Linux-Misc Digest #886, Volume #19 Sun, 18 Apr 99 13:13:17 EDT
Contents:
Re: FreeBSD vs. Linux ("Cameron Spitzer")
Re: Boot problem (kernel 2.2.6) (Killingtime "@drum14.freeserve.co.uk>)
Re: Kernel installation question (Bruce Daane)
Re: Linux command: bg and fg (brian moore)
problem with my soundblaster AWE64 ("Simon")
hal91 (ted)
Re: FreeBSD vs. Linux (Tom Christiansen)
ppp and SuSE 6.0 ("Mitchell Scott")
Re: This is not an oppinion - just an interesting fact. (Walter Strong)
missing file for gnome (Bob Tennent)
----------------------------------------------------------------------------
From: "Cameron Spitzer" <[EMAIL PROTECTED]>
Crossposted-To: comp.unix.bsd.freebsd.misc
Subject: Re: FreeBSD vs. Linux
Date: 18 Apr 1999 14:50:48 GMT
In article <[EMAIL PROTECTED]>,
Donn Miller <[EMAIL PROTECTED]> wrote:
>Tak Pui Lou wrote:
>> Dana Booth wrote:
>> > Try Windows NT, it's the best,
>> Did you post this to entertain other people?
>
>An even bigger joke is that NT is POSIX compliant.
Is FreeBSD POSIX compliant? Has anybody actually run the
certification suite on it?
Cameron
------------------------------
From: Killingtime <"matt<NO SPAM>"@drum14.freeserve.co.uk>
Subject: Re: Boot problem (kernel 2.2.6)
Date: Sun, 18 Apr 1999 12:08:29 +0100
Check the site where you got the kernel from. You may find that there are other
dependencies that need updating, such as gcc, and the bin utils.
mox wrote:
> I don't use lilo, my HD don't support it.. maybe because it is too big
> (20gig)
> I boot from a floppy made by (cat zImage > /dev/fd0)
> My old kernel version (2.0.36) works fine like that.. I think the problem is
> more with configuration
>
> Killingtime @drum14.freeserve.co.uk> <"matt<NO SPAM> wrote in message
> news:[EMAIL PROTECTED]...
> > If you recompiled the Kernel and then moved it to /vmlinuz did you type
> 'lilo'.
> > This may solve your problem. You will need to use your boot floppy to get
> > access to your partition.
> > mount root=/dev/hxx
> >
> > mox wrote:
> >
> > > Anyone have an idea why my system halt right after the boot disk says:
> "Ok,
> > > now booting the kernel..."
> > >
> > > I just installed the kernel v2.2.6..
> > > I don't know why.. many ppl said to me to remove some things into my
> config,
> > > I made what the asked me to remove, but I have no results... the problem
> > > can't be my system.. (p2 333 / 20gig / 96 ram)
> > > I think the problem is my kernel configuration, cause the v2.0.36 boot
> fine.
> >
------------------------------
From: Bruce Daane <[EMAIL PROTECTED]>
Crossposted-To:
alt.uu.comp.os.linux.questions,alt.os.linux,alt.os.linux.caldera,comp.os.linux.setup,linux.redhat.install,linux.redhat.misc,linux.sources.kernel
Subject: Re: Kernel installation question
Date: Sun, 18 Apr 1999 13:27:08 -0700
:)
------------------------------
From: [EMAIL PROTECTED] (brian moore)
Crossposted-To: comp.os.linux.x
Subject: Re: Linux command: bg and fg
Date: 18 Apr 1999 08:30:48 GMT
On Sun, 18 Apr 1999 01:03:02 -0700,
Jimmy Navarro <[EMAIL PROTECTED]> wrote:
> I use Linux Red Hat 5.2. I'm wondering if after launching X Window by
> "startx" and do Ctl+Alt+F1 afterwards to login with text_based workspace
> but decided to go back to X Window later. Is it possible to foreground
> X Window process or putting current session to background? Here's my
> PID's to see what I meant...
Yep, just do Ctl-Alt-F7.
You should start x with 'startx &', though, or the vty you started X
from will be sorta useless. Or better yet, run xdm, gdm or kdm and
login straight to X.
> It may be a stupid question, huh? Somebody had asked me after an Linux
> Install SIG and I can't give an another. Somehow there's neither "man
> fg" nor "man bg" in Linux Red Hat, I dunno for other distro... (=:
fg and bg are entirely different things and won't help you in this case.
