Linux-Misc Digest #730, Volume #24 Tue, 6 Jun 00 16:13:11 EDT
Contents:
Re: Serious fragmentation under Linux ("Art S. Kagel")
atd job number ([EMAIL PROTECTED])
Re: DELL's Linux price is HIGHER than Win98 (John Hasler)
gnome modemlights ("Gerald Pollack")
FWD: Linux now support Ultra ATA/100 ("Alex Lam.")
Re: large disks - partitions problem with RH 6.1 (Christoph Kukulies)
Re: Sound with Inspiron 3800? ([EMAIL PROTECTED])
Backward compatible AGP video card? (Allen Ashley)
Re: large disks - partitions problem with RH 6.1 (Christoph Kukulies)
Re: Linux uses lots of memory? (Steve Lamb)
Re: Linux uses lots of memory? ("Robert L.")
Re: LILO in partition rather than MBR (John in SD)
Re: Linux uses lots of memory? ("Peter T. Breuer")
Re: Can a 486 handle PPPoE + ipMASQ for 3 comps ?
Re: RealPlayer 7: Cannot open the audio device... (Ties Verkuil)
Re: Linux uses lots of memory? (Martin Herrman)
gnome/gdm woes (Janet)
Install Error - kmod: failed to exec /sbin/modprobe... ([EMAIL PROTECTED])
Linux vs. FreeBSD (Melvin Toy)
Re: correction (was Re: forgot root password) (Leo Howell)
----------------------------------------------------------------------------
Date: Tue, 06 Jun 2000 14:01:41 -0400
From: "Art S. Kagel" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Serious fragmentation under Linux
MH wrote:
>
> Dances With Crows wrote:
> >
> > On Mon, 05 Jun 2000 21:27:59 -0700, MH
> > <<[EMAIL PROTECTED]>> shouted forth into the ether:
> > >I've more than a few posts regarding fragmentation under Linux. Most of
> > >the responses have been to the effect that "Linux doesn't have a
> > >fragmentation problem". I beg to differ.
> > >On a recent reboot, I noticed that I had 11.1%, 15.4%, and 19.8%
> > >"non-contiguous" files.
> >
> > As I'm sure N+1 others will point out, this isn't a problem unless you're
> > suffering horrible filesystem performance.
> >
> > Anyway, the "non-contiguous" report can be somewhat misleading. Linux
> > (and Unix in general) manages disk space differently from DOS. Linux
> > tries to keep all the blocks of a file near each other on the disk. When
> > a new file is created, at least 8 blocks are pre-allocated for it, even if
> > the file is only 1 byte in size. If you looked at the raw disk, you'd
> > probably see files spread out pretty evenly across the disk, with some
> > buffer space between files.
> >
> > Of course, files can get split up, but the ext2fs driver tries to keep
> > things relatively sane. So you might have 32K of file1, then 64K of
> > file2, then the next 32K of file1... etc. Files are kept within the same
> > "block group" if at all possible, where a block group is generally 8192
> > contiguous blocks on the disk. A block under Linux is at least 1K and
> > often 4K.
> >
> > In contrast, DOS filesystems use the first available block they find on
> > the disk. So if you have a DOS filesystem like so:
> > FILE1.TXT 8K -- FILE2.TXT 4K -- FILE3.TXT 16K
> > and you delete FILE2 and then create a 32K FILE4, you'd have:
> > FILE1.TXT 8K -- FILE4.TXT 4K -- FILE3.TXT 16K -- FILE4.TXT 28K
> > whereas under Linux, FILE4 would be placed right after FILE3, where
> > there's more free space. That way, FILE4 could be internally contiguous,
> > though there'd be a 4K section of free space between FILE1 and FILE3.
> >
> > I'm sure that people who are more familiar with ext2 internals could
> > explain this better, but the practical upshot is, "Don't worry about
> > non-contiguous files unless performance starts suffering!"
