Subject: Linux-Misc Digest #423
From: Digestifier <Linux-Misc-Request@senator-bedfellow.MIT.EDU>
To: Linux-Misc@senator-bedfellow.MIT.EDU
Reply-To: Linux-Misc@senator-bedfellow.MIT.EDU
Date:     Mon, 13 Dec 93 16:13:24 EST

Linux-Misc Digest #423, Volume #1                Mon, 13 Dec 93 16:13:24 EST

Contents:
  Linux locking under X (John Aughey)
  Austin Code Works: Linux CD-ROM (Andrew Krieg: Man or Android?)
  Re: OGI Speech Tools for Linux (Klaus Weidner)
  Re: Linux rsh BIG RAGGEDY HOLE ( Alfred Longyear)
  Re: Elm 2.4.23 and ISO-8859-1 charset on Linux (Arjan de Vet)
  Re: LGX List of Problems #5 (Grant Edwards)
  Using TERM w/TCP-IP link (Larry Arbanas)
  Re: Elm 2.4.23 and ISO-8859-1 charset on Linux (Arjan de Vet)
  Re: Linux Consortium (Theodore Ts'o)
  adjtime and ntpd.... (Bjorn Vidar Remme)
  Re: Linux Consortium (Phil Hughes LJ Editor)
  Linux {Oranization|Consortium|Foundation|?} (Phil Hughes LJ Editor)
  X-Windows (ADAM HUSIK)
  Re: Linux rsh big raggedy hole (Jan Willem Hubbers)

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

Crossposted-To: comp.os.linux
From: augheyjh@mentor.cc.purdue.edu (John Aughey)
Subject: Linux locking under X
Date: Mon, 13 Dec 1993 17:05:38 GMT

I just installed Linux and I'm having trouble keeping it sane under X.
It starts X OK, but after a while, it will just lock.  The first time, it
was sitting in xlock, and the second time I was loading emacs.  It was
completely locked up!  Even Num lock, caps lock and scroll lock wouldn't
work.

I'm running a 486/33DX with 8 megs of real and 12 megs of swap.

Any suggestions?


-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
/*  John H. Aughey                             | 3.14159265358979323846264338
**  jaughey@sonata.cc.purdue.edu               |   32795028841971693993751058
**  Note: finger account for PGP public key    |   20974944592307816406286208

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

From: ultron@csd4.csd.uwm.edu (Andrew Krieg: Man or Android?)
Subject: Austin Code Works: Linux CD-ROM
Date: 13 Dec 1993 17:19:16 GMT

I am ready to put Linux on my system.  Is a CD-ROM with the Linux sources a
good route to go for an initial installation?  The Austin Code Works is
offering a CD-ROM w/ Linux (version 0.99.10), 386BSD (version 0.1) and NetBSD
sources all on 1 disc.  I realize there are new patches that come out
regularly.  Can they be easily installed onto an existing version, or will I
just end up having to download a whole new version again?  I'm wondering if
$40 is worth the aggregation of downloading and installing 30+ floppies
(assuming I can upgrade easily).
-- 
=     Andrew Krieg      |                                               =
= krieg@ct.med.ge.com   |  Treguna   Mekoides   Trecorum   Satis   Dee  =
=          or           |                                               =
= krieg@point.cs.uwm.edu|                                   - Astoroth  =

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

From: klaus@snarc.greenie.muc.de (Klaus Weidner)
Subject: Re: OGI Speech Tools for Linux
Date: Mon, 13 Dec 1993 12:49:14 GMT

jedubins@unix.amherst.edu (Just a fellow traveller...) writes:
>A note of warning: there still are a few bugs in the package, but they
>are mostly in the X-graphical tool Lyre(and auto_lyre which is a scripted
>version of Lyre).  You'll often get a floating point exception error
>when trying to use Lyre.  I'm not sure what's going on here either.  If
>you would like to help email Tilo at tilo@cs.TU-Berlin.DE.  I'm sure he'll be
>able to send you his diffs or a list of his patches.  The general *nix source
>release is on speech.cse.ogi.edu in /pub/tools/ogitools.v1.0.tar.Z

I think I found the `floating exception' bug - just add the statement
`#include <stdlib.h>' at the top of the files `bin/lyre/main.c' and
`bin/auto_lyre/main.c'. It now displays a beautiful two-dimensional
spectrum for the data files.

