Linux-Misc Digest #788, Volume #20 Fri, 25 Jun 99 16:13:08 EDT
Contents:
Re: Documentation, or the lack thereof (Alan Curry)
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News (Terry Carmen)
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News ("John Hughes")
Re: NT the best web platform? ("John Hughes")
SuSE GmbH Plans IPO (Harold Stevens ** PLEASE SEE SIG **)
long filenames unsupported on mounted cd's (Uri Kerbel)
fstab and mount troubles (Aaron)
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News (Terry Carmen)
Re: Screen fonts size in console view (Karl-Heinz Zimmer)
upgrade to RH 6.0 broke my printer ("gregg aronson")
HELP!! - How to prevent process swapout on Linux. (Jae Song)
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft Retest
News (Terry Carmen)
Re: UNIX / LINUX Compatibility (John Girash)
Re: libs all around (gtk+) (Paul Kimoto)
Re: fstab and mount troubles (Ted Sikora)
Re: Docbook? Linuxdoc? Re: Documentation issues. (Cameron L. Spitzer)
Re: Does lilo exist for DOS? (Doug DeJulio)
Re: COL225: Upgrading a 2.2.5 kernel to 2.2.9 (Julius Longauer)
----------------------------------------------------------------------------
Crossposted-To: comp.unix.bsd.misc,gnu.misc.discuss
Subject: Re: Documentation, or the lack thereof
From: [EMAIL PROTECTED] (Alan Curry)
Date: Fri, 25 Jun 1999 18:31:08 GMT
In article <[EMAIL PROTECTED]>,
Tom Christiansen <[EMAIL PROTECTED]> wrote:
>In gnu.misc.discuss, [EMAIL PROTECTED] (Alan Curry) writes:
>:With man pages as the only motivation, nobody wants to learn roff.
>
>Writing papers. But you're right, nobody wants to learn anything.
>Then again, I have little sympathy for people who don't want to learn.
Please have sympathy for those who want to learn, but don't have time to
learn both roff and tex, and choose to study the better one.
>:In summary, if you don't want man to continue its slow death, here's how you
>:start:
>: 1. decouple the man utilities from the roff format
>: 2. produce a version of man(1) with search options, like "show all pages
>: matching *sh in section 1", or "dump the SYNOPSIS section from ytalk(1)."
>: When you do this, do not destroy pipeline-friendliness!
>
>We've largely done that. Pod is the first part, and a zillion little
>happy CLI tools I've written make up the second.
Good. Next step: convince other programmers (those who are currently churning
out texinfo or you-must-run-the-program-and-point-and-click-the-help) to
switch to pod.
>:At this point the info and SGML people will still be whining about the lack
>:of hyperlinks.
>
>They have. They lost. I'd rather write in Latin in a candle-lit
>scriptorium than in docbook or similarly heavy SGML-smelling
>unpleasantries. Gets in my way.
They didn't lose. They fragmented the documentation by inventing new formats
that suck. The losers are the reasonable people who want a system with a
single set of pipeline-friendly documentation tools that work on _all_ the
docs. And we'll continue to lose, until we make man attractive to them.
Making pod a real alternative to roff would be a step in the right direction.
By this I mean that man(1) should be able to do pod2text on pages it finds in
$MANPATH/pod1/foo.1.gz, and cache the results in $MANPATH/cat1/foo.1, just
like it does for roff pages.
On the other side of things, we need to give the point-and-clickers a way to
make that [Help] button they love so much work _with_ the system
documentation instead of against it. It should probably invoke pod2html piped
into the user's chosen HTML viewer. This is not hard, but if it's not in
libgnome and/or libkde nobody is going to use it.
--
Alan Curry |Declaration of | _../\. ./\.._ ____. ____.
