Linux-Misc Digest #43, Volume #25 Tue, 4 Jul 00 19:13:02 EDT
Contents:
Re: A confusing, but interesting topic... ("David E. Smith")
Re: kppp amd redhat (Hamid Misnan)
Re: Corel PhotoPaint & RH 6.2 problems (Mike Frisch)
Re: Netscape and RedHat 6.2 (Charles Leslie)
Re: Help! root directory mounted as readonly. What to do? (David)
Re: Memory Leak in 2.2.x series? (Vilmos Soti)
Re: Corel PhotoPaint & RH 6.2 problems (Prozessor)
Re: tool for joining various (text) files, editing and splitting them (Kai
=?iso-8859-1?q?Gro=DFjohann?=)
Re: Debian 2.1 upgrade failed !!! (Barry Samuels)
Re: Problems with fsck on new disk (David Efflandt)
Re: RAID syncronization problem (Max TenEyck Woodbury)
Re: compiling gnumeric ("Kevin Vandersloot")
Re: inn, suck, nntpsend (Sylvain POURRE)
Re: A confusing, but interesting topic... ([EMAIL PROTECTED])
Re: Corel PhotoPaint ([EMAIL PROTECTED])
Re: ANSI Colors in a C program under Linux... (William Wueppelmann)
----------------------------------------------------------------------------
From: "David E. Smith" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,nf.comp.linux
Subject: Re: A confusing, but interesting topic...
Date: 4 Jul 2000 19:35:18 GMT
In our last shocking episode of comp.os.linux.misc,
Hendrix revealed a dark secret:
> Someone told me that Red Hat just installs everything from a binary (RPM
> I guess) image. Maybe the boot disk provides an interface to activate
> the CD so that it acts like a working system to install and configure
> the new system...
Essentially, yes.
Each of the hundreds or thousands of individual programs that makes up a
Linux distribution is pre-compiled, usually for a "lowest common
denominator" sort of environment. (For Intel chips, this usually means
compiling things so that they'll run on a 80386. This guarantees that for
people still using these older computers, even the newest distribution
will -- theoretically -- run.)
Each of those programs is packaged up in some way that makes sense. Most
of the stuff in /bin might be one package, /sbin might be another. Or they
might bundle ls, cp, touch, and other basic file-manipulating programs in
one package. Whatever. The point is, they're usually bundled into
packages. (There's nothing that requires this other than the wavering
sanity of the people that create and maintain distributions.)
A typical distribution CD will contain a "mini-linux" system to be written
to a floppy disk, which has enough drivers to read the CD. Or the CD
itself might be bootable, for those with newer computers that support such
things. Or both.
...dave
------------------------------
From: [EMAIL PROTECTED] (Hamid Misnan)
Crossposted-To: comp.os.linux.networking
Subject: Re: kppp amd redhat
Date: 4 Jul 2000 20:10:12 GMT
On 01 Jul 2000 18:52:01 EDT, [EMAIL PROTECTED] wrote:
>Hi all, I just installed RH6.2 update-2 on my machine (KDE option with GUI
>intallation) and the first thing I've noticed is thakppp can only be run by
>root.When I try to run kppp as a user it compalins that the server refuses
>to connect to 0:0 or something to that effect. As far as I can see there is
>no permissions problem kppp is 777.
>Does anyone know how to fix that?
Here is what I've in my system, it doesn't ask for a password and I've been
using this config since day one I use RH 6.2:
[mhmsys@n30207 mhmsys]$ cat /etc/pam.d/kppp
#%PAM-1.0
auth sufficient /lib/security/pam_permit.so
auth required /lib/security/pam_pwdb.so
session optional /lib/security/pam_xauth.so
account required /lib/security/pam_permit.so
[mhmsys@n30207 mhmsys]$ cat /etc/security/console.apps/kppp
USER=root
PROGRAM=/usr/sbin/kppp
SESSION=true
[mhmsys@n30207 mhmsys]$ ls -la /usr/bin/kppp
lrwxrwxrwx 1 root root 13 May 21 19:32 /usr/bin/kppp -> consolehelper*
[mhmsys@n30207 mhmsys]$ ls -la /usr/sbin/kppp
-rwsr-xr-x 1 root root 373288 Mar 7 21:46 /usr/sbin/kppp*
--
|Mohd Hamid Misnan | http://www.mhmsys.com | [EMAIL PROTECTED] |
|AMD-RH6.2 & iMac Bondi Blue & PalmVx & eTrex | [EMAIL PROTECTED] |
- The three Rs of Microsoft support: Retry, Reboot, Reinstall.