--
Brian Moore | "The Zen nature of a spammer resembles
Sysadmin, C/Perl Hacker | a cockroach, except that the cockroach
Usenet Vandal | is higher up on the evolutionary chain."
Netscum, Bane of Elves. Peter Olson, Delphi Postmaster
------------------------------
From: "Simon" <[EMAIL PROTECTED]>
Subject: problem with my soundblaster AWE64
Date: Sun, 18 Apr 1999 16:46:09 +0200
I can't install my Soundblaster AWE64 on Linux (2.0.35).
Could anyone give me the /etc/isapnp.conf file ?
Is there anything else I should know about this installation ?
Thank you
[EMAIL PROTECTED]
------------------------------
Date: Sun, 18 Apr 1999 08:51:44 +0000
From: ted <[EMAIL PROTECTED]>
Subject: hal91
I have the two disks for hal91..Tha root disk loads ok but the data disk
when loaded gives an error message..Can't find an ext2 file system ondev
02:00...I made sure I could mount the floppy before I dd'd the dat.img
to it` and it was ok but after the transfer I tried to mount it and got
the no ext2
file system....Anyone tell me where I am going wrong ??
--
Regards Ted
g3tpi.ampr.org Mandrake linux
g3tpi@notbbs 44.131.147.8
Ted Wager 10509 261032
Beware of geeks bearing gifts
------------------------------
From: Tom Christiansen <[EMAIL PROTECTED]>
Crossposted-To: comp.unix.bsd.freebsd.misc
Subject: Re: FreeBSD vs. Linux
Reply-To: [EMAIL PROTECTED] (Tom Christiansen)
Date: 18 Apr 1999 06:23:44 -0700
[courtesy cc of this posting sent to cited author via email]
In comp.os.linux.misc, Donn Miller <[EMAIL PROTECTED]> writes:
:An even bigger joke is that NT is POSIX compliant.
Judge the truth of that one for yourself.
--tom
+-------------------------------------------------------------+
| Open Systems, POSIX, and Windows NT - Another Point of View |
| by Heinz Lycklama <[EMAIL PROTECTED]> |
+-------------------------------------------------------------+
"It's Official: Windows NT is Open"
Editorial by Michael Goulde in the November 1995 issue of Open Computing.
"Feds declare NT 'open system'; UNIX takes a hit"
ComputerWorld news headline, July 31, 1995.
"NT is a FIPS-2 certified system, and as such is a 'POSIX-compliant'
operating system." Stated as fact in "redacted decision" by judge from
GSBCA.
What's going on here? What do these statements from recent trade
publications and the judge's "redacted decision" have to do with the
facts? Are any of them true? How did the GSBCA judge come to this
conclusion in the Coast Guard Standard Workstation III award? The
flurry of activity following the US Government bid protest judgment
handed down in June, 1995 has been remarkable. Are thirty lawyers
better qualified to define an "open system", let alone a
"POSIX-compliant operating system" than the technical experts who
produced the POSIX standards and the POSIX.1 Testing Policy?
What we have here is the trade press badly misrepresenting the decision
that was handed down by the US Government, analysts repeating what the
trade press is reporting without doing any real analysis, the judge in
this case making statements which are contrary to the spirit and law of
the NIST POSIX.1 Testing Policy, and at least one "POSIX expert"
agreeing to the judge's statement, even though it is contrary to the
NIST POSIX.1 Testing Policy.
The press has done a great job of clouding the issue of "open systems".
First it was ComputerWorld with its article stating that the government
declares that Windows NT is "open". The November 1995 issue of Open
Computing has an editorial written by Michael Goulde of the Patricia
Seybold Group with the title "It's Official: Windows NT Is Open".
Michael states that the "GSA Board of Contract Appeals declared that
Microsoft Windows NT is an open system." THEY SAID NO SUCH THING! Where
do these analysts/reporters get this from? What's a user to believe?
In his article on "Open Systems, POSIX, and NT" published in the
December, 1995 issue of ;login:, Stephe Walli provides much of the
background of this protest case involving the award of the US Coast
Guard Standard Workstation III (CGSW) RFP to Unisys. He also provides a
summary of the "findings of fact" with some discussion from the judge's
redacted decision. I won't bother to repeat the "findings of facts",
but I do disagree with his conclusions. This article is written to
explain my views on why the decision handed down in this protest bid is
incorrect, and sets a bad precedent for those who promote the use of
open systems -- suppliers and users alike.