[EMAIL PROTECTED]|bigotries (should| [ | | ] / _> / _>
==============+save some time): | \__/ \__/ \___: \___:
Linux,vim,trn,GPL,zsh,qmail,^H | "Screw you guys, I'm going home" -- Cartman
------------------------------
From: [EMAIL PROTECTED] (Terry Carmen)
Crossposted-To:
comp.os.linux.networking,omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Fri, 25 Jun 1999 17:55:34 GMT
On Thu, 24 Jun 1999 22:28:58 GMT, [EMAIL PROTECTED] (Timothy
Kelley) wrote:
>On Thu, 24 Jun 1999 18:18:58 GMT, [EMAIL PROTECTED] (Terry Carmen)
>wrote:
>
>>FWIW, the most reliable servers I've ever seen were Netware 3.x and 4,
>>where they were simply installed, stuck in a closet and ignored. Many
>>have been running for years with no reboots or failures.
>
>Yeah but when there's a problem in netware it is usualy a real big
>problem. I think it is even worse than NT in many respects. Most
>people think netware is so stable because it really isn't doing much
>of anything.
I don't see that as a problem. There are worse things than having a
whole machine who's purpose in life is to give you files when you ask
for them and keep out Evil People.
OTOH, when it goes, it's usually a disk failure, and yeah, it's
usually very bad.
Terry
"It's much easier to develop software using actual technology, instead of just made-up
stuff."
------------------------------
From: "John Hughes" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.networking,omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Fri, 25 Jun 1999 19:30:14 +0100
Well you couldnt have configured it properly. We are running an NT box for
web serving and email. Also 2 other servers with SQL Server and we havent
had any crashes in 2 years.
yan seiner <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Terry Carmen wrote:
> >
> > While it's very easy to bash NT and come up with amazing statistics
> > supporting either operating system, I should mention that NT is very
> > stable if you do a proper install on certified hardware and don't load
> > it up with a bunch of crap.
> >
> > If you take NT certified hardware, install NT, the web server of your
> > choice and a recent service pack, then log off and walk away from the
> > console, it will run quite nicely for a very long time.
>
> Hmmm. I was running NT server on a name brand system, stock install,
> the only two packages that were non-MS were wingate and seagate backup.
> I made sure the service packs were installed (with R&RAS you have no
> choice.)
>
> As MS-Proxy is way too expensive, and NT does not come with a backup
> that's any good, I'd say that's a minimum of stuff.
>
> No IIS, no nothing. The ONLY thing this server was supposed to do was
> provide file service and proxy service. That's it.
>
> It never ran longer than a week without a crash of some sort.
> Typically, it was R&RAS that quit working; file service was pretty
> stable (except for the time that the server froze, and took the entire
> NTFS with it. Seagate backup proved its worth that day).
>
> I reinstalled the entire server 3 times in one year. After 18 months of
> this, the overhead of maintaining the damn thing become too great.
>
> >
> >
> > This is a central flaw in the design of Windows, and one of it's
> > dirtiest secrets, that there are a virtually infinite number of
> > possible DLL combinations, depending on what's been installed, and
> > they're not all compatible. However, if you install a good set and
> > leave it alone, it's usually OK.
> >
>
> Absolutely - I agree that the entire Win32 paradigm is fundamentally
> flawed at its core levels. Every software package mucks about with the
> core OS, both in the registry and the DLLs. It is impossible to design
> in stability, or even make a realistic assumption of system behavior,
> when the admin has no control over what is actually running. Worst of
> all, there is no way to truly uninstall a package once the DLLs have
> been changed a few times.
>
> Yan
------------------------------
From: "John Hughes" <[EMAIL PROTECTED]>
Crossposted-To: comp.infosystems.www.servers.unix,comp.os.linux.advocacy
Subject: Re: NT the best web platform?
Date: Fri, 25 Jun 1999 19:33:18 +0100
FREE means nothing if your spending more time configuring. NOT saying thats
the case with Linux but this FREE mentality just doesnt cut it in business.
What really matters is that it delivers needed business services.
Ted Gamron <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> The best benchmark. LINUX is Free.
>
>
> Donovan Rebbechi wrote:
>
> > On Thu, 24 Jun 1999 17:48:54 +0100, John Hughes wrote:
> > >There are numerous benchmarks showing NT to be faster. Where are the
Linux
> > >ones?
> >
> > (a) The recent PC Week benchmark showed linux/SAMBA beating NT at
serving NT
> > clients.
