CHn3zJ.3GL@icarus.ci.net> mirons@icarus.ci.net (Michael A. Irons) writes:


>       My understanding of C.O.L.Development was that is was for
>kernel related work.  As this newsgroup seems to be drifting from that
>towards applications how about a sibling newsgroup with a more 'in
>your face' obvious name like C.O.L.kernel?


Good idea!

Someone should start an RFD on the new group soon!

--
Kai
wpp@marie.physik.tu-berlin.de
Advertisement by Microsoft in a well-known German magazine:
        If you don't like our programmes, then make your own ones.
However, they expect you to use Microsoft products for this -:)

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

Crossposted-To: comp.infosystems.www,comp.windows.x,comp.windows.x.i386unix,comp.windows.x.motif,gnu.misc.discuss,comp.sources.d
From: de5@de5.ctd.ornl.gov (Dave Sill)
Subject: Re: Merry $*!@ing Christmas!
Date: Wed, 8 Dec 1993 13:34:10 GMT

In article <CHnw0r.Fus@ucc.su.OZ.AU>, dawes@physics.su.OZ.AU (David Dawes) writes:
>In article <1993Dec6.153330.29733@ornl.gov> de5@de5.ctd.ornl.gov (Dave Sill) writes:
>>
>>Hogwash.  The producers are free to ignore their customers desires at their
>>own peril, but users are free to ask for or expect whatever they want.  (If
>>they don't get what they expect, they're also free to stop using the
>>product, find one that does meet their needs, hack the code, etc.)  It's
>>called "feedback", and most implementors appreciate it even if they 
>>can't make everyone happy.
>
>Freeware developers appreciate constructive feedback, but they don't
>appreciate demands from those who choose to use their software.  From
>my experience, expectations of users are too often presented to the
>developers in the form of demands, and often rudely.

Two wrongs don't make a right.  Sure, some people have been rude to the
Mosaic developers--I don't condone that at all.  However, the developers
responded in kind--even to those who were nothing but polite to them.

>I think your use of "customer" to describe the users of free software
>is inappropriate.

Whether they pay (directly) or not, or don't even pay indirectly through
their taxes, people that use Mosaic are still NCSA's customers, and they
deserve to be treated with respect--provided they are respectful
themselves.

>What exactly is the "peril" producers face?  That fewer
>people use their software?

That's certainly one peril.  Another is a backlash resulting in reduced
funding by corporate donors and government agencies.

>So what?

I'd be willing to bet that NCSA is *quite* concerned about the popularity
of their products.  If they aren't popular, they won't attract attention,
which will jeopardize future funding.

>Are you speaking as a user of free software or a developer?

Both.

-- 
Dave Sill (de5@ornl.gov)             Computers should work the way beginners
Martin Marietta Energy Systems       expect them to, and one day they will.
Workstation Support                                            -- Ted Nelson
URL http://gatekeeper.dec.com/archive/pub/DEC/DECinfo/html/dsill.html

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

From: hsieber@Mathematik.TU-Chemnitz.DE (Holm Sieber)
Crossposted-To: comp.os.linux.admin,comp.os.linux.misc,comp.os.linux.help
Subject: Re: EXT2 filesystem unrecoverable error - HELP!!!!!
Date: Wed, 8 Dec 1993 14:15:39 GMT

In article <ecarpCHK3EA.Do8@netcom.com> ecarp@netcom.com (Ed Carp) writes:

>Then it STOPS!  It won't continue, it won't fix the filesystem.  When the
>system comes up, and it hits the directory where the problem is, the
>damned kernel PANICS!

>kernel panic: EXT2-fs panic: ext2_read_inode: unable to read i-node block

I have the same problem. The reason is a damaged harddisc. But I don't
know how I can repair the problem.
       //    Holm Sieber         Tel.: +49(0)371/411784        \\
       ||    Gieszerstr. 5             (FRG/Chemnitz/..)       ||
       ||    09130 Chemnitz                                    || 
       \\    e-mail: holm.sieber@mathematik.tu-chemnitz.de     //

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

From: passat@cs.tu-berlin.de (Ulf Karbstein)
Subject: [Q] ISODE port for Linux ??
Date: 8 Dec 1993 15:20:21 GMT

Hello !

I am looking for an ISODE port for linux. Is there any version available
via FTP or someone working on it ?!?!?!?