> >
> > --
> > Matt G / Dances With Crows \###| You have me mixed up with more
> > There is no Darkness in Eternity \##| creative ways of being stupid?
> > But only Light too dim for us to see \#| Beer is a vegetable. WinNT
> > (Unless, of course, you're working with NT)\| is the study of cool. --MegaHAL
>
> I'm not sure your explanation makes any sense, unless "non-contiguous"
> means something else in the Linux world than it does in the DOS world,
> or in the English-speaking world for that matter. I understand
> "non-contiguous" to mean bits of a single file located on blocks
> separated by other empty blocks or blocks containing bits of other
> files. Since files (blocks) are read sequentially, "non-contiguous"
> necessarily implies a degradation in performance since more blocks have
> to be traversed to read (or write) a given file.
There are several reasons why file fragmentation is less of a problem under
Linux ext2fs and other UNIX filesystems than it was under DOS (and windoze
IS better about this than DOS for similar reasons). Let me try to explain.
On a FAT partition each file is represented by a linked list of block
pointers in the FAT and a fragmented file can mean jumping around the FAT
table and around the disk quite a bit. Because native DOS does not cache,
smartdisk asside, and because DOS is single user, and so is Windoze, there
is a big impact on system performance if the drive heads have to be moved
about alot.
On the other hand Linux and other UNIX filesystems represent a file with
an inode which, unless the file is very large, contains pointers to
contiguous blocks of disk assigned to the file/inode. All of the blocks
pointed to by each inode entry, the 8K referred to, are contiguous. Now
any file larger than 8K will have multiple entries to other contiguous
groups of blocks and each entire group is read into the system cache in
one operation using read-ahead. Dances explains how Linux tries to keep
each group contiguous with the last group in a file or at least close.
This sometimes helps by reducing head movement, however, since Linux is
a multi-user operating system, head position after a read is mostly
irrelevant as other tasks and users will also be reading and writing that
drive and the buffer cache will be flushing older data out also moving the
drives' heads. This is why one does not notice as quickly any performance
impact from 'fragmented' files. The normal head movement that is part of
a multi-tasking multi-user system masks most performance impact and the
intelligent cacheing and inode design improve performance by normally
fetching the next block from disk to the cache before it is called for
so that applications do not notice any slowdown. Intelligent controllers
and disks (which mitigate for Windoze also somewhat), elevator sorting of
requests, out-of-order retrieval, etc all go to make fragmentation of all
but the worst kind irrelevant.
Art S. Kagel
------------------------------
From: [EMAIL PROTECTED]
Subject: atd job number
Date: Tue, 06 Jun 2000 18:00:01 GMT
hi!
is there a way to get the number of my atjob after invoking atd.
i have a lot of jobs in the queue but sometimes i have to delete one.
so i want to store the job number in a database.
thx
matthias
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: John Hasler <[EMAIL PROTECTED]>
Subject: Re: DELL's Linux price is HIGHER than Win98
Date: Tue, 6 Jun 2000 15:30:20 GMT
Gerald Willmann writes:
> I heard one prominent economist involved with the last MS proceedings
> hint at the possibility that they charge a huge fixed amount and then
> actually pay Dell et al for each copy of an MS OS that they install on a
> PC. Sounds incredible but ...
Sounds quite plausible (which is not to say that it is true). All
software, including Windows, is free: it costs Microsoft not one penny for
Dell to install one or a million copies of Windows. However, every copy of
Windows Dell installs creates one more potential customer for a copy of
Microsoft Office.
--
John Hasler
[EMAIL PROTECTED]
Dancing Horse Hill
Elmwood, Wisconsin
------------------------------
From: "Gerald Pollack" <[EMAIL PROTECTED]>
Subject: gnome modemlights
Date: Tue, 06 Jun 2000 18:24:43 GMT
I'm having trouble with the modemlights applet (gnome 1.2). When I click
on the large green button my ppp connection is established, but the
portion of the applet interface containing the button disappears (i.e. I
can't click it again to disconnect). The receive and send lights do not
become active, and it appears as if the applet has simply "died". gnome
1.2 / sawfish 0.27.2; mandrake 7.0.
------------------------------
From: "Alex Lam." <"here,there"@everywhere.com>
Crossposted-To: comp.os.linux.hardware
Subject: FWD: Linux now support Ultra ATA/100
Date: Tue, 06 Jun 2000 11:52:06 -0700
========================fwd==========================
Posted 06/06/2000 3:51pm by Annie Kermath
Linux beats Microsoft to support superfast disks
Linux support is now available for the Ultra ATA/100 bus
specification, Quantum announced today. Linux
is the only OS so far to support the new spec, and three of the
five current controller 1chipsets - Intel,
CMD and Promise - with AMD and HighPoint in the wings.
Full Story: http://theregister.co.uk/000606-000006.html
======================================================
Alex Lam.
------------------------------
From: Christoph Kukulies <[EMAIL PROTECTED]>
Subject: Re: large disks - partitions problem with RH 6.1
Date: 6 Jun 2000 19:00:12 GMT
Christoph Kukulies <[EMAIL PROTECTED]> wrote:
: I have problems installing RH 6.1 on a 40 GB IDE disk.
: I installed Win98 first and after trapping into the <1024 cylinders
: problem with bootable partitions have to be within I created
: a 2 GB / , 16 MB /boot , a 4 GB DOS and a 500 M swap partition.
: The rest of 30514 MB were made a /home partition.
: I did that all with the GNOME installer booted via a NFS install
: disk. (the boot disk was from the redhat site and contained already an
: upgrade which as it seems doesn't fix the problem)
Sorry, somehow a portion of my post got lost.
The problem is:
LILO got overwritten, when I installed something like OSBS208BETA
a multi boot loader.
Now I'm looking for a way to establish LILO just on the partition
rather than in the MBR.
Is that possible?
Or is the way always LILO->lilo.conf->multi-OS-boot?
I dont't want to use loadlin and I also want to
avoid booting one of the OSs to reach the other.
The other problem, which is virulent again after I did some minor
twaeking to the partition table, is that
the 6.1 installer upgrade still cannot cope with a 2GB DOS partition
type 6, a 4GB extended DOS and the rest empty on a 39GB IDE disk.
: --
: Chris Christoph P. U. Kukulies [EMAIL PROTECTED]
--
Chris Christoph P. U. Kukulies [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Sound with Inspiron 3800?
Date: Tue, 06 Jun 2000 18:53:48 GMT
Hi Gerard,
For configuring the sound on your Inspiron 3800, you might want to try
"sndconfig --noprobe --noautoconfig" as root and choose:
- the card type: ESS1688
- IRQ = 5
- DMA1 = 1
Let me know if it works.
The sound configuration files on my I3800 (created with "sndconfig
--noprobe" after initializing the sound card in DOS and using Loadlin to
boot) are:
- the file /etc/sysconfig/soundcard :
# THIS FILE IS WRITTEN BY SNDCONFIG
# PLEASE USE SNDCONFIG TO MODIFY
# TO CHANGE THIS FILE!
# There should be no spaces at the start of a line
# or around the '=' sign
CARDTYPE=ESS1688
- and the file /etc/conf.modules :
alias parport_lowlevel parport_pc
options mad16 io=0x530 irq=5 dma=1 dma16=0 joystick=1
post-install awe_wave /bin/sfxload /etc/midi/GU11-ROM.SF2
alias sound-slot-0 sb
options sound dmabuf=1
alias midi opl3
options opl3 io=0x388
options sb io=0x220 irq=5 dma=1 mpu_io=0x330
The file /etc/isapnp.conf does not exist on my system.
I hope that this helps.