> > (b) An older PC Week benchmark on cheaper hardware showed
linux/SAMBA doing
> > better
> > (c) Who cares anyway ? Benchmarking static server performance is
pointless.
> > If anyone cared about the *speed* of webservers, Apache would not be the
> > MARKET LEADER. Apache is much slower than other UNIX webservers, but it
> > is still the most popular because of its functionality, reliability,
> > and security.
> >
> > >Apart from mouthing off why doesnt the Linux community get some
benchmarks
> >
> > The linux community isn't in the business of performing sleazy stunts
such
> > as the Mindcraft fiasco. *Regardless* of of what the linux community
do,
> > independent industry observers ( such as PC Week ) will continue to
perform
> > benchmarks. The kernel developers are doing the right thing by doing
what they
> > can to make sure that linux performs, rather than buying into sleazy
> > publicity stunts.
> >
> > --
> > Donovan
>
------------------------------
From: [EMAIL PROTECTED] (Harold Stevens ** PLEASE SEE SIG **)
Subject: SuSE GmbH Plans IPO
Date: Fri, 25 Jun 1999 18:53:02 GMT
FWIW:
http://linuxtoday.com/stories/7107.html
Regards, Weird (Harold Stevens) ** IMPORTANT EMAIL INFO **
1. As antispam, I have completely disabled my "adam" email account.
2. Please vent inconvenience at Cyberpromo and their Satanic spawn.
3. Please look for (wyrd) at raytheon, dotted with com. NO UCE/UBE.
4. I detest UCE/UBE. I support CAUCE; http://www.cauce.org HR 1748.
Standard Disclaimer: My opinions, and not Raytheon Systems Company.
------------------------------
From: [EMAIL PROTECTED] (Uri Kerbel)
Subject: long filenames unsupported on mounted cd's
Date: Fri, 25 Jun 1999 18:05:41 GMT
Despite have selected CDROM support under xconfig, when I mount cd's,
long filenames are not supported, i.e. the tilde (~) is used to
shorten them.
I am currently using kernel 2.2.9 on an Intel machine.
Does anyone know how to rectify this please?
TIA
Uri
--
[EMAIL PROTECTED]
===================================================
For free Java applets, utilities and scripts to enhance your site,
visit - http://www.kalbi.demon.co.uk/
===================================================
------------------------------
From: [EMAIL PROTECTED] (Aaron)
Subject: fstab and mount troubles
Date: Fri, 25 Jun 1999 18:38:01 GMT
I am having trouble mounting my windows drive. I got it to work once,
then I tried to add it to fstab and now I can't get it to work at all.
The drive is hda1. My original mount command was:
mount -t vfat /dev/hda1 /mnt/drive1
That worked the first time. Then I decided I wanted it to be
automatically mounted at startup and I also wanted it in /drive1
instead of mnt/drive1, so I added following to fstab and created the
/drive1 directory.
/dev/hda1 /drive1 vfat user 0 0
ftstab also contained the folowing line:
/dev/hda
What does that do? Do I need to worry about the spacing on the fstab
lines? Anyway, the automount didn't work and now I can't mount it
manually or the new mount point /drive1 or the old mount point
/mnt/drive1. The error I get is:
mount: wrong fs type, bad option, bad superblock on /dev/hda1, or too
many mounted files systems.
Any clues or advice?
------------------------------
From: [EMAIL PROTECTED] (Terry Carmen)
Crossposted-To:
comp.os.linux.networking,omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Fri, 25 Jun 1999 17:56:38 GMT
On 24 Jun 1999 16:23:27 -0700, [EMAIL PROTECTED] (Jason O'Rourke) wrote:
>Terry Carmen <[EMAIL PROTECTED]> wrote:
>>While it's very easy to bash NT and come up with amazing statistics
>>supporting either operating system, I should mention that NT is very
>>stable if you do a proper install on certified hardware and don't load
>>it up with a bunch of crap.
>
>What the heck is certified hardware?
It's hardware listed by Microsoft as having been tested with the OS
and found to work properly.
> Is a HP Kayak workstation certified?
Beats me. I never looked for it.
Terry
"It's much easier to develop software using actual technology, instead of just made-up
stuff."