Greetings, Ulf

-- 
PS:Dieser Text enthaelt keine das Netz schaedigenden Rueckstaende; die Bits
   sind vollstaendig elektronisch abbaubar. Er traegt daher den gruenen Strich.

   the GREEN LINE    ==============

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

From: neal@ctd.comsat.com (Neal Becker)
Subject: Re: Appletalk and LINUX Update
Date: 08 Dec 1993 15:37:55 GMT

We use uar on hpux with cap with fine results.  uar may be easier to
port.  I believe it needs access to raw sockets to work.

Here is the readme for UAR:

                                 UAR
                        UNIX AppleTalk Router

                     The University of Melbourne
                          djh@munnari.OZ.AU
                            October, 1992
                            version 1.0b8


NB: This is BETA code. It has been tested under SunOS 4.1.2, AIX 3.1.5
    IRIX 4.0.4 and HP-UX 8.07 only. Most testing involved routing between
    Phase 2 networks. Mixed Phase interfaces and Phase 1 only testing has
    been limited. Use with care. The packet filter code is based on the 
    ARNS package which supports the hosts listed below. Please do not
    redistribute this code until it has been officially released ...


UAR is a 'UNIX AppleTalk Router' supporting Phase 1 or Phase 2 AppleTalk
routing between multiple ethernet interfaces on a UNIX workstation.  UAR
also supports CAP services which are attached to one of the interfaces
and AppleTalk packet tunneling over an IP internet.

The UAR router currently supports Phase 1 and Phase 2 EtherTalk networks
connected to SUN, DEC ULTRIX, SGI IRIX, Sony NEWS 4.2, HP-UX 8.07 and IBM
RS6000 AIX workstations, and Phase 1 only on Sony NEWS pre-4.2 workstations.

The router is configured from a configuration file, if no file is provided
the router will try and glean the necessary information from other routers
on the network (ie: it acts as a non-seed router). The default file name is
/usr/local/lib/cap/uar.conf, for example:

# configuration file for UAR
#
interface le0
        node 244
        network 83.5
        networklo 83.5
        networkhi 83.6
        zone unimelb-CompSci
        zonelist unimelb-CompSci
        zonelist unimelb-Maths
        zonelist unimelb-Stats
        phase 2
        cap on
interface le1
        network 83.4
        zone "unimelb-CompSci"
        zonelist "unimelb-CompSci"
        phase 2
        cap off

There should be a "zonelist" entry for each of the zones on a Phase 2
network (including the default zone).  NB: this is not the same as the
full zone list for the entire network. The "zone" entry should contain
one of these listed zone names and is used as the default zone name for
UAR services such as CAP. If UAR is running on a routerless network then
the "zone" entry also becomes the default zone name for the cable.

The configuration information in uar.conf *must* be identical to that
configured into other routers for the local network.

Please report any problems via email to djh@munnari.OZ.AU


UAR INSTALLATION:

To install the UAR router, 'make uar' and install it in
/usr/local/cap.  The arguments to uar are as follows:

        -d number       bit positions in <number> print debugging messages
                        for various subsystems.  For details see uar.h.

        -l logfile      use "logfile" for logging output /* NYI */.

        -1 | -2         use either Phase 1 (default) or Phase 2.

        -c              don't use DDP checksums (default is on).

        -C              attach CAP to the first interface. Normally CAP
                        services are specified in the configuration file.
                        Use -C only for non-seed routers (no config file).

        <interface list>

EG:
        uar -d 3967 -l logfile -2 ie0 ie1


CAP

To run UAR with CAP, Configure CAP for use with UAB. On hosts where UAB
support in Configure is not available or CAP is to be used with Phase 2,
compile CAP for IPTalk (select the default answers for all Configure
questions) and edit the m4.features file to include the lines

# + PHASE2 add support for AppleTalk Phase 2
define(`specialcflags',concat(specialcflags,` -DPHASE2'))
#

Run 'gen.makes' to create the makefiles. cd to cap60/lib/cap and edit
the file 'makefile'. Change the line

LAPOBJ=abkip.o abddp.o abnbp.o atalkdbm.o

to

LAPOBJ=abmkip.o abddp.o abnbp.o atalkdbm.o
         ^

Then compile CAP as usual, 'make programs'  etc ...
Don't run 'uab', run 'uar' then 'atis', 'atlook', 'getzones'  etc ...


TUNNELING

UAR supports a simple method for tunneling AppleTalk packets over an IP
internet.  That is, EtherTalk networks separated by IP-only routers may
be joined seamlessly by running UAR on UNIX hosts connected to each net.

The only restriction is that network numbers at each location must be
unique across the extended EtherTalk network. ie: network number remapping
is not supported.