Nick
In article <8hielk$981$[EMAIL PROTECTED]>,
"Gerard Milmeister" <[EMAIL PROTECTED]> wrote:
> In article <8hgs1m$e69$[EMAIL PROTECTED]>, [EMAIL PROTECTED] wrote:
> > Hi,
> >
> > I have an Inspiron3800, 700Mhz PIII, 256mb RAM, 12GB hard drive,
> > etc. I needed
> > to use both Win98 and Linux RH6.2 and I had the same problems,
> > but I fixed them :-)
> >
> > The solution was to remove LILO from MBR (with "fdisk /mbr" in
DOS
> > at the C:
> > prompt, read the Lilo-howto and Loadlin-Howto for more info)and
to
> > use Loadlin
>
> I have no Windows partition, and I don't intend to make one. So I have
to
> wait for another solution.
> Thanks anyway.
>
>
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (Allen Ashley)
Subject: Backward compatible AGP video card?
Date: 6 Jun 2000 18:57:25 GMT
I have found, to my dismay, that current ATI video cards are not backward
compatible with earlier cards. By that I mean that direct control commands
to the earlier video cards do not affect the current cards correctly. This
is important to me because I would like to utilize SVGATextConfig to
change the console display. In the progression for the ATI cards
from Mach32->Mach64->Rage128 it is not possible to take the chipset and
clocks specification from the Mach64, say, and apply it to the Rage128.
Is there a manufacturer, or chipset family, that is backward compatible?
If I buy the latest S3 chipset will it be possible to configure it as,
say, an S3 Virge? Achieving the highest video performance isn't important
to me.
Thanks for any response.
------------------------------
From: Christoph Kukulies <[EMAIL PROTECTED]>
Subject: Re: large disks - partitions problem with RH 6.1
Date: 6 Jun 2000 19:01:36 GMT
John in SD <[EMAIL PROTECTED]> wrote:
: On 5 Jun 2000 09:44:30 GMT, Christoph Kukulies
: <[EMAIL PROTECTED]> wrote:
:>
:>I have problems installing RH 6.1 on a 40 GB IDE disk.
: The 2.2.12 kernel supplied with RH6.1 does not support disks beyond 32Gb.
: Newer kernels do.
I thought the installer upgrade did. But it gets confused
by the partition table.
:>
:>I installed Win98 first and after trapping into the <1024 cylinders
:>problem with bootable partitions have to be within I created
:>a 2 GB / , 16 MB /boot , a 4 GB DOS and a 500 M swap partition.
:>The rest of 30514 MB were made a /home partition.
: As long as the 16Mb /boot partition lies below the 1024 cylinder limit, LILO
: v.21 should work fine. But the kernel may have problems with the /home
: partition in the upper part of the disk. Getting a newer kernel is the only
: solution.
: --John Coffman
:>
:>I did that all with the GNOME installer booted via a NFS install
:>disk. (the boot disk was from the redhat site and contained already an
:>upgrade which as it seems doesn't fix the problem)
: LILO version 21.4.3 (06-May-2000) source at
: ftp: sd.dynhost.com dir: /pub/linux/lilo
--
Chris Christoph P. U. Kukulies [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Steve Lamb)
Crossposted-To: comp.os.linux.setup,linux.help,linux.redhat.misc
Subject: Re: Linux uses lots of memory?
Reply-To: [EMAIL PROTECTED]
Date: Tue, 06 Jun 2000 19:05:22 GMT
On Tue, 6 Jun 2000 10:53:34 GMT, M. Buchenrieder <[EMAIL PROTECTED]> wrote:
>Run, yes. But certainly not with XFree86 running (at least, not in
>a useable state).
Pshaw. Be realistic, please.
>Don't run XFree86 with less than 64 MB at all.
Don't tell my IBM Thinkpad 360csa with a whopping 20Mb of RAM. It ran X
just fine. I found it quite usable.
>Once installed, however, the pure Linux system will run on 8MB - 16 MB
>just fine. It depends, of course, on what apps you're running and what
>services you do need.
I got the same laptop, a Debian machine, to boot in two and was quite
functional. All depends on your definition of "functional", doesn't it?
--
Steve C. Lamb | I'm your priest, I'm your shrink, I'm your
ICQ: 5107343 | main connection to the switchboard of souls.
===============================+=============================================
------------------------------
From: "Robert L." <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,linux.redhat.misc
Subject: Re: Linux uses lots of memory?
Date: Tue, 06 Jun 2000 19:05:54 GMT
"Mogens Kjaer" <[EMAIL PROTECTED]> a écrit dans le message news:
[EMAIL PROTECTED]
> Yes, netscape uses a lot of memory... But you can't blame the Linux
> kernel for this.
>
> I don't think anyone would run X and netscape on a 4Mb or 8Mb machine.
>
I have an 8Meg system, X run ( slow but run anyway for some game). If i
start netscape, it take ~45s. Changing webpage ( local ) ~10s.
I'll buy used parts latter today or tomorrow, so i'll have 16Meg instead.
------------------------------
From: John in SD <[EMAIL PROTECTED]>
Subject: Re: LILO in partition rather than MBR
Reply-To: [EMAIL PROTECTED]
Date: Tue, 06 Jun 2000 19:13:32 GMT
You have a couple of problems with this configuration.
1. RH6.1 kernel 2.2.12 cannot deal with disks larger than 32Gb. You need a
more recent kernel to fix this problem.
2. LILO v.21, which ships with RH6.1 cannot deal with cylinders above 1023.
You need a BIOS that supports EDD packet calls on int 0x13, and a newer boot
loader such asLILO 21.4.3 or Nuni. The former is available from my ftp site.
--John Coffman
On 6 Jun 2000 16:59:21 GMT, Christoph Kukulies
<[EMAIL PROTECTED]> wrote:
>I'm fighting to install Win98 concurrently with Redhat Linux 6.1 on the same
>(39GB) IDE disk.
>
>After the Redhat GNOME installer messed the partition table several times
>I now installed a plain DOS 2GB partition (to install Win98 later)
>having Linux on a partition after the 2GB (in reach of the 1024 sectors
>of BIOS)
>
>After installing Linux (had installed the DOS partition and DOS first)
>the Linux installation overwrote my MBR with Lilo, obviously.
>
>I then fdisk/mbr'ed the partition again and installed a boot manager
>which I know from older times, OSBS208BETA (or something like that).
>
>When I choose the first partition labeled 'linux' in the boot manager
>I cannot boot that partition. It hangs.
>
>Not knowing very much about the LILO architecture, is it possible to
>install LILO such that it sits only on the Linux partition and not in
>the MBR?
LILO version 21.4.3 (06-May-2000) source at
ftp: sd.dynhost.com dir: /pub/linux/lilo
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,linux.help,linux.redhat.misc
Subject: Re: Linux uses lots of memory?
Date: 6 Jun 2000 19:09:43 GMT
In comp.os.linux.misc Steve Lamb <[EMAIL PROTECTED]> wrote:
: On Tue, 6 Jun 2000 10:53:34 GMT, M. Buchenrieder <[EMAIL PROTECTED]> wrote:
:>Run, yes. But certainly not with XFree86 running (at least, not in
:>a useable state).
: Pshaw. Be realistic, please.
:>Don't run XFree86 with less than 64 MB at all.
: Don't tell my IBM Thinkpad 360csa with a whopping 20Mb of RAM. It ran X
: just fine. I found it quite usable.
Hey .. my 486sx50 is doing fine with X under 8MB of ram Right Now.
(if "fine" means "I can run 3 rxvt's and a console plus a fvwm file
manager, and still edit and telnet!" There are about 9 processes
running in this machine, I kid you not).
Peter
------------------------------
From: <[EMAIL PROTECTED]>
Subject: Re: Can a 486 handle PPPoE + ipMASQ for 3 comps ?
Date: Tue, 6 Jun 2000 15:10:26 -0400
So... my cpu is good enough for it.Roger that.
And I'm already scouting around for more memory . If I can live without
swapping , it would be great.
Neil Wolvin <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> I have a 486 dx33 520 mb hdd with 20 meg ram firewall for 4 computers, 3
> windows 1 linux with bellatlantic dsl. On occasion all 4 are connected to
the
> internet at the same time and I find no slowdown in performance. Before I
set
> this box up as a forewall I used it to experiment and I had 8meg or ram
and was
> running X. I found 8megs were not enough with X so I added some more
memory and
> performance improved a little. I later removed X and set up this box as
my
> firewall for a 56k dialup. When dls became available, I installed pppoe
and
> made the switch. I suspect you will be ok with 8megs of ram. If not, you
can
> pick up some additional memory.
>
> Good luck
> Neil Wolvin
>
> [EMAIL PROTECTED] wrote:
>
> > Hello,
> > I'm considering getting Bell Sympatico 's hse service. I've got 4
> > computers at present in a little lan , and they masq through the 486 ,
which
> > is currently using a 56K dialup.
> >
> > I am considering using Sympatico, but the pppoe has me wondering if my
old
> > 486 can handle it.
> > It's a 486 66mhz with 8MB ram and a 500 MB hdd. WOuld this be
sufficient to
> > masquerade 3 computers at the same time?
> >
> > Or would it be fine for atleast 2 computers to masq through at the same
time
> > . My searches indicate that the roaring penguin pppoe client is the most
> > recommended one out there, and I will be using it. Would any one know
what
> > if any ) would the throughput difference be between using the rp-pppoe
as a
> > user app and compiling it into the kernel ?
> >
> > Or is it that the processor is fine but the memory is insufficient ?
> > Services that are disabled :
> > samba ,xfs, nfs, named, apmd, pcmcia ( despite the warning in the
install
> > program, things are fine ), gpm, unnecessary getty's , arp, and a host
of
> > others that I'd not installed when I installed Linux .
> >
> > TIA
> > joseph
>
------------------------------
From: [EMAIL PROTECTED] (Ties Verkuil)
Subject: Re: RealPlayer 7: Cannot open the audio device...
Date: Tue, 06 Jun 2000 19:09:09 GMT
Hi,
I have changed the permissions for /dev/mixer and /dev/mixer1 as you
suggested.
I also changed the permissions for /dev/audio, /dev/audio1, /dev/dsp
and /dev/dsp1.
... I still get the same error message.
I am wondering I have only got mixer and mixer1, and you have got
mixer, mixer0, ...and mixer4, can this be the problem???
Greetings,
Ties Verkuil.
On Mon, 5 Jun 2000 06:44:42 +0700, <[EMAIL PROTECTED]> wrote:
>I did not used rp7.linux20.libc.i386.b2.rpm but load rp7 directly from
>internet.
>
>Please take a look at
>
>crw-r--r-- 1 root sys 14, 0 Jul 18 1994 mixer
>crw-rw-rw- 1 root root 14, 0 Jun 4 19:56 mixer0
>crw-rw-rw- 1 root sys 14, 16 Jul 18 1994 mixer1
>crw-rw-rw- 1 root root 14, 32 Jun 4 19:58 mixer2
>crw-rw-rw- 1 root root 14, 48 Jun 4 19:58 mixer3
>crw-rw-rw- 1 root root 14, 64 Jun 4 19:58 mixer4
>
>If so, then
>
>supat1:/dev# chmod a+w mix*
>supat1:/dev# su sss
>supat1:/dev$ realplay&
>
>Because realplay not only need /dev/audio permission but also /dev/mix
>
>Other sound tools may need only /dev/dsp permission.
>
>Regards,
>zxc
>
>
>=CA=D8=BE=D1=B5=C3=EC =BF=E9=D2=C3=D8=E8=A7=CA=D2=A7 (supat faarungsang) =
> Kasetsart Univ., Nakorn Pathom,
>mailto:[EMAIL PROTECTED] http://sss.agri.kps.ku.ac.th/
>office: (034)351892 home: (034)351843 fax: (02)5791120
>mobile: (01)4127930 lab:(034)281053-6 ext 3434
>voicechat: http://supat.dhs.org/ttt.html or http://go.to/supat
>=A4=C7=D2=C1=B4=D5=A4=B9=E0=C3=D2=B9=D5=E8=B4=D5=E3=B4 =B4=D5=B9=D3=E9=E3=
>=A8 =B7=D5=E8=E3=CB=E9=E1=A1=E8=A4=B9=B7=D1=E9=A7=BB=C7=A7 =
> :)
>
>On Sun, 4 Jun 2000, Ties Verkuil wrote:
>
>> Hi All,
>>=20
>> I have installed rp7.linux20.libc.i386.b2.rpm on Red Hat 6.0 and 6.1.
>> When I start RealPlayer, I get in both causes the error message:
>> Cannot open the audio device. Another application may be using it.
>> My soundcard is a Sound Blaster PCI 64 (ES 1371).
>> Which is working fine with XMMS etc.
>>=20
>> Does anyone knows a solution for this problem?
>>=20
>> Thanks in advance,
>>=20
>> Ties Verkuil.
>>=20
>> E-mail adress: Replace home with hotmail.
>>=20
>>=20
>
------------------------------
From: [EMAIL PROTECTED] (Martin Herrman)
Crossposted-To: comp.os.linux.setup,linux.redhat.misc
Subject: Re: Linux uses lots of memory?
Reply-To: [EMAIL PROTECTED]
Date: 06 Jun 2000 19:21:12 GMT
On Tue, 06 Jun 2000 19:05:54 GMT, Robert L. <[EMAIL PROTECTED]> wrote:
>
> I have an 8Meg system, X run ( slow but run anyway for some game). If i
> start netscape, it take ~45s. Changing webpage ( local ) ~10s.
> I'll buy used parts latter today or tomorrow, so i'll have 16Meg instead.
I have 233mhz, 32 mb ram, running Xfree 86 3.3.6.. I am using
the windowmanager icewm and it is really fast!!! So.. no problems
with less than 64 megs.. bull shit.. (ehm.. but restart netscape
regularly when having not that much memory, but I guess that also
is true with a 128 mb machine ;-)
Martin
P.S. someone with some optimize methods? I disabled 3 of my 6 consoles
in the inittab file, am using a small windowmanager (icewm, black box
would do as well) and i'm using pine and slrn as clients.. netscape
as browser (sometimes lynx) and licq.. that will run fine.. but
i'm an optime freak you know :-)
--
Linux Gebruikers Handleiding v1.2 : http://2mypage.cjb.net
Linux RedHat 6.1 Kernel 2.2.14 Toshiba P233 MHz, 32 Mb RAM
9:10pm up 5 days, 1:45, 3 users, load average: 0.00, 0.06, 0.07
Western Civilization, that would be a good idea!
------------------------------
From: Janet <[EMAIL PROTECTED]>
Subject: gnome/gdm woes
Date: 06 Jun 2000 12:36:19 -0700
Hi,
I recently installed the Helix version of Gnome 1.2. I am having an awful
time trying to get everything to play nicely. Basically, some users would
like to use gnome, and some would like to use non-gnome sessions. So, I
would prefer to have everyone just type startx, but when I put
gnome-session in the .xinitrc file and try to startx, the computer has a
tendency to freeze (pingable but can't login remotely). So then I thought
I would just use gdm, but when I try to start an Xsession from gdm, it
hangs as well. This time I can ssh in and kill gdm, so I do. But then
when I try to restart it, the screen flashes and I can see on alternate
flashes the console, where it's saying stuff about freeing AGP pages. It
seems to be starting a new gdm process every time it flashes because the
pid of gdm keeps changing, but I manage to kill that. And then when I try
to do a normal startx into fvwm, I get a signal 11 error. Help!
Janet
------------------------------
From: [EMAIL PROTECTED]
Subject: Install Error - kmod: failed to exec /sbin/modprobe...
Date: Tue, 06 Jun 2000 19:26:14 GMT
I'm trying to install SuSE on a Motorola Compact
PCI PowerPC MCP750 single board computer.
I'm getting the following "kmod" error as my boot
image executes (only applicable lines listed
here):
Linux/PPC load:
initrd_start located > 16M
Uncompressing Linux...done.
Now booting the kernel
PReP architecture
...
Boot arguments:
...
Starting kswapd v 1.5
Serial driver version 4.27 with MANY_PORTS
MULTIPORT SHARE_IRQ enabled
ttyS00 at 0x03f8 (irq=4) is a 16550A
ttyS01 at 0x02f8 (irq=3) is a 16550A
...
scsi : 0 hosts.
scsi : detected total.
...
RAMDISK: Compressed image found at block 0
EXT2-fs warning: checktime reached, running e2fsck
is recommended
VFS: Mounted root (ext2 filesystem)
kmod: failed to exec /sbin/modprobe -s -k
char-major-4, errno = 2
Any ideas? A previous post mentioned that
"char-major-4" refers to the ttySx devices. I'm
using an HP 700/96 terminal connected to the COM1
port. I've tried using "console=ttyS0,9600n8",
"root=/dev/sda2", "root=/dev/sda1", and many other
variations of these as boot arguments, but I still
get this error.
My CDROM & Hard disk are both SCSI, but the
install says 0 scsi hosts were detected - is this
a problem?
Also, how can I determine if my hard disk is sda1,
sda2, etc? I don't have an OS installed yet
(that's what I'm working on)...just the PPC-Bug
Motorola debugger. Anyone out there PPC-Bug
experts?
Thanks for all your help!
-Dave
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Melvin Toy <[EMAIL PROTECTED]>
Subject: Linux vs. FreeBSD
Date: Tue, 06 Jun 2000 19:48:58 GMT
This may bring up an old argument.
I been playing w/ linux for over a year and like it for a server very
much. I've been starting hearing more and more about FreeBSD. First
from a FreeBSD hard-core who tells me if you have a web server running
FreeBSD & another running Linux you won't see a performance difference
w/ a thousand hits a day. However if you get a million hits a day,
FreeBSD leaves Linux behind in the dust. He also sees Linux in a few
years becoming bloated like NT because stuff is being slapped on Linux
left and right w/o any control. He says, "FreeBSD is slow to change but
we're very stable." I asked one of our Linux instructor about this. He
didn't know too much about FreeBSD but it didn't surprise him because he
knows FreeBSD is more or less bulletproof and Linux still has security
problems as a server. He was surprise FreeBSD didn't get more
publicity. I had another instructor pointed out though that most
development is being done on Linux. Then I read this big article over
the Memorial Day weekend on FreeBSD and all these companies like Yahoo,
Hotmail,etc running all their servers using FreeBSD.
I'm about to set up 12 web servers w/ RH linux but now I'm not sure.
I know this is an old argument but does it still hold today? Comments?
Melvin
------------------------------
From: Leo Howell <[EMAIL PROTECTED]>
Subject: Re: correction (was Re: forgot root password)
Date: Tue, 06 Jun 2000 20:47:14 +0100
Minko Markov wrote:
> That was a stupid thing to say, sorry. Of course,
> passwd will ask for the old password. So, you have
No, it won't. Not for root, anyway. That's how root can change user's
passwords without knowing what they are.
------------------------------
** 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
******************************