------------------------------
From: Karl-Heinz Zimmer <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Screen fonts size in console view
Date: Fri, 25 Jun 1999 17:49:04 GMT
Am 25.06.1999, 18:42:29, schrieb [EMAIL PROTECTED] (Knarf):
> I would like to change the size of the screen font (make
> it smaller) when I work in text editor (emacs) so I can
> have a larger view.
> i have tried stty but with no success...
A look at ''The Linux Console Tools'' might help.
http://www.mygale.org/~ydirson/en/lct/
will take you to the pages...
HTH,
Karl-Heinz
--=20
"If you think of MS-DOS as mono, and Windows as stereo, then Linux is
Dolby Pro-Logic Surround Sound with Bass Boost
and all the music is free." Brendan Bradley
------------------------------
From: "gregg aronson" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux
Subject: upgrade to RH 6.0 broke my printer
Date: 25 Jun 1999 18:36:50 GMT
I had my printer working fine under RedHat 5.2 along with my Internet
connection and some other stuff. Then I upgraded to 6.0. I didn't notice a
problem until I installed WordPerfect and tried to print a test document.
Printing failed. Then I tried to do a regular lpr type print of an ASCII
file, and that failed too. I could also no longer pront from Acrobat
Reader, which I could do before.
I tried using the printtool to uninstall and reinstall my printer, but I
got a message indicating that the system couldn't find LP0, LP1, or LP2. It
looks to me like linux can't find my parallel port. Does anyone know how to
fix this. I really don't want to do a reinstall of Redhat 6.0 and blow all
my configurations away.
Thanx for any help anyone might be able to give me.
--gregg
I block spambots. You can write me directly at:
garonson at sns dot com
------------------------------
From: Jae Song <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: HELP!! - How to prevent process swapout on Linux.
Date: Fri, 25 Jun 1999 15:31:53 -0400
First, description of our machine config. Linux 2.2.5-22 (SMP) on a
dual Pentium II - Xeon 400 MHz.
We're running multiple processes on this machine and all of these
processes access the shared memory segment (ipc facility). In order to
main validity of the data, each process locks out all other processes
while reading and writing. (For example, we would not want a process to
be in the middle of reading a block when another process is writing to
the same block. And, vice versa.) But, it seems that after a process
locks the shared memory segment, it is getting swapped out by the system
scheduler before it finishes its task (reading or writing and then
unlocking the memory segment so that other processes can get at the
memory segment). So the result of this is, other processes can't get
access to the shared memory segment because the process that had locked
it had been swapped out and won't unlock it until it is swapped back in.
My question is, is there any system call that I can make to allows a
process to be immune to the swap-out for a certain amount of time? I
know that under IRIX, there is prctl(...) with PR_RESIDENT that can make
a process immune to swapouts. But, this doesn't seem to be implemented
in Linux yet.
I would like to do something like:
_system_call_to_make_process_to_be_immune_to_swapout
lock_shared_memory_segment
read_or_write_to_memory_segment
unlock_shared_memory_segment
_system_call_to_cancel_swapout_immunization
Any help would be greatly appreciated.
Jae.
------------------------------
From: [EMAIL PROTECTED] (Terry Carmen)
Crossposted-To:
comp.os.linux.networking,omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: Mindcraft
Retest News
Date: Fri, 25 Jun 1999 17:53:40 GMT
On Thu, 24 Jun 1999 14:24:09 -0700, [EMAIL PROTECTED] () wrote:
>>
>>While it's very easy to bash NT and come up with amazing statistics
>>supporting either operating system, I should mention that NT is very
>>stable if you do a proper install on certified hardware and don't load
>>it up with a bunch of crap.
>
> But loading a machine up with crap is what it's there for.
> This is especially true in with Microsoft based machines
> as they're supposed to be better due to the wider range of
> crap available for them.
I actually prefer simpler servers.
My very favorite is Netware 3.1 with no 3rd party NLMs.
Just start it up, configure it, and lock the door.
I've got installations that haven't been touched in ages, and have
never crashed, and in some cases the customer doesn't even know where
the server is.
Terry
"It's much easier to develop software using actual technology, instead of just made-up
stuff."
------------------------------
From: John Girash <[EMAIL PROTECTED]>
Crossposted-To:
alt.unix,alt.unix.geeks,alt.unix.wizards,alt.unix.wizards.free,ca.unix,comp.os.linux.development.apps,comp.os.linux.questions
Subject: Re: UNIX / LINUX Compatibility
Date: 25 Jun 1999 15:17:54 GMT
In comp.os.linux.misc anthony classick <[EMAIL PROTECTED]> wrote:
> If it runs on your computer, it may not run on the host even if they are
> both Linux! Sorry to give you such grim news, but I have tested "BASH"
> scripts on my SuSE linux machine, uploaded it to www.one.net's server,
> which is Linux (unknown version), and they failed! I had to test my CGI
> scripts "live" (with an html page that used the cgi script, and a
> browser), making alterations after seeing the browser output.
Just a thought, but are you sure they were being run under bash (and not csh
or ksh or whatever) on one.net? That'd be my first guess as to where the
imcompatibility lies especially for scripts (as opposed to binaries).
> Don't count on ANY commonality.
Well, uh, yeah. Test it! Test test test test test. I like tests.
followsup culled to something more reasonable
jg
------------------------------
From: [EMAIL PROTECTED] (Paul Kimoto)
Subject: Re: libs all around (gtk+)
Date: 24 Jun 1999 23:53:25 -0500
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>, Vic Mortelmans wrote:
> I will add another question as well: I'm trying to install the
> gtk+-libs, and the installation script is spreading (huge) .a-files in
> my lib-dirs. Do I need them?
>
> What are .a-files anyway? a.out-libs? Something to do with
> static/dynamic?
Yes, they are the "static libraries", used by the linker (at the
end of the compilation procedure) to get the object code, which
(in this case, as opposed to the "dynamic" case) is copied into
the compiled executable.
For most garden-variety libraries, they are redundant with their
dynamic equivalents (but glibc2 is a slight, but very important[!],
exception).
--
Paul Kimoto <[EMAIL PROTECTED]>
------------------------------
From: Ted Sikora <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: fstab and mount troubles
Date: Fri, 25 Jun 1999 19:29:51 GMT
Aaron wrote:
>
> I am having trouble mounting my windows drive. I got it to work once,
> then I tried to add it to fstab and now I can't get it to work at all.
> The drive is hda1. My original mount command was:
>
> mount -t vfat /dev/hda1 /mnt/drive1
>
> That worked the first time. Then I decided I wanted it to be
> automatically mounted at startup and I also wanted it in /drive1
> instead of mnt/drive1, so I added following to fstab and created the
> /drive1 directory.
>
> /dev/hda1 /drive1 vfat user 0 0
>
> ftstab also contained the folowing line:
>
> /dev/hda
>
> What does that do? Do I need to worry about the spacing on the fstab
> lines? Anyway, the automount didn't work and now I can't mount it
> manually or the new mount point /drive1 or the old mount point
> /mnt/drive1. The error I get is:
>
> mount: wrong fs type, bad option, bad superblock on /dev/hda1, or too
> many mounted files systems.
>
> Any clues or advice?
Try:
/dev/hda1 /drive1 vfat rw defaults 1 1
in /etc/fstab.(Make sure the /drive1 directory exists.)
Under /proc/filesystems you do have:
umsdos
msdos
vfat
listed?
--
Ted Sikora
[EMAIL PROTECTED]
http://tsikora.tiac.net
------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Crossposted-To: comp.unix.bsd.misc,gnu.misc.discuss
Subject: Re: Docbook? Linuxdoc? Re: Documentation issues.
Date: 25 Jun 1999 18:58:32 GMT
In article <[EMAIL PROTECTED]>, Richard Kettlewell wrote:
>[EMAIL PROTECTED] (Cameron L. Spitzer) writes:
>> One of the reasons the LILO mini-HOWTO is out of date is that I have
>> never seen any documentation about the Linuxdoc format. I've
>> written other HOWTO maintainers and they have told me they couldn't
>> find any, either, and just imitated other HOWTOs.
>
>The Debian sgml-tools package contains some documentation on Linuxdoc
>- if you don't have a Debian system yourself you can either unpack the
>source archive with gzip/tar or the binary package with ar/gzip/tar.
>Start at http://www.debian.org/ .
I posted that from a Debian-2.0 system with everything sgml-related
installed.
There's a thing called /usr/doc/sgml-base. It's about "entity management."
A classic case of useless "documentation." That is, documentation for
people who don't need any documentation because they already know
more about the subject than is revealed in the "documentation."
There's a thing called /usr/doc/sgml-data. Nothing in there but a
changelog and a copyright.
There's a /usr/doc/sgml-tools/html which I didn't find before because
the files are hidden with the unusual extension html.gz. It's from
1996 and documents SGML-Tools version 0.99.0.
I have ported my copy to work with Apache, by gunzipping and
then fixing the internal links.
Buried deep in there, I find what might be a useful file:
http://petra.greens.org/about/sgmltools/sgml-tools/html/guide-3.html
Now we can play that fun game for Linux users, guess the context.
I wonder if guide-3.html still applies. What's the relationship between
Docbook and Linuxdoc? Has Linuxdoc been updated since November '96?
Has it been replaced by Docbook? Is guide-3.html an exhaustive list
of tags recognized by sgmltools? Or is it just the tags Matt Welsh
was using when he burned out on the Linux Documentation Project.
See the notes at the very top and very bottom of
http://www.sgmltools.org/
>
>The Debian package cites http://pobox.com/~cg/sgmltools/dist/ as the
>location of the upstream source.
What does that mean? How does a package cite something?
I found http://www.debian.org/Packages/stable/text/sgml-tools.html
but there is no mention of http://pobox.com/~cg/sgmltools/dist/
there.
Cameron
------------------------------
From: [EMAIL PROTECTED] (Doug DeJulio)
Subject: Re: Does lilo exist for DOS?
Date: 25 Jun 1999 12:03:44 -0400
In article <7kvbga$92r$[EMAIL PROTECTED]>,
Villy Kruse <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>,
>Ken Kwasnicki <[EMAIL PROTECTED]> wrote:
>>Hi All,
>>
>>Does anybody know if there's a dos version of lilo, or some way to
>>modify the lilo configuration in the master boot record from dos?
>
>
>Afraid not. The lilo command needs help from the kernel to get the
>absolute disk addresses of every block in the boot images, that is,
>vmlinuz, and maybe the initrd file. You could probably make a dos
>program that could do this, but it needs to know everything about the
>linux file systems to do this.
What about something that *just* handles the "lilo -R" functionality
of LILO?
I've got multiple OSes, and sometimes I want to select which one will
boot next so I can reboot the machine and walk away for a bit. I can
do a "lilo -R BeOS" when I'm in Linux, but not when I'm in Windows or
BeOS.
--
Doug DeJulio | mailto:[EMAIL PROTECTED]
HKS, Incorporated | http://www.hks.net/~ddj/
------------------------------
From: Julius Longauer <[EMAIL PROTECTED]>
Subject: Re: COL225: Upgrading a 2.2.5 kernel to 2.2.9
Date: Fri, 25 Jun 1999 18:07:03 +0200
Chris Raper wrote:
>
> On Fri, 25 Jun 1999 16:55:10 +0200, Julius Longauer
> <[EMAIL PROTECTED]> wrote:
>
> >> make[1]: Entering directory `/usr/src/linux-2.2.5/arch/i386/boot'
>
> >Oops, obviously you have installed the 2.2.9 kernel right over
> >the 2.2.5 kernel source tree.
>
> Hi Julius
>
> Sounds a good idea - will I lose my configuration from the old kernel
> and have to set a new one? Is there any way of keeping the kernel
> config from 2.2.5 but in 2.2.9?
>
Of course :-) You can find the current kernel configuration in
/usr/src/linux-2.2.5/.config. Copy it to /usr/src/linux-2.2.9/ and
run 'make oldconfig' within this directory. oldconfig runs
a line-oriented interface, which automatically re-uses old config
values and prompts you only, if there are any new config options not
known in the former kernel. After that you can directly start with
'make dep' and so on.
Julius
------------------------------
** 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
******************************