I was directly involved in the bid protest trial as an expert witness on
POSIX related issues for the protesters (proud to represent the side of
"open systems", I might add). As the founder of the original /usr/group
Standards Committee which spawned the IEEE POSIX standards efforts, I
care a great deal about how POSIX is viewed and used in the industry.
Someone familiar with this protest made the statement that "Now that
Windows NT has won a large bid by following the rules that the UNIX
community created, UNIX people are crying foul -- That's not right!".
My contention is that Windows NT "won" this bid by the not-so-subtle use
of a "bait-and-switch" policy, and not by rules that the UNIX community
created. Let's look a little deeper into the issues of POSIX compliance
and "open systems" surrounding this protest.
1) Is Windows NT POSIX-compliant?
For that we need to go to the NIST POSIX Testing Policy. This policy
recognizes that the POSIX.1 API's can be hosted by a number of different
configurations, one being a "cooperating-hosted system". In the
definition of terms it states that a "cooperating-hosted system" is "A
Single computer system that provides the functionality of both the
development system and the host implementation with a SINGLE operating
system, and provides the FIPS 151-2 conforming implementation with
ANOTHER operating system." [CAPS added for emphasis.] This definition
was introduced to accommodate the testing of the Windows NT POSIX
Subsystem. No problem here -- the intent is clear when you look at the
three other configurations that had been dealt with by the NIST POSIX
Testing Policy heretofore (native implementation, hosted implementation,
and cooperating system). Windows NT supports multiple operating system
environments, e.g., Win32, OS/2 and POSIX, and thus a new test
configuration definition was required.
In the Certificate of Validation issued by NIST, the implementation
tested is the "Microsoft Windows NT POSIX Subsystem, Version 3.5". It
should also be noted that there are some major deficiencies listed, i.e.
General Terminal Interface devices
Mountable File Systems
Modem Control
Appropriate Privileges
on the Certificate of Validation. These deficiencies carry no legal
binding but it does indicate that the POSIX Subsystem of Windows NT
barely squeaked through the tests.
The Windows NT POSIX Subsystem is the validated product, the "another
operating system" in the definition of cooperating-hosted introduced in
the NIST POSIX Testing Policy. The "single operating system" in this
case is the Windows NT Win32 Subsystem -- that is the development system
that was used to compile the POSIX test suites. The implementation
under test, i.e., the validated FIPS 151-2 product, as identified in the
NIST POSIX Testing Policy, and in the Certificate of Validation is the
"Windows NT POSIX Subsystem". The NIST POSIX Testing Policy says that
"The product identified represents the operating system tested." This is
correctly identified in the Certificate of Validation as the "Windows NT
POSIX Subsystem".
Since the "Windows NT POSIX Subsystem" is certified to be
POSIX-compliant, does this mean that Windows NT is POSIX-compliant? No,
the definitions in the NIST POSIX Testing Policy are very clear that
this is not what is meant. NIST never intended it this way, and NIST
personnel have testified to that. Consider this analogy with the
cooperating system, in which the development system (which is used to
compile the POSIX test suites) and the target system (which is used to
run the POSIX test suites) are two separate computers -- if the target
system is certified to be POSIX-compliant, does that make the
development system POSIX-compliant? I don't think anyone would argue
that -- but that's exactly what is being claimed for the Windows NT
system. The claim is that because the POSIX Subsystem is POSIX compliant
- therefore Windows NT is POSIX compliant.
2) Did the Windows NT-based solution proposed meet the CGSW III RFP?
One of the major requirements of the CGSW III RFP is that certain
applications (email and RDBMS) run under the POSIX operating system. We
interpret this to mean that these applications must run under control of
the POSIX-compliant operating system. For the Windows NT platform
proposed, that means the POSIX Subsystem, which is the operating system
environment which provides the POSIX.1 services. The email and RDBMS
products proposed by the offeror run in the Win32 Subsystem. So how does
this meet the requirements of the RFP?