------------------------------
From: [EMAIL PROTECTED] (Mike Frisch)
Subject: Re: Corel PhotoPaint & RH 6.2 problems
Date: Tue, 04 Jul 2000 20:20:22 GMT
On 4 Jul 2000 12:30:32 -0500, Dave Brown <[EMAIL PROTECTED]> wrote:
>The install log claims that the install was successful. (...Although I
>got no icons on my window manager menus. I have to start from the
>command line.) I'm looking for answers as well.
Don't expect icons... all window managers have their own method of
configuring the menus. Presumably, Corel will support KDE/GNOME's common
interface, but if you're not running either of these, chances are good
that you'll be out of luck.
------------------------------
From: Charles Leslie <[EMAIL PROTECTED]>
Subject: Re: Netscape and RedHat 6.2
Crossposted-To: comp.os.linux.questions,comp.os.linux.redhat,comp.os.linux.x
Date: Tue, 04 Jul 2000 20:23:11 GMT
Wish that was the case, I have 384 Megs of Ram and about 512 Meg of swap=
=20
space running on a PIII 700mhz.
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 7/4/00, 10:59:00 AM, [EMAIL PROTECTED]=20
(Chiefy) wrote regarding Re: Netscape and RedHat 6.2:
> On Tue, 04 Jul 2000 07:02:25 GMT, Charles Leslie <[EMAIL PROTECTED]> =
wrote:
> >I have the weird behavior from Netscape. For some strange reason,
> >whenever, I'm using Netscape, my Xserver crashes. I'm using the defa=
ult
> >window manager, Gnome, with Redhat 6.2 and have installed the Metro-X=
> >3.4.5 xserver.
> Hello, not sure if this is related,
> I experienced problems with Netscape and RedHat 6.0. I had only 20Mb o=
f
> memory, and it was pointed out that 32Mb was a minimum for reliable
> service, more if used with Gnome.
> I never did get the memory, so I'm not sure if that would have solved
> things, but it might be worth considering.
> Regards
> Martin.
------------------------------
From: David <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Help! root directory mounted as readonly. What to do?
Date: Tue, 04 Jul 2000 13:18:41 -0700
Thanks for your replies, guys. I rebooted off of a emergency disk and ran
e2fsck manually. A couple of inodes were inconsistent, it fixed it and now
it's fine again. During "e2fsck" I got the message that the file "wtmp" had
a deleted/unused inode. It told me that it was put into "/lost+found". That
was easy.
Thanks again - David.
Fabian Gebhardt wrote:
>
> You should not repair a mounted disk.
>
> Try to boot a rescue Linux and run 'e2fsck /dev/hda5' manually.
>
> After you repaired it you can try to boot from it.
>
> HTH
> --
> CU, Fabian Gebhardt
>
> E-Mail: [EMAIL PROTECTED]
> ICQ#: 77948091
> Homepage: http://www.ki.tng.de/~gebhardt
> Schul-Seite: http://www.ebg.org
------------------------------
Subject: Re: Memory Leak in 2.2.x series?
From: Vilmos Soti <[EMAIL PROTECTED]>
Date: Tue, 04 Jul 2000 20:26:59 GMT
"Jeff Shultz" <[EMAIL PROTECTED]> writes:
> The problem is that when I start and end programs (esp. Netscape) I
> never get back as much memory as they've used. I _never_ get back
> any buffer memory (what is this?) and it goes on to eat up multiple
> hundreds of megabytes of my RAM. Cache gets fairly large too... but
> buffer just keeps growing and growing...
>
> It's ridiculous to be running nothing but Netscape, KDE Task Manager
> and xosview and to have 80mb being used for program memory, 149mb
> used for buffer, the rest used for cache... and already be 5mb into
> my swap file.
>
> Anyone have any ideas?
Linux uses free memory as cache. This is what happens in your case.
When you need memory and you don't have enough free, then the kernel
will free some of the cache. This is normal and is good. This is
dynamic caching. You don't have to set aside any part of memory
as disk cache.
Vilmos
------------------------------
From: Prozessor <[EMAIL PROTECTED]>
Subject: Re: Corel PhotoPaint & RH 6.2 problems
Date: Tue, 04 Jul 2000 20:49:28 GMT
Mike Frisch wrote:
> =
> On 4 Jul 2000 12:30:32 -0500, Dave Brown <[EMAIL PROTECTED]=
> wrote:
> >The install log claims that the install was successful. (...Although =
I
> >got no icons on my window manager menus. I have to start from the
> >command line.) I'm looking for answers as well.