It seems that the compiler expected `atof()' to return an int, took
4 bytes from the stack and converted that to a float, instead of
taking the top value out of the co-processor stack. 

(P.S.: somebody ought to explain the difference between a NULL pointer
and ASCII NUL to the program author... that is the cause for most of the
integer <-> pointer warnings produced.)

It looks like a great package, now I'll start experimenting...

ciao, Klaus
-- 
\ klaus@snarc.greenie.muc.de--kweidner@physik.tu-muenchen.de--2:246/55.4
\ .signature error -- quote dumped

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

From: longyear@citrus.SAC.CA.US ( Alfred Longyear )
Subject: Re: Linux rsh BIG RAGGEDY HOLE
Date: 3 Dec 93 03:27:58 GMT

hobbit@babyoil.ftp.com (*Hobbit*) writes:

>If this has been discussed already, ignore this msg.

It has to most degree.

>The Linux boxes I've examined so far come with

>       root::0:0:random-string:/:/bin/sh

>in /etc/passwd.  This is fine since most incoming servers use the shadow
>mechanism, but rshd does NOT check /etc/shadow.  It just lets you in.

>At the very least, put a * in all the passwd-fields in your passwd file...
This won't work unless you do have shadow passwords.

>This is being CCed to CERT and CIAC.  Let's see what they do with it.
I doubt that they will do anything. It is not their function to comment on
people's stupidity.

(Remove the cross posting becase it doesn't belong anywhere else but in
the linux.admin group.)

It is valid to ship the system with the root password being empty. The error
is what most people do after that -- they simply leave it that way. It comes
from running MSDOS. "That system has no security, so why should this one ???"

Secondly, you should have a version of rshd which uses shadow passwords. All
of the programs which expect to do logon validation should use shadow
passwords. This is being made a standard function of the next version of the
shared library (to include the shadow suite and use it where appropriate.)

However, the fact that the root user has no password is not a fault of the
operating system nor the packager. It is a fault of the end-user who does
the installation. If people care about root access to their system then they
should set the password for root. That is simply part of the installation
of the system. If they don't then they get what they want -- no security for
the root account.
-- 
longyear@citrus.sac.ca.us       csusac.ecs.csus.edu!citrus!longyear

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

From: devet@adv.win.tue.nl (Arjan de Vet)
Subject: Re: Elm 2.4.23 and ISO-8859-1 charset on Linux
Date: 9 Dec 1993 00:31:05 +0100

In article <2e36ot$1tm@adv.win.tue.nl>, I <devet@adv.win.tue.nl> wrote:

>Did anybody succeed compiling Elm 2.4.23 in such a way that ISO-8859-1
>characters are properly displayed using the builtin+ viewer? less works OK
>as an ISO-8859-1 viewer, so the problem must be in the builtin+ viewer.
>
>I found out that isprint() and isascii() return 0 (false) for characters
>above 127 on Linux. Because of this Elm replaces characters above 127 with
>a `?'.  But even with removing this isprint() and isascii() check Elm still
>doesn't display ISO-8859-1 characters correctly: an e with a ' accent is
>displayed as a white solid box :-(.

I got some reactions by mail (thanks!) and the (temporary) solution was
adding a

        setlocale (LC_CTYPE, "ISO-8859-1")

at the beginning of main(). The call

        setlocale (LC_ALL, "")

does not seem to work on Linux yet. It should take the value of the
environment variable LANG to set all LC_* variables (according to the
Linux manual pages).

Arjan

--
Arjan de Vet                             <Arjan.de.Vet@adv.win.tue.nl> (home)
Eindhoven University of Technology, the Netherlands <devet@win.tue.nl> (work)

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

From: grante@hydro.rosemount.com (Grant Edwards)
Subject: Re: LGX List of Problems #5
Date: Mon, 13 Dec 1993 17:17:47 GMT

Adam J. Richter (adam@adam.yggdrasil.com) wrote:
: In article <1993Dec11.000838.7153@rosevax.rosemount.com>,
: Grant Edwards <grante@hydro.rosemount.com> wrote:
: >Adam J. Richter (adam@adam.yggdrasil.com) wrote:
: >
: >: > The LGX linux header files are different (and incompatible) with the
: >: > pl13 kernel sources on sunsite.unc.edu.
: >: [TOO VAGUE TO VERIFY.]
: >
: >If you like, I can send you a list of header files that are not the
: >same as the those in the pl13 sunsite kernel sources.  I don't know
: >that it matters -- the LGX headers are compatible with the LGX
: >sources.  I assume there are differences because LGX contains an older
: >version of pl13.

:       What are you talking about?  There is no such thing as "an
: older version of pl13."

Oh.  I thought that the letters after the number indicated successive
revisions of a patch level, e.g. pl13r, pl13q, and so on.  Perhaps
not.

: If you are the author of this bug report, what do you mean by
: "incompatible?"

I mean that the sunsite pl13 sources will not compile if you use the
header files on the LGX CD (and, I suppose, vice versa: the LGX kernel
souces won't compile with the pl13 header files from sunsite).

: The kernel distributed with LGX has a number of additional device
: drivers, so there will be additions, of course.

That's a different issue: those additional device drivers caused
problems when you configured them out.  They caused errors when the
kernel was linked -- the initialization of the device driver table in
the kernel was still happening, even though the routines to which the
table entries pointed weren't compiled in.

: The only other improvement made in LGX that comes to mind is the
: elimination of the i_dev field from the inode data structure, which
: is designed to make it easier to implement certain types of
: filesystems.  This necessitated a some minor changes to some file
: cacheing routines, but that's about it.

** change (whether and improvement or not) == incomptable **

The change you made in the inode structure means that some of the
"vanilla" pl13 source files won't compile.  I _think_ there was one
other incompatibility that I ran into, but maybe the inode structure
was the only one.

It wouldn't have mattered, but the proc filesystem files didn't get
installed.  By the time I figured out what was missing, I didn't have
a CD-Drive, so I got the proc filesystem code from the pl13 sources on
sunsite.  I copied the proc stuff into the LGX source stree and tried
to compile.  No Dice.

I grabbed the entire kerenel source tree from sunsite and tried to
compile.  Still no go.  This time it was becuase /usr/include/linux
was not a symbolic link to /usr/src/linux/include, so I was still
using the LGX headers rather than the ones that came with the kernel
sources from sunsite.

After I fixed the symbolic links so that I was using the header files
that came with the sources, things worked fine.

In your next release, I suggest that you document the fact the your
"pl13" kernel isn't the same as the "pl13" kernel that you get from
ftp sites.

--
Grant Edwards                                 |Yow!  I'm dressing up in an
Rosemount Inc.                                |ill-fitting IVY-LEAGUE
                                              |SUIT!!  Too late...
grante@rosemount.com                          |

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

From: arbanas@cpchq.cpc.gmeds.com (Larry Arbanas)
Subject: Using TERM w/TCP-IP link
Date: 13 Dec 1993 17:42:58 GMT

Is it possible to run Term over an existing ethernet based TCP/IP
link between 2 computers, instead of over a serial/modem link?

This would have application with machines that are blocked from direct 
Internet access but can telnet to an Interent connected machine.

Thanks

--
R. Larry Arbanas           | voice:          +1 (313) 575-4423           
                           | email:          arbanas@hqs.mid.gmeds.com
                           | email-personal: arbanas@bovaris.northville.mi.us


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

From: devet@adv.win.tue.nl (Arjan de Vet)
Subject: Re: Elm 2.4.23 and ISO-8859-1 charset on Linux
Date: 12 Dec 1993 20:27:54 +0100

In article <2e828u$40@renux.frmug.fr.net>,
Rene COUGNENC <rene@renux.frmug.fr.net> wrote:

>> Did anybody succeed compiling Elm 2.4.23 in such a way that ISO-8859-1
>> characters are properly displayed using the builtin+ viewer? less works OK
>> as an ISO-8859-1 viewer, so the problem must be in the builtin+ viewer.
>
>This is a common problem whith all versions of elm :-((
>Here is the patch, it is not against pl23, but should work whith some offset:

The real problem is the uncomplete setlocale implementation of the Linux
C-library. But Mitchum DSouza is working on it.

Arjan

--
Arjan de Vet                             <Arjan.de.Vet@adv.win.tue.nl> (home)
Eindhoven University of Technology, the Netherlands <devet@win.tue.nl> (work)

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

From: tytso@athena.mit.edu (Theodore Ts'o)
Subject: Re: Linux Consortium
Date: 13 Dec 1993 15:12:30 -0500
Reply-To: tytso@athena.mit.edu (Theodore Ts'o)

   From: markline@henson.cc.wwu.edu (Mark Line)
   Date: 11 Dec 93 07:28:34 GMT

   To the developers: *Talk* to us. Or put up with the consequences. And
   read Frankenstein over the holidays.

OK, I also think the term "Linux Consortium" is at the very minimum, in
bad taste.  It presumes to try to speak for "the Linux community"
(whatever that means) when it obviously doesn't.  It falls in the same
category as past attempts to try to establish the "Linux Filsystem"
(xiafs, which is used by very few users), and the "Linux Standard
Distribution".

Call it something else, like the Linux Distribution Review Group; that
is much less offensive than the "Linux Consortium". 

                                                        - Ted

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

From: bjornv@brosme.dhmolde.no (Bjorn Vidar Remme)
Subject: adjtime and ntpd....
Date: 13 Dec 1993 18:11:02 GMT
Reply-To: bjornv@brosme.dhmolde.no (Bjorn Vidar Remme)


Hi....

I tried to compile some NTP-stuff (xntp3.3b.tar.Z) with no sucess.....
It fails when trying to compile adjtime (loads of warnings).

Any suggestions how to solve this problem ...(I dont even no what the problem
                                               REALLY is ....)

                                 


                                   Bjorn Vidar Remme (bjornv@brosme.dhmolde.no)

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

From: phil@fylz.com (Phil Hughes LJ Editor)
Subject: Re: Linux Consortium
Date: Sun, 12 Dec 1993 19:27:27 GMT

Lars Wirzenius (wirzeniu@klaava.Helsinki.FI) wrote:
: markline@henson.cc.wwu.edu (Mark Line) writes:
: > Is LJ the one that's been promised off and on sporadically to appear
: > some day, or is this something I've missed?

: Yup.  I don't know what their current status is.

You don't?  Aren't you the guy who approved the posting a few
weeks ago to comp.os.linux.announce?  The short version is we are
real, we are collecting articles, press releases, advertisers, subscribers,
...  The first issue will be out in February, 1994.
For more info, send e-mail to mail-server@fylz.com and put
    send lj/info.sheet
in the body of the message.

There are other documents available, the line
    dir lj
will get you a table of contents.

: > Why are
: > 
: > (a) Magnus publishes a review in UnixWorld or some such; and
: > (b) a group of LC collaborators publish a rivew in UnixWorld or some
: >     such
: > 
: > so different that people are prepared to support the one but not the
: > other?

: You completely miss the point.  The problem is not publishing reviews.
: The problem is giving pompous, official-looking "Stamps of Approval".
: They may not be meant to be pompous and official-looking, but that's
: what they are.  Forget the approval business and the whole thing starts
: sounding reasonable.

: There's no problem in having a group of people do reviews together.  In
: fact, it's a good idea --- you get much more experience into the review.
: The latest name suggestion I've seen, "Linux Review Group", still sounds
: too official to me; I'd prefer if the word Linux didn't apprear in the name
: at all.  In fact, I don't think a name is needed in the first place,
: because the group would probably work better as just a group and not an
: organization.

Linux development has been done in a very open manner.  Probably more
so than anything else of this size.  Unless the "stamp of approval"
has some sort of very open approval process I agree with Lars, it
looks like "big government".  Linux doesn't need this but it does
need two things:
   - some sort of coordination of development (which pretty much
     exists) so people know what other people are doing and have
     some standards to work by
   - something that conveys an "external image" of Linux

I think the second problem is the one that people want to address.
From talking to non-Linux people about Linux the most common concern
I hear (from Unix professionals, DOS users and vendors who might
consider porting their products to Linux) is that Linux is unreliable.
This is because of the very open development.  SCO won't publish
a list of their bugs whereas the bugs in Linux are discussed in
a public forum.

For development, this public forum is very good.  Bugs get found and
fixed very fast without a huge paid testing staff.  But it doesn't
do much for the image of Linux in the "commercial world".
This is one of the main reasons that we have started Linux Journal.
I personally have been working with Linux for about a year.  I believe
in it.  I use it.  The mail server I described above runs on it.
But I see the need to project an image that shows how real
it is if we are to get vendors (for example, intelligent comm board
vendors) to jump on the bandwagon.

This brings up the question of whether or not we want Linux to become
a "commercial product".  I think the problem is in definition.  If
we want Linux to prosper it will need to be somewhat popular.  And
those of us in the Linux community cap benefit.  For example, we
get:
  - an operating system we want with source code
  - support of peripheral equipment from vendors
  - a platform to use to build end-user products (a job)

End users (commercial enterprises who just want a solution) can
get a more cost-effective solution to their problem with available
local support.

If we agree that it is ok to be "commercial" then the next step is
how to compete in the OS market.  Clearly, Linux isn't going to do
that with full-pages ads in magazines.  In fact, even getting press
releases published on something inexpensive or free in magazines that
are supported by ads from Microsoft, SCO, et al, is very difficult.

Thus, the next reason for Linux Journal.  If our focus in clearly
the Linux community we don't have this sort of conflict. The question
that remains to me is whether or not we need something in addition
to Linux Journal to convey the "external image".  I think that is
up to the rest of you to decide.  Here is what we intend to offer:
   - editorials (this discussion is a good example)
   - beginning tutorials
   - technical articles
   - columns on various subjects including
       - FSF
       - Debian
       - letters
       - new product announcements
       - product reviews
       - programming tips
       - questions and answers
   - interviews
   - Linux in the "real world"
   - advertisements

Finally, let me say that having a paper journal with real articles and
real ads helps the Linux community a whole lot.  It will be much easier
to convice your boss, your customer, your neighbor or a vendor that
they should do something with Linux if you can show them a magazine.
It's also easier to read than comp.os.linux.* when you are on the bus. :-)

I would like to encourage people to let us know what they think
of our ideas.  What we are doing is based on a survey I sent out
to comp.os.linux about 6 months ago.  I also encourage people to
join in and help.  If you have an idea for a article, have some
comments or want to join our product review team, let me know.
-- 
Phil Hughes, Editor, Linux Journal, P.O. Box 85867, Seattle, WA 98145-1867 USA
E-mail: phil@fylz.com   Phone: +1 206 524 8338 FAX: +1 206 526 0803

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

From: phil@fylz.com (Phil Hughes LJ Editor)
Subject: Linux {Oranization|Consortium|Foundation|?}
Date: Sun, 12 Dec 1993 19:41:00 GMT

I was going to write an editorial for Linux Journal on the current
discussion for a Consortium/Foundation/Organization/... but I
realized that maybe some people would like to write it for me. :-)
Makes my job easier and maybe we can get more opinions in instead
of my interpretation of a lot of opinions.

In any case, there will be something in the first issue of LJ about
this.  If you want to submit an opinion piece, go for it.  I would
like to have that input by January 3. I will attempt to include
as much input as I can--either within an article I write or as
separate pieces--depending on what I get for input.
-- 
Phil Hughes, Editor, Linux Journal, P.O. Box 85867, Seattle, WA 98145-1867 USA
E-mail: phil@fylz.com   Phone: +1 206 524 8338 FAX: +1 206 526 0803

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

From: adam.husik@njcc.wisdom.bubble.org (ADAM HUSIK)
Subject: X-Windows
Date: 9 Dec 93 19:45:59 GMT


    Guess What?! I finally got X-windows to run in mono mode with my
laptop, and a mouse. Thanks a lot for the great replys I got on how to
partition my HD using mke2fs. My only problem now is that I get the
basic Xmono mode running, but then only have a empty screen with no
menu's, or ways to run a program. Where do I go from here to run the
sample programs that came with my SLS package?

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

From: jwhub@sci.kun.nl (Jan Willem Hubbers)
Subject: Re: Linux rsh big raggedy hole
Date: Mon, 13 Dec 1993 18:01:44 GMT

In <1993Dec8.184430.11035@cronkite.ocis.temple.edu> xeno@velcro.cis.temple.edu writes:


>from what i understand that is a genereic sls bug, i do not think it has anything to do
>with the mcc, however i could be wrong...
>just adding my own two cents !!! :)
Just check /etc/password
>christopher k neitzert                 xeno@velcro.cis.temple.edu
>-all standard disclaimers apply--
--
 Jan-Willem Hubbers         ( For HiFi [no, not out of a computer] )
 Internet: jwhub@cs.kun.nl  ( For /usr/local/gnu/bin/bash -login   )
 Linux   : jwhub@bsdnet     ( For bsdnetters/dial-in only :-)      )
 Fidonet : 2:284/122.2      ( For slow mail :-(                    )

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


** 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-Misc-Request@NEWS-DIGESTS.MIT.EDU

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

    Internet: Linux-Misc@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-Misc Digest
******************************