One of the other major objectives of the RFP was to provide a platform
for portable applications using the NIST Application Portability Profile
(APP) as a framework. Certain standards were selected from this APP for
the CGSW III RFP. These include:
GOSIP (FIPS 146-1)
SQL (FIPS 127-1)
XVT
C (FIPS 160)
Ada (FIPS 119)
Pascal (FIPS 109)
POSIX.1 (FIPS 151-2)
The intent of the NIST APP (and of the POSIX Open System environment
(OSE) as defined in the POSIX.0 Guide for Open Systems Environment, upon
which the NIST APP is based) is that the APIs defined by the standards
be part of an integrated environment so that a portable application can
use any and all of the APIs which are part of the APP. The Windows
NT-based solution provides the GOSIP, SQL, XVT, C, Ada, and Pascal
standards in the Win32 Subsystem, and only C and POSIX.1 in the POSIX
Subsystem. This makes it impossible to write a portable program that
uses all of these APIs in an integrated manner so that the application
can be ported to another POSIX.1 compliant platform. So the solution
proposed defeats the intent of the NIST APP, the government's own
proposed framework for developing portable applications.
Clearly the Windows NT-based solution proposed does not meet either the
letter, the intent, or the spirit of the CGSW III RFP. How did this
happen? If the CG wanted Windows NT, they should not have written
"POSIX operating system" into the requirements, or determined a need for
portable applications for that matter. They should have stated up front
that a proprietary solution such as Windows NT was acceptable. Other
bidders spent millions of dollars to put together bids that complied
with the POSIX and portability requirements.
The government, NIST specifically, spent millions of our tax dollars to
define procurement procedures which would meet the needs of various
government agencies. Part of the effort was to define an Application
Portability Profile which would provide a framework for writing portable
applications, and give the agencies the choice of selecting from
multiple suppliers, knowing that their current applications would still
run on any new platform that they might acquire in the future. This is
called investment protection.
Investment protection was in fact a major objective for the CG since
they wanted to move from a proprietary CTOS environment to an "open
environment" that would give them choice of suppliers and solutions in
the future. The CG has moved from one proprietary platform to another
with the Windows NT solution. The CG also wanted to be able to import
applications written by other government agencies for their "open
platforms". The Windows NT-based solution also defeats this purpose
since portable POSIX compliant applications cannot be ported to the
Windows NT platform.
We even heard the argument that "The language `run under' was used by
the Coast Guard to prevent bidders from proposing solutions of these
applications that were run under emulation." Give me a break! The
English language is not that imprecise that one would believe that the
word "POSIX" was introduced so that we should interpret "POSIX Operating
System" as meaning that this disallows emulation. There is no mention of
emulation in the RFP. This really stretches credibility! In fact, even
the Win32 Subsystem in Windows NT supports Windows applications running
in 16-bit mode using emulation. This interpretation of the RFP language
would even disallow Windows NT as a solution!
Microsoft has a desire to capture as much of the government market for
computing platforms and applications as possible. [They have that right,
but they need to play by the same rules as other suppliers do.] Windows
NT was designed with the government market in mind. As stated in
Chapter 1 of the book "Inside Windows NT" by Helen Custer of Microsoft,
"To meet the government's POSIX procurement requirements, NT would be
designed to provide an OPTIONAL [CAPS added for emphasis] POSIX
application execution environment." This is exactly what Microsoft has
done -- added an optional POSIX subsystem purely for the purpose of
passing a POSIX.1 test suite, but serving no other purpose. No
commercial Microsoft or third party products that use the POSIX
Subsystem have been introduced. It was never intended to be useful. In
fact one can remove the POSIX Subsystem, and all commercial applications
will run just fine.
Let's call a spade a spade. Bottom line, this is a "bait-and-switch"
policy: the POSIX Subsystem is the bait, and the Win32 Subsystem is the
switch. "Yes, we have POSIX, but please use our Win32 Subsystem, i.e.,
WOSA [defined below] instead - it's the only one that really works." Is
this a marketing sham or what? It's like writing a contract to have a
house built with 110 volt sockets. Your contractor builds the house with
220 volt sockets, but with only two 110 volt sockets. The test is --
will your toaster work on the 110 volt socket? Yes, but if you plug in
two appliances on the two 110 volt sockets, a fuse is blown. Oh by the
way, you can plug in as many 220-volt appliances as you want, but you
can only buy those appliances from a factory in Redmond!
3) What makes a system open?
