Linux-Misc Digest #582, Volume #26 Mon, 18 Dec 00 18:13:02 EST
Contents:
Re: Can any1 tell me why I can't print postscript ? (Emmanuel Beranger)
Network kits that work w/ Linux? (Andrew Purugganan)
Re: How is a GNU/Linux modem user supposed to do his offline www browsing? ("Dan
Jacobson")
Re: Do Linux ext2 partition need defrag? (Jean-David Beyer)
Memorex CDRW ATIP/calibration problem? (Oleg Krivosheev)
Re: Do DMA mode relate to SCSI HDD? (Oleg Krