Linux-Misc Digest #375, Volume #25                Mon, 7 Aug 00 18:13:02 EDT

Contents:
  Re: UDMA IDE Drive stops network transfers (Stewart Honsberger)
  Re: Size of /var/lib/rpm - why so big? (Stewart Honsberger)
  Re: mp3 normalizer for Linux? (Kevin E Cosgrove)
  Re: Xdefaults (Fester)
  Re: psychology of linux ("Philo")
  Re: FWD: Red Hat's CFO abandoning ship. (Robert Krawitz)
  Re: FWD: Red Hat's CFO abandoning ship. (Robert Krawitz)
  3c59x using IRQ 0, how to change? ("Jason")
  Web BBS with stories & Video (Scorpius)
  Re: psychology of linux (Stephen Hui)
  Re: 3c59x using IRQ 0, how to change? (mst)
  Putting 2GB into several CD's (Julio Gonzalez-At)
  GUI Software??? Please Help ([EMAIL PROTECTED])
  Re: FWD: Red Hat's CFO abandoning ship. (Johan Kullstam)
  Re: Help burning CDR (iso) image (Patrick Lamb)

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

From: [EMAIL PROTECTED] (Stewart Honsberger)
Crossposted-To: comp.os.linux.hardware
Subject: Re: UDMA IDE Drive stops network transfers
Reply-To: [EMAIL PROTECTED]
Date: Mon, 07 Aug 2000 20:52:32 GMT

On 7 Aug 2000 19:04:07 GMT, Dances With Crows wrote:

After using the -u1 flag my drive performance seems to have improved,
especially in the buffer-cache reads. They now look as follows;

>>I just tried it on my drives, both UDMA 33, and got the following results;
>>/dev/hda:
>> Timing buffer-cache reads:   128 MB in  2.69 seconds = 47.58 MB/sec
>> Timing buffered disk reads:  64 MB in  4.73 seconds = 13.53 MB/sec

blackdeath:~ # hdparm -tT /dev/hda

/dev/hda:
 Timing buffer-cache reads:   128 MB in  2.22 seconds = 57.66 MB/sec
 Timing buffered disk reads:  64 MB in  4.32 seconds = 14.81 MB/sec

>>/dev/hdb:
>> Timing buffer-cache reads:   128 MB in  2.55 seconds = 50.20 MB/sec
>> Timing buffered disk reads:  64 MB in  7.78 seconds =  8.23 MB/sec

blackdeath:~ # hdparm -tT /dev/hdb

/dev/hdb:
 Timing buffer-cache reads:   128 MB in  2.22 seconds = 57.66 MB/sec
 Timing buffered disk reads:  64 MB in  7.60 seconds =  8.42 MB/sec

I've also enabled Write cacheing on both drives. That, from what I
understand, can cause data-loss after a power failure, but I'm going to
test it for a while and see if performance improves.

The above 57.66 number seems a bit interesting to me - both drives DMA
buffers read at the same speed? Is that normal?

The info on my drives looks like such;

blackdeath:~ # hdparm -i /dev/hda

 Model=Maxtor 90648D3, FwRev=GAS74812, SerialNo=A36200NC
 Config={ Fixed }
 RawCHS=12556/16/63, TrkSize=0, SectSize=0, ECCbytes=29
 BuffType=3(DualPortCache), BuffSize=512kB, MaxMultSect=16, MultSect=16
 DblWordIO=no, OldPIO=2, DMA=yes, OldDMA=2
 CurCHS=12556/16/63, CurSects=12656448, LBA=yes, LBAsects=12656448
 tDMA={min:120,rec:120}, DMA modes: mword0 mword1 mword2
 IORDY=on/off, tPIO={min:120,w/IORDY:120}, PIO modes: mode3 mode4
 UDMA modes: mode0 mode1 *mode2
 Drive Supports : ATA/ATAPI-4 T13 1153D revision 17 : ATA-1 ATA-2 ATA-3 ATA-4

blackdeath:~ # hdparm -i /dev/hdb

 Model=FUJITSU MPA3043ATU, FwRev=9519, SerialNo=01168939
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
 RawCHS=9042/15/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=0(?), BuffSize=0kB, MaxMultSect=16, MultSect=16
 DblWordIO=no, OldPIO=2, DMA=yes, OldDMA=0
 CurCHS=9042/15/63, CurSects=8544690, LBA=yes, LBAsects=8544940
 tDMA={min:120,rec:120}, DMA modes: mword0 mword1 mword2
 IORDY=yes, tPIO={min:240,w/IORDY:120}, PIO modes: mode3 mode4
 UDMA modes: mode0 mode1 *mode2
 Drive Supports : ATA-1 ATA-2 ATA-3

>VIA MVP3 chipset w/VIA UDMA support enabled in kernel 2.2.16, hdparm
>-u1 -c1 -m16 -d1 applied to both drives.

I'm now using all of those settings. It appears as if my Sector Count was
already set to the MaxMultSect value for both drives by default.

>The figures are even better with the 2.4.x test kernels,

My kernel version is in my .signature. I must say I'm quite impressed with
the performance of these drives, if those figures are correct. I'm going
to reccomend some of these settings to a friend running a UDMA66 drive and
see what numbers he can come up with.

>but overall performance is still worse with 2.4 thanks to the ongoing
>VM problems.

Oh? I've noticed snappier performance since switching to 2.4. Memory
management appears to be much better, and disk cacheing is greatly
improved.

"Ongoing VM problems"? Would you mind elaborating?

>My early '97 laptop gets 3M/sec even with everything tweaked for hdparm,
>so 13M sounds great to me....

I dunno, that's just what hdparm tells me. Since applying some of the early
tweaks to my HDD's, moving 3-6MB MP3 files between drives/partitions has
GREATLY improved. I'll have to try moving a 300MB VMWare .dsk file around
and see how fast MC tells me it's moving.

-- 
Stewart Honsberger (AKA Blackdeath) @ http://tinys.cx/blackdeath
[EMAIL PROTECTED]  (Remove 'thirteen' to reply privately)
Humming along under SuSE 6.4, Linux 2.4.0-test5

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

From: [EMAIL PROTECTED] (Stewart Honsberger)
Crossposted-To: alt.os.linux.mandrake
Subject: Re: Size of /var/lib/rpm - why so big?
Reply-To: [EMAIL PROTECTED]
Date: Mon, 07 Aug 2000 20:56:51 GMT

On Mon, 07 Aug 2000 20:30:30 GMT, Vlar Schreidlocke wrote:
>But how do I resize (make larger) my /var partition without trashing
>everything in it?

Copy everything off of it, being careful to read the man page of cp to
copy all file permissions, to a temporary holding directory. (I used
/var_temp, a directory created within the / partition). Umount the
existing /var partition and kill it with fdisk. Create a new /var partition
of the desired size (this may require deleting/resizing/moving other
partitions around, which can be done with PowerQuest's Partition Magic;
I just don't trust it completely with ext2 partitions).

Once you've created the new partition, create a filesystem on it (man mke2fs),
mount it, and copy the files back (again being certain to maintain file
permissions) and you're ready to roll.

I modified my /var partition as mentioned above on a running Linux system
(name, proxy, web, FTP, SMTP, POP3, NTPD, etc..) without having to reboot.

Good luck - and be sure to read carefully all man pages along the way!
Losing your /var partition can be hazardous to your health. ;>

-- 
Stewart Honsberger (AKA Blackdeath) @ http://tinys.cx/blackdeath
[EMAIL PROTECTED]  (Remove 'thirteen' to reply privately)
Humming along under SuSE 6.4, Linux 2.4.0-test5

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

From: [EMAIL PROTECTED] (Kevin E Cosgrove)
Subject: Re: mp3 normalizer for Linux?
Date: Mon, 07 Aug 2000 20:59:44 GMT

The script below could let one adjust the peak levels of the
files to all be at maximum.  But, peak level is *NOT* a good
indicator of listener "volume".  Rather, RMS amplitude is a
better indicator of listener volume.  Recently RMS amplitude was
added to sox's "stat" feature.  I forget the first version number
to include this new feature, but it's definitely in 12.17.
Here's a sample of the new stat output.

    Samples read:          27378806
    Length (seconds):    620.834603
    Scaled by:         2147483647.0
    Maximum amplitude:     0.946472
    Minimum amplitude:    -0.912659
    Mean    norm:          0.062141
    Mean    amplitude:     0.000117
    RMS     amplitude:     0.091823
    Maximum delta:         0.806030
    Minimum delta:         0.000000
    Mean    delta:         0.041814
    RMS     delta:         0.061465
    Rough   frequency:         4698
    Volume adjustment:        1.057

The script below should be coded so that the file with the
largest "Maximum amplitude" is increased in volume by its
"Volume adjustment".  After that, the rest of the files should
be increased in volume so that their "RMS amplitude"s match that
of the first files new "RMS amplitude".  This will be a good
starting point for equalizing the perceived volumes.

G'luck...


In article <[EMAIL PROTECTED]>,
 [EMAIL PROTECTED] (Mike Castle) writes:
> 
> How about a simple shell script?
> Hmmm... something like:
> 
> #!/bin/sh
> for v in $*; do
>       mp3towav $v $v.wav
>       adj=`sox $v.wav stat -v 2>&1`
>       sox $v.wav -v ${adj} -t wav - | wavtomp3 - $v.tmp
>       mv $v.tmp $v
> done


-- 
Unless otherwise noted, the statements herein reflect my personal
opinions and not those of any organization with which I may be affiliated.

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

From: [EMAIL PROTECTED] (Fester)
Subject: Re: Xdefaults
Reply-To: [EMAIL PROTECTED]
Date: Mon, 07 Aug 2000 21:03:23 GMT

On Mon, 07 Aug 2000 19:03:39 -1000, Ron Nicholls <[EMAIL PROTECTED]> wrote:
>Although I have configured gnome from the defaults, I don't have an
>Xdefaults file in my home directory. I do have one in root.
>
>So how can I have modified my home desktop and not produced an Xdefaults
>file.

Gnome (1.2, at least) uses three subdirectories:

Program Settings: ~/.gnome/
Desktop Icons:    ~/.gnome-desktop/
Stored Passwords: ~/.gnome_private/

(and, of course, that last one is chmod 700)

-- 
-- Fester

 "Anyone can create 'smart' sentences 
  by using 'one', 'thus', or 'hence'." 
  -Tomo 
 
 "Thus, one is inclined to laugh at 
  anything Tomo says hence." -Dexter
======================================



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

From: "Philo" <[EMAIL PROTECTED]>
Subject: Re: psychology of linux
Date: Mon, 7 Aug 2000 16:00:10 -0500

even tho it was 46 years ago...i recall vividly the day i learned to ride my
bike.
it was the same day i tried riding one handed on gravel!

learning linux may be more difficult than riding a bike on gravel...
but at least you won't come home with a bloody face :)

Philo

ps: both tasks will give you a lot od satisfaction once mastered.



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

From: Robert Krawitz <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,gnu.misc.discuss
Subject: Re: FWD: Red Hat's CFO abandoning ship.
Date: 07 Aug 2000 17:17:24 -0400

blowfish <[EMAIL PROTECTED]> writes:

> Robert Krawitz wrote:

> > > Take Boeing as an example.  If they go the free software route, they can
> > > save millions in software costs, but they decided against it, because
> > > it's not practical for such a big international corp to switch
> > > everything, tens of thousands of employees in numourous countries, and,
> > > the trainning costs and time loss will far outcosted the cost saving in
> > > free software.
> > 
> > That's quite a different issue.
> 
> That's part of the total costs too. Isn't it!?

Of course it is, but that has nothing to do with free vs. proprietary
software; it's a matter of frictional losses in retraining on anything.
-- 
Robert Krawitz <[EMAIL PROTECTED]>      http://www.tiac.net/users/rlk/

Tall Clubs International  --  http://www.tall.org/ or 1-888-IM-TALL-2
Member of the League for Programming Freedom -- mail [EMAIL PROTECTED]
Project lead for The Gimp Print --  http://gimp-print.sourceforge.net

"Linux doesn't dictate how I work, I dictate how Linux works."
--Eric Crampton

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

From: Robert Krawitz <[EMAIL PROTECTED]>
Crossposted-To: gnu.misc.discuss
Subject: Re: FWD: Red Hat's CFO abandoning ship.
Date: 07 Aug 2000 17:19:20 -0400

blowfish <[EMAIL PROTECTED]> writes:

> Robert Krawitz wrote:
> > 
> > blowfish <[EMAIL PROTECTED]> writes:
> > 
> > > Also. You did contradicted your own idea of *free software*, as you have
> > > taken the money from Debian as a reward.
> > >
> > > Isn't that your codes should be *FREE* as in money - free, unlike
> > > non-GNU-GPL codes-which costs money, as in *FREE BEER* too!?
> > 
> > No, "free" as in "free software" only refers to "free speech"
> > (liberty), not "free beer".  It's perfectly legal to sell GPL'ed
> > software for whatever the market will bear.  The only thing the seller
> > has to do is provide the source code, and not restrict further
> > distribution beyond what the GPL specifies.
> > 
> No. Speech to most *real* human means expressing ideas, communicating to
> other *real* humans.

Fine.  Let's skip "free speech", then, and say that "free software"
refers to "liberty of code".  "Free speech" is merely an analogy, anyway.

-- 
Robert Krawitz <[EMAIL PROTECTED]>      http://www.tiac.net/users/rlk/

Tall Clubs International  --  http://www.tall.org/ or 1-888-IM-TALL-2
Member of the League for Programming Freedom -- mail [EMAIL PROTECTED]
Project lead for The Gimp Print --  http://gimp-print.sourceforge.net

"Linux doesn't dictate how I work, I dictate how Linux works."
--Eric Crampton

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

From: "Jason" <[EMAIL PROTECTED]>
Subject: 3c59x using IRQ 0, how to change?
Date: Mon, 07 Aug 2000 21:12:06 GMT

I can't get my LInux  6.2 to get on the network. Can't ping anything and
can't ping the box from another pc.
Looking at the /var/log/messages, when it loads the Cyclone driver for my
3com Fast Etherlink XL PCI, it says "might not work on IRQ=0" and it does
try IRQ 0. How do I change it to IRQ 11 (which is what I think it should
be)?

Probably a easy programming answer.

Thanks,

Jason



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

From: Scorpius <[EMAIL PROTECTED]>
Crossposted-To: alt.bbs.unixbbs,alt.bbs.internet,alt.bbs
Subject: Web BBS with stories & Video
Date: Mon, 07 Aug 2000 14:20:25 -0700

http://www.story.com
“ACCEPTANCE “
This months NEW story.
Come Visit The WSB Internet Café.
Short Stories, Poetry & Web Video...


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

From: Stephen Hui <[EMAIL PROTECTED]>
Subject: Re: psychology of linux
Date: Mon, 07 Aug 2000 16:28:04 -0500

Right now, as a student about to graduate from university, I like to
keep a competitive edge over my classmates.  I learn cutting-edge,
high-demand languages (especially related the Web development world,
**except** for ASP...), and I learn all about networks and system
administration (particularly UNIX/Linux).  I'm even thinking about
working through a Linux certification course.

I could forsee that my competitive-edge-itch will probably become simple
curiosity in the future (i.e. I wonder if I can do this, or I wonder how
I can to do that) because I'll be learning everything I need to know for
my job on the job.

The other aspect of it (which will never fade) is sheer exploratory
science--diving into an unknown world and discovering things as if they
have never been discovered before.  Linux was **COMPLETELY** foreign to
me when I dove in head-first four years ago.  Since then, I have amassed
more knowledge than I probably care to know, a lot of which has come out
of resources like this newsgroup and other online resources.  As I
continue to use UNIX/Linux regularly I'm discovering more and more about
it.

And yet another aspect of it is genuine interest.  I find reading about
networking rather fascinating.  Just the fact that a team of people
invented this thing called the Internet about 30 years ago boggles my
mind.  And to see where the technologies have gone since they were
invented is even more fascinating, and then imagining where they might
go....

In a nutshell, to me at least, having this kind of motivation is an
"acquired taste".  Most of my friends don't/can't understand how I can
be interested in all the things I'm interested in.  Once a person gets a
taste for something they can be really passionate about, it really
starts to make sense.  They just have trouble extending it to
computers.  :o)

Stephen.



shawn wrote:
> 
> greetings earthlings
> 
> Ok, this is an obscure and hard to answer question, but i want to know
> what motivates the most hardcore geeks among us. what makes you go out
> and learn that new language, what makes you interested in learning new
> systems, networking, that type of thing. I really have trouble
> grasping/understanding a concept without knowing all the guts behind the
> curtain so to speak, so if you could give me any relevant information it
> would be appreciated.
> 
> shawn

-- 
Stephen Hui, ARL:UT, Austin, Texas

Computer Terms: Programmer - A red-eyed, mumbling mammal
capable of conversing with inanimate objects.

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

From: mst <[EMAIL PROTECTED]>
Subject: Re: 3c59x using IRQ 0, how to change?
Date: Mon, 07 Aug 2000 17:31:49 -0400

Jason wrote:
> 
> I can't get my LInux  6.2 to get on the network. Can't ping anything and
> can't ping the box from another pc.
> Looking at the /var/log/messages, when it loads the Cyclone driver for my
> 3com Fast Etherlink XL PCI, it says "might not work on IRQ=0" and it does
> try IRQ 0. How do I change it to IRQ 11 (which is what I think it should
> be)?
> 

Try to pass it as an option when you load the module.

MST

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

From: Julio Gonzalez-At <[EMAIL PROTECTED]>
Crossposted-To: dc.org.linux-users
Subject: Putting 2GB into several CD's
Date: Mon, 07 Aug 2000 17:36:47 -0400
Reply-To: [EMAIL PROTECTED]

Hi everyone,
I thought the following problem would be already solved.
But now I'm sort of loosing hope.

I want to backup 2GB data into several CD's. Ideally
ther would be a way for me to

1) Compress the files before writing
2) Split the resulting strucutre into 650MB

I tried using the -L option for the gnu tar
to no avail. This option seems to make sense
only when writing to tapes.

compressing would be easy, I just need to duplicate
the data before writing, but I don't know
how to split the data into chunks in order to
use mkisofs. Maybe I'll need to write my
own script ?

thanks all of you

Julio

--
"In the world, there are only matters which have not been understood,
there is nothing which cannot be understood."
                    Qigong Phenomena booklet




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

From: [EMAIL PROTECTED]
Subject: GUI Software??? Please Help
Date: Mon, 07 Aug 2000 21:26:37 GMT

Hello, does anyone know of a gui software installer for linux, or a
good web site that explains how to install linux software?

Also, are there any utilities that can do a GUI removal or uninstall of
junk files and programs?  Hopefully, there is something like Norton's
uninstaller, cache cleaner, and temp file cleaner.  Could also be like
*soft's disk cleaner that is included in Win 98.

Any help would be appreciated.



Sent via Deja.com http://www.deja.com/
Before you buy.

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

Crossposted-To: comp.os.linux.setup,gnu.misc.discuss
Subject: Re: FWD: Red Hat's CFO abandoning ship.
From: Johan Kullstam <[EMAIL PROTECTED]>
Date: Mon, 07 Aug 2000 21:52:23 GMT

blowfish <[EMAIL PROTECTED]> writes:

> Christopher Browne wrote:
> > 
> > Centuries ago, Nostradamus foresaw a time when blowfish would say:
> > >Christopher Browne wrote:
> > >> Centuries ago, Nostradamus foresaw a time when blowfish would say:
> > >> >Johan Kullstam wrote:
> > >> >> if you can't own it, you can't be stealing it right?
> > >> >>
> > >> >Here goes your twisted logic again...
> > >> >
> > >> >Okay. A just robbed a bank. Then you go and robbed A, took the money
> > >> >that A robbed from the bank.
> > >> >
> > >> >You didn't robbed the bank directly, but you're in process of the bank's
> > >> >money by robbing A.
> > >> >
> > >> >Does that makes you a lesser criminal??????????????
> > >>
> > >> Money is a construct for which ownership is pretty intrinsic.
> > >>
> > >> Its _essence_ is as an expression of owned value.
> > >>
> > >> Thus, any argument surrounding the notion of things that _cannot be
> > >> owned_ cannot be applied to money, at least not without taking _great_
> > >> care to form syllogisms to indicate the lack of ownership.
> > >
> > >How/Why money cannot be owned?
> > >
> > >If you've created something, anything, and people likes it, they pay you
> > >money for it, or for a copy of it. Then, you've *EARNED* that money, and
> > >that money belongs to you. You have every legal title over that money.
> > >
> > >Or if your parents have earned the money, and passed it down to you,
> > >then, you have inherented the money bevause of your parents. Even if you
> > >didn't work for it yourself.
> > >
> > >That money is still yours. And you're the legal owner of that money.
> > 
> > I don't disagree with all that; the notion of being "owned" is pretty
> > intrinsic to money.
> > 
> > The point is that if the original premise concerned something that
> > _couldn't_ be owned, it is problematic to try to build analogies using
> > money, which _is_ owned.
> > 
> The *only* things that cannot be owned are the things that you are not
> entilted to.
> 
> Let me ask you this:

no, let me ask you this:

what is a strawman?

let's see you stick to the subject instead of proffering an endless
stream of strawmen, hmm?

-- 
J o h a n  K u l l s t a m
[[EMAIL PROTECTED]]
Don't Fear the Penguin!

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

From: Patrick Lamb <[EMAIL PROTECTED]>
Subject: Re: Help burning CDR (iso) image
Date: Mon, 07 Aug 2000 17:03:56 -0500

Dances With Crows wrote:
> 
> On Mon, 07 Aug 2000 08:21:02 -0500, pdlamb wrote:
> >I need some help burning an image to a CD.  I haven't been able to burn
> >it with cdrecord, using a command line like 'cdrecord speed=4 dev=0,0,0
> >-v pinstripe-powertools-i386.iso' -- makes the best coaster you've ever
> >seen!  I've had my best luck mastering and burning images with xcdroast,
> >but I can't figure out how to specify an image file.
> 
> How did you make the .iso file in the first place?  And can you check
> the .iso using the loopback, like so?
>   mount -t iso9660 file.iso /mnt/other -o loop,ro
> should mount the .iso on /mnt/other --if it coughs up error messages
> unrelated to the lack of loopback support in your kernel (Put it in!) or
> the nonexistence of /mnt/other (mkdir it), then the .iso file is more
> than likely corrupt.  To make an .iso file out of a data CD, just do:
>   dd if=/dev/cdrom of=file.iso bs=2k

The file was a download from the Redhat site (powertools 7 beta).  I can
read it fine as loopback.

> xcdroast is a frontend to cdrecord and mkisofs--nothing more, nothing
> less.  What works in xcdroast *must* work in cdrecord.  Does cdrecord
> cough up any errors while burning?

Sounds good, but unfortunately all the documentation is code - there is
no documentation, help files, etc.  Even the README is dedicated to
describing compilation and installation.  No errors during the cdrecord
run.  So, clue me in: how do you specify an image file to be used with
xcdroast?  Or does xcdroast feed a command line to cdrecord that is
substantially different from the one I listed above?

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


** 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
******************************

Reply via email to