> =
> Don't expect icons... all window managers have their own method of
> configuring the menus. Presumably, Corel will support KDE/GNOME's comm=
on
> interface, but if you're not running either of these, chances are good
> that you'll be out of luck.
I=B4ve tried it under KDE anf Gnome as well and no icons appeared.
But you have the possibility to make the menuentry or an icon on the
desktop by yourself by calling the script /usr/bin/potopaint.
Another thing:
Did anyone of you succeed in running PhotoPaint using a
preassuresensitve graphicstablet?
There are tablet settings in the Options Dialog but they are inactive.
Cursormovement works but preassuresensitivity don=B4t.
i=B4m running a Wacom Artpad II using the XFree Wacom Driver.
Has anyone ideas?
------------------------------
From: [EMAIL PROTECTED] (Kai =?iso-8859-1?q?Gro=DFjohann?=)
Crossposted-To: comp.os.unix.misc,comp.unix.questions,comp.unix.shell
Subject: Re: tool for joining various (text) files, editing and splitting them
Date: 4 Jul 2000 21:05:38 GMT
Uwe Brauer <[EMAIL PROTECTED]> writes:
> [EMAIL PROTECTED] (Kai Gro�johann) writes:
>
> > Using Emacs, it might be possible to write some Lisp which does this.
> > Hm. You'd have to assign a text property giving the name the file
> > came from, and then when saving the buffer, you'd have to go through
> > all the files and save each region. Hm. Yes, that could work.
>
> I guess with Lisp it would be complicated and slow.
I'm pretty sure it wouldn't be all that slow. As to
complicated... hm.
(defun ub-append-file-to-buffer (f)
(interactive "fAppend file to buffer: ")
(goto-char (point-max))
(let ((b1 (point))
(b nil)
(e nil))
(insert (format "*** File %s follows ***\n" f))
(setq b (point))
(add-text-properties b1 (- b 1) (list 'face 'bold))
(insert-file-contents f)
(setq e (point-max))
(add-text-properties b e (list 'ub-file f))))
(defun ub-split-buffer-to-files ()
(interactive)
(let (b e)
(goto-char (point-min))
(unless (get-text-property (point) 'ub-file)
(setq b (next-single-property-change (point) 'ub-file)))
(while (not (eobp))
(when b
(setq e (or (next-single-property-change b 'ub-file) (point-max)))
(write-region b e (get-text-property b 'ub-file))
(goto-char e))
(setq b (next-single-property-change (point) 'ub-file)))))
Wasn't all that complicated. Here's how to use it: you switch to a
fresh buffer, then say M-x ub-append-file-to-buffer RET for each of
the files you want. Then you edit as you like, and when you're
finished, you say M-x ub-split-buffer-to-files RET and the changed
stuff will be written to the files.
The following problems still exist with this code:
* There is a bug that requires there to be at least one character
before the first file. Therefore, I changed the code to insert the
`*** File foo follows***' stuff -- this makes sure that there is
text before the first file.
I was too lazy to find this bug, but I think you might like the bold
stuff, anyway, since it tells you where are the boundaries of the
files.
* The code doesn't do anything special with MULE. So you might run
into problems if one of the files in the buffer uses a Latin-1
encoding, say, and another uses Latin-2, say, and still another uses
Shift-JIS (Japanese). If you need this, feel free to contact me.
* The user interface leaves something to be desired. Hm. Maybe there
should be a M-x find-multiple-files RET command and the splitting
should be done automatically when hitting C-x C-s. Oh, well.
As to slowness: I think editing speed is impaired only by the
slightest amount, and probably the effect on I/O is negligible.
What do you think?
kai
--
I like BOTH kinds of music.
------------------------------
From: [EMAIL PROTECTED] (Barry Samuels)
Subject: Re: Debian 2.1 upgrade failed !!!
Date: 4 Jul 2000 21:19:39 GMT
On Sun, 4 Jul 3900 04:47:05, [EMAIL PROTECTED] wrote:
Many thanks for your tip. I have now started the salvage
operation.
May I take this opportunity to congratulate you on your little
disc with the BIG heart!
Best wishes
Barry Samuels
my real e-mail address is:
barry (dot) samuels
@
btinternet (dot) com
> > I have tried Tomsrtbt floppy as that has both lilo and libc on
> > it. So after mounting my Debian partition on /mnt and entering
> > 'lilo -r /mnt ' lilo does run but displays the following:
> > 'first boot sector doesn't have a valid LILO signature'.
>
> (As noted in the FAQ,) use:
>
> chroot /mnt /sbin/lilo
>
> instead of lilo -r /mnt because you will get a version conflict between
> boot.b and the lilo executable...