That depends on who writes the definition. Portability, interoperability and
user portability are three agreed-upon key requirements. How well do products
on the market meet these requirements? This has become a very subjective
discussion. The POSIX open system environment (OSE) has four major goals:
application portability
application interoperability
data portability
user portability
with the resulting benefits of:
integration of components from multiple vendors
efficient development and implementation
efficient porting of applications
The NIST APP, which has a strong resemblance to the POSIX OSE, was developed
for the government market to make large government procurements more
cost-effective and efficient, and to promote portability and interoperability
between IT solutions adopted by various government agencies. The key here is
that suppliers and users must agree on an application framework to meet the
stated goals and achieve the benefits listed above.
Computing systems and applications which meet the above stated goals
meet the needs of the government agencies. Application portability only
works if the application uses an integrated set of APIs that fit within
a well-defined applications framework such as the NIST APP or POSIX OSE.
The X/Open application profiles also match the NIST APP and POSIX OSE
very well. Most UNIX systems, and even proprietary operating systems
with integrated "open systems environments", delivered today provide a
consistent set of "open systems" APIs agreed to by players in the open
systems industry. These systems provide open platforms suitable for
applications portability and interoperability.
Given these application profiles/frameworks, openly defined by all
participants in the process, any system vendor can build computing
platforms to meet the requirements, and any ISV can build applications
which fit into the framework. The user has a choice of system providers
and a choice of applications providers. The framework is open and not
controlled by any one vendor. This model fits the government's
standards-based procurement needs and does not lock the government into
any one vendor. This is openness in the purest sense of the word. The
specifications for all important "open system" component interfaces such
as POSIX, X Windows, TCP/IP, CORBA, and now the World Wide Web, were
determined by cooperation among industry suppliers.
4) Is Windows NT "open"?
By whose definition? Does it support application portability? Only if
you move an application from one Windows NT platform to another. Porting
an application from Windows NT to a UNIX system or vice versa is not
easy because the set of APIs used on one system are not necessarily
supported on the other system. UNIX systems provide an integrated set of
APIs which match the requirements and framework of the NIST APP. Windows
NT provides a different set of APIs which do not meet the framework
requirements of the NIST APP, but rather fit within the Microsoft
defined Windows Open System Architecture (WOSA). The POSIX.1 APIs
provided by the Windows NT POSIX Subsystem do NOT fit into the WOSA
framework (by design).
With Windows NT, we have a model where the application architecture or
framework (WOSA) and the APIs are controlled by one vendor - Microsoft.
The user can buy computing platforms and applications from any supplier
who provides Windows NT and applications that fit the WOSA architecture,
all owned by one vendor. Open systems is an attempt to break this
control by one vendor.
By adding the POSIX Subsystem as an appendage to Windows NT, and then
declaring that Windows NT itself is POSIX compliant, Microsoft has
corrupted the concept of "open". Microsoft advertises Windows NT as a
POSIX-compliant operating system, thereby subverting the meaning of
POSIX compliance. Other operating systems suppliers such as IBM, HP,
DEC, and Tandem have added POSIX.1-compliant interfaces to their
proprietary systems in an integrated manner, but these POSIX.1-compliant
environments were meant to be, and are, used by their customers to build
portable and interoperable applications. Microsoft has no such intent,
in providing the POSIX Subsystem appendage to Windows NT.
Users are free to buy proprietary solutions controlled by one supplier.
But if this is what the user intends to procure, then the RFP should
state clearly that all solutions will be considered, open or
proprietary. Don't use POSIX-compliance as a ruse. Don't even use the
term "POSIX compliant" in the RFP if it carries no meaning.
5) Possible responses by the open systems community.
The CGSW III RFP was awarded improperly to a supplier which responded
with a non-compliant operating system, Windows NT. It is a mistake to
let this award stand because of the precedence it sets. Here are some
options that suppliers interested in open systems have:
a) Mount a concerted effort to overthrow this award. By letting it stand the
meanings attached to "POSIX compliant" and "open" remain confused in the
user's mind. From a technical point of view, the U.S. Government's ruling
doesn't have a leg to stand on. The "finding of fact" quoted in the
introduction to this article is in fact false, according to the NIST POSIX
Testing Policy.
b) Mount an open systems marketing effort to shed light on what's really
happening in order to educate confused users. This will encourage users
to write RFP's that result in the procurement of open systems solutions.
RFP's must be written with a lot more precision than they have in the
past. Open systems give users choice.
c) Work with users to strengthen the demand for open systems solutions.
X/Open, OSF and UniForum are in a position where they can help influence
the writing of RFP's that result in the procurement of open systems
solutions. RFP's must be written with much more precision to avoid the
problems encountered with the CG III RFP.
d) Work with NIST to strengthen the RFP requirements writing procedures
to assure that the government acquires open systems solutions that meet
the NIST APP specifications. The government has spent millions of
dollars to develop the NIST APP and the test suites that are used to
measure conformance. Let's not let this effort go to waste.
e) Strengthen industry cooperative efforts to avoid unnecessary
fragmentation and to counter the inroads being made by Windows NT. A
number of steps have been taken by industry players to strengthen the
role of open systems technologies in the past year. We need more open
systems technologies such as X Windows, TCP/IP, NFS, World Wide Web,
CORBA, Java, etc. to give users a choice in buying open systems
solutions from more than one vendor.
6) Postscript
Michael Goulde recants statement. [Open Computing, December 1995 issue,
page 10.]
Open Computing magazine closes its doors. [Unigram X, Issue 567,
December 4-8, 1995.]
Stephe Walli does penance - builds the real McCoy. [Details at
UniForum'96 in San Francisco.]
Sun Microsystems pours hot Java on Microsoft and writes script for
the new game.
Microsoft: "Let's hope Anne Bingaman doesn't read about this."
30 lawyers agree to redefine "open system" by spelling the second word
as s-e-a-s-o-n.
--
Hackers should be judged by their hacking, not bogus criteria such as degrees,
age, race or position. -- From _Hackers_ heroes of the computer revolution --
------------------------------
From: "Mitchell Scott" <[EMAIL PROTECTED]>
Subject: ppp and SuSE 6.0
Date: Sun, 18 Apr 1999 00:24:42 +0100
Reply-To: "Mitchell Scott" <[EMAIL PROTECTED]>
Hi
I am trying to find the easiet way to set up SuSE 6.0 with an internet
connection and I have been doing this in yast Network base configuration.
"ppp modem" I give my ISP phone number my login-name and my password
(account with virgin.net)
yet when I do a ppp-up I get a connection and then the server unexpectedly
dies. If i try and connect using wvdial I get the message "No Carrier" When
I try with "minicom" I can hear the connection but the cursor prompt just
keeps dropping a new line when you type anything on the command line and
nothing happens in fact I have to "control, alt, delete" out of that one!
I have several books on Linux "Hands on Linux" by Sobell "Running Linux"
OREILLY and of course the SuSE Linux hand book, I have tried but with no
luck to follow all the different examples off connecting to the internet
what is your advise to me
by the way virgin.net offer no tech support for Linux?
Have been working with Linux for about three months now and I realise once
it has been configured correctly it is a much more powerful o/s system than
windows but... why is it written in such a way that configuring peripherals
such as printers ,modems, soundcards, scanners etc is so compilicated to new
users, after weeks and weeks of trying to configure my Lexmarc 3000 printer
I have no control over the printing I can only get it to print in 300x300
landscape ! I have given up on the soundcard and scanner configs I would
like to be able to use the mail utility but obviously the modem must be set
up first.
Thanks for any help on these matters
Mitchell
------------------------------
From: [EMAIL PROTECTED] (Walter Strong)
Subject: Re: This is not an oppinion - just an interesting fact.
Date: 18 Apr 1999 16:26:59 GMT
Desmond ([EMAIL PROTECTED]) wrote:
: --------------B6953DE0E78465371D0872DA
: Content-Type: text/plain; charset=us-ascii
: Content-Transfer-Encoding: 7bit
: I saw a friend of mine compiling and running KDE under NT
: --------------B6953DE0E78465371D0872DA
: Content-Type: text/html; charset=us-ascii
: Content-Transfer-Encoding: 7bit
: <HTML>
: I saw a friend of mine compiling and running KDE under
: <U>NT</U></HTML>
: --------------B6953DE0E78465371D0872DA--
Could this same friend teach you how to post to a ng? That would
be interesting.
------------------------------
From: [EMAIL PROTECTED] (Bob Tennent)
Subject: missing file for gnome
Date: 18 Apr 1999 16:28:58 GMT
Reply-To: rdt(a)cs.queensu.ca
I've upgraded all the gnome packages and now it won't start.
I get the following message in .xsession-errors:
GImLib ERROR: Cannot find palette file /usr/etc/im_palette.pal
I can tell by using locate that I *had* this file before I upgraded.
Which package removed it? Where do I find it now?
Bob T.
------------------------------
** 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: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