IP tunneling is specified in each uar.conf file as follows

interface tnnl
        node 253
        network 83.3
        zone unimelb-CompSci
        peer "253 @ 128.250.97.86"
        peer "252 @ 128.250.73.40"
        phase 2

Each UAR host participating in IP tunneling must contain identical
network, zone, peer and phase entries and each must have unique node
numbers.  The peer entries are used to map node numbers to IP addresses.
On multiple interface UNIX hosts, a peer entry for the local node number
(253 in the above example) is used to specify which interface is to be used.

For security, UAR hosts are not permitted to participate in tunneling unless
their IP addresses are listed as peers in each uar.conf file.


STATISTICS

UAR can be requested to dump the AARP, RTMP, ZONE and statistics tables
by sending signals to the UAR process.

The signals are

        HUP  dump the RTMP table to /usr/tmp/uar.rtmp
        QUIT dump the AARP table to /usr/tmp/uar.aarp
        USR1 dump the ZONE table to /usr/tmp/uar.zone
        USR2 dump the stats table to /usr/tmp/uar.stats


NOTES

UAR runs as a UNIX user process, it's not designed for continuous,
high-throughput situations.  For this purpose you should consider
an external dedicated hardware gateway, for example, a Cisco.

Some hosts only support a "small" number of multicast addresses per
ethernet interface. Under SunOS, this is 64, under AIX (maybe) 10.
If your local Phase 2 network range contains more than this number of
different zones, you will have problems using UAR.  Note that this is
not the total number of zones on your network but the number of zones
defined for the single network number range on the attached cable.  At
present the zone multicast limits are not known for other machines that
UAR runs on.  Although Apple supports up to 255 zone names per network,
you should perhaps consider that your network is not well partitioned if
you are using anywhere near this number ...


NOTICE

Copyright (c) 1993, The University of Melbourne.
All Rights Reserved. Permission to publicly redistribute this
package (other than as a component of CAP) or to use any part
of this software for any purpose, other than that intended by
the original distribution, *must* be obtained in writing from
the copyright owner.

This software is supplied "as is" without express or implied
warranty.  May not be resold.


========================
Notes for IBM AIX users:

You MUST manually set the define "AIX" in Makefile.

If you use the interfaces named in 'netstat -i' you may receive the message
"en1: A file or directory in the path name does not exist."  Instead, use
the network interfaces named "ent0" or "ent1" etc.

==========================
Notes for Sony NEWS users:

You may know that recently Sony Corp. released NEWS-OS4.2 which supports 
AppleTalk protocols. ARNS runs on both NEWS-OS4.2 and pre-4.2 NEWS-OS that
supports /dev/ether0 device. This note describes how to use UAR on SONY NEWS.

1. /dev/ether0

Please make sure that /dev/ether0 exists. If not, you can make this device
file by issuing the following command in the directory /dev:

        % sh MAKEDEV ether0

2. Restrictions on pre-4.2 users

You cannot use phase 2 option (-2 option) on this OS version. This is because
multicast address cannot be set on /dev/ether0. The only way to use phase 2
is to upgrade to NEWS-OS4.2.

3. Restrictions on 4.2 users

You cannot use UAR and AppleTalk facility provided by OS4.2 at the same time
if you want UAR to speak phase 2. Don't run UAR in phase 2 mode if atconfig
is up. To make sure that atconfig is down, issue the following command:

        % atconfig en0 down

On the other hand, you can run UAR in phase 1 mode (use -1 option) and 
AppleTalk facility without any problem.

4. General Restrictions

Because UAR uses /dev/ether0 and the device file cannot be opened at the same
time by more than one processes, you can run only one of the following programs
on the same host:

        UAR
        ARNS
        uab for CAP6.0
        rarpd (on pre-4.2 NEWS-OS)

Rarpd on NEWS OS4.2 now uses BPF and can run simultaneously with processes
using /dev/ether0.

5. Compiling on pre-4.2

Just invoke make command.

6. Compiling on OS4.2

You must put -Dsony_phaseII option on compiling uar.c and pf.c files. Edit
Makefile for these lines. Then make it.

7. Usage

On pre-4.2 OS, the following is the typical way to invoke ARNS:

        % uar -l logfile -1 /dev/ether0

The -1 and -i options with /dev/ether0 argument are necessary.

On OS4.2, use the following instead:

        % uar -l logfile -2 -i /dev/ether0

Then uar will speak phase 2. 

Yasuaki Honda
Sony Computer Science Laboratory Inc. 
honda@csl.sony.co.jp

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


** 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
******************************