>
> > barry (dot) samuels
> > @
> > btinternet (dot) com
> >
> >
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
------------------------------
From: [EMAIL PROTECTED] (David Efflandt)
Crossposted-To: alt.os.linux,comp.os.linux.setup,alt.linux
Subject: Re: Problems with fsck on new disk
Date: 4 Jul 2000 21:23:56 GMT
Reply-To: [EMAIL PROTECTED]
On Mon, 03 Jul 2000, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>I am having problems adding a new harddrive to my system. It was working
>fine in Windows. It is a Western Digital Caviar 34000
>specs:
>7752 Cylinders
>16 heads
>63 spt
>4000.7 megabytes
>
>I tried to fdisk it. It runs fine, but When I ran an fsck it complained
>about not being able to find the superblock. I tried iterations of the
>superblock backup. No luck. I finally ended up doing a mk2efs /dev/hdb
>
>Now I can run fsck on it. I fdisk deleted all partitions. I then ran
>fsck on it. If I run an "fsck -y /dev/hdb" it runs clean.
>
>If I run a "fsck -y -c /dev/hdb" it fails. I get a ton of errors.
You cannot fsck a partition until there is a filesystem on it. But then
instead of mke2fs /hdb1 (or other partition), you tried to write a
filesystem on the entire drive, deleted the partitions with fdisk, and
then tried to fsck an entire drive with no partitions on it. And you
wonder why fsck is confused?
You might want to grab your WD floppy, use the diagnostic program that can
wipe the drive to all zeros, and start over. Create some partitions with
fdisk, mke2fs each partition and then play with fsck on the partitions
(NOT the entire drive).
--
David Efflandt [EMAIL PROTECTED] http://www.de-srv.com/
http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
http://hammer.prohosting.com/~cgi-wiz/ http://cgi-help.virtualave.net/
------------------------------
From: Max TenEyck Woodbury <[EMAIL PROTECTED]>
Subject: Re: RAID syncronization problem
Date: Tue, 04 Jul 2000 17:43:28 -0400
Update--
> Jun 30 20:32:17 oscar kernel: SCSI disk error : host 1 channel 0 id 0 lun 0 return
>code = 18000002
> Jun 30 20:32:17 oscar kernel: Info fld=0x20920, Current sd08:21: sense key Aborted
>Command
> Jun 30 20:32:17 oscar kernel: Additional sense indicates Scsi parity error
^^^^^^^^^^^^^^^^^
On the assumption that this is a BUS error, not a disk error, I wrapped
the SCSI cable in bubble wrap. That seems to have made the problem worse,
not better. Just what does that error indicate?
Further, I added another swap tray and put the 7200 RPM drive in it
instead of replacing the 10K drive. It seems to have liked the changes
because it doesn't give errors anymore. It's about 1/4 synced at this
moment.
Again, is there a better forum for this kind of question?
[EMAIL PROTECTED]
------------------------------
From: "Kevin Vandersloot" <[EMAIL PROTECTED]>
Subject: Re: compiling gnumeric
Date: Tue, 04 Jul 2000 21:43:43 GMT
In article <[EMAIL PROTECTED]>, David
Grogan <[EMAIL PROTECTED]> wrote:
> I am trying to compile gnumeric 0.56 and am
> having trouble. I am running Mandrake 7.0 with
> a 2.2.16 kernel. In the gnumeric README, it
> says that I need
>
> gnome-libs-1.0.57 gnome-print-0.20
> gnome-xml-1.8.7 libglade-0.13
>
> So I went and found these rpms and installed
> them. I installed gnome-libs-1.0.58 because I
> couldnt find 1.0.57 and figured they would be
> backwards compatible. When I do the
> ./configure, I get this error
>
> checking for gnome-config... no checking for
> gnomeConf.sh file in /usr/local/lib... not found
> configure: error: Could not find the
> gnomeConf.sh file that is generated by
> gnome-libs install
>
> I can't find gnome-config or gnomeConf.sh
> anywhere on my system or my mandrake cd. Can
> anyone tell me why this isn't working and how I
> should go about fixing it? Thanks a lot.
>
> David
>
You need the devel packages if you want
to compile programs. So get the
gnome-libs-devel-1.0.58 etc.
------------------------------
From: Sylvain POURRE <[EMAIL PROTECTED]>
Subject: Re: inn, suck, nntpsend
Date: Tue, 04 Jul 2000 23:54:42 +0200
Hi Jim
I'll try to answer you but it's difficult because I'm french and I don't
use INN since I'm connected to Internet with a cable modem.
>
>. I have initiated a download of the last 150
> articles for each group we subscribe to, but now when the news gets > sucked
>down, it doesn't have a directory structure (e.g.
> spool/articles/comp/os/linux/misc) for the messages to go into and it
> complains.
> How do I go about recreating the structure? Do I have to manually create
> each directory or is there a command to do it for me.
If you've called suck with -c -n -H -K and -br options, all the news are
in a single file located in the /usr/lib/suck directory. You can open
this file for reading the news (cat, less, vi ...). The news separator
is #!rnews XXXX (where XXXX is the size in bytes). Now you must feed INN
with the sucked batch file : rnews batch_file_name. After that you must
be able to read the news with our news reader.
> Also if I want to subscribe to a new group from my newsfeed, is there an
> automated process or do I have to edit the sucknewsrc file manually and
> create the relevant directories?
If you call suck with the -hl and -A options it will read the INN active
file. You must type: suck ISP_name -c -n -A -hl localhost -br
batch_file_name.
> For posting articles the sysadmin used nntpsend,
Sorry I used rpost
>
> Thanks.
Do
I hope it helps
----
Sylvain
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: A confusing, but interesting topic...
Crossposted-To: alt.os.linux,nf.comp.linux
Date: Tue, 4 Jul 2000 20:16:01 +0100
In comp.os.linux.misc Hendrix <[EMAIL PROTECTED]> wrote:
> Hi there,
> How is linux installed from a distro? For instance, is the packages
> kept as an image, or is it compiled from source everytime the distro is
> installed on a new system...
On bootup, a REM disk is initialised and a mini-linux disk image is loaded
and uncompressed into the RAM disk. This then starts the main install
routines (such as YAST. YAST2 or whatever Red Hat and others use).
This routine then goes through all the partitioning, hardware detection and
installation. No recompilation is done. It's too processor intensive for one
thing (and remember, linux can just as easily be installed on a 486 as an
Athlon).
> I can see doing a new compile being a very good option, hence, the
> software can probe the new system in attempt to customize the kernel, and
> other programs, to work better on that partitcular machine...
Rather than recompiling the kernel (which can take hours), all modules tend
to be included with the distro.
--
______________________________________________________________________________
| [EMAIL PROTECTED] | |
|Andrew Halliwell BSc(hons)| "The day Microsoft makes something that doesn't |
| in | suck is probably the day they start making |
| Computer science | vacuum cleaners" - Ernst Jan Plugge |
==============================================================================
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Corel PhotoPaint
Date: Tue, 4 Jul 2000 20:18:12 +0100
David Turley <[EMAIL PROTECTED]> did eloquently scribble:
> Corel is full of shit when it claims this is a Linux app. It's just the
> windows version running under wine.
> Good enough reason to avoid all Corel apps.
Apart from wordperfect 8, of course.
--
______________________________________________________________________________
| [EMAIL PROTECTED] | |
|Andrew Halliwell BSc(hons)| "ARSE! GERLS!! DRINK! DRINK! DRINK!!!" |
| in | "THAT WOULD BE AN ECUMENICAL MATTER!...FECK!!!! |
| Computer Science | - Father Jack in "Father Ted" |
==============================================================================
------------------------------
From: [EMAIL PROTECTED] (William Wueppelmann)
Crossposted-To: alt.os.linux,comp.lang.c
Subject: Re: ANSI Colors in a C program under Linux...
Date: Tue, 04 Jul 2000 23:05:40 GMT
In our last episode (Tue, 04 Jul 2000 08:26:22 GMT),
the artist formerly known as Richard Bos said:
>Hendrix <[EMAIL PROTECTED]> wrote:
>
>> I think that a C program can be written with ANSI escape
>> characters in order to colorize a screen, and I believe that this would
>> be included in the ANSI Standard of the C language... It isn't graphics
>> functions, it just uses the ANSI escape sequences to print...
>
>Nope. ANSI wrote more standard than just the one, you know. The ANSI C
>standard is one; the ANSI escape characters are (I presume) from
>another, and have nothing to do with ANSI C (now by preference called
>ISO C, btw, since it was ISO that did the last standard, ANSI being one
>member of ISO).
>FWIW, my computer (an MS-DOS box) will not display ANSI escapes as
>intended, even though I compile ANSI C all the time.
More to the point, the C language itself doesn't have I/O functions, and
the C standard libraries make no assumptions about what kind of output
device (VTU, teletype, etc.) is used.
Curses will provide you with the ability to use colour, if the display
supports it.
--
It is pitch black.
You are likely to be spammed by a grue.
------------------------------
** 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
******************************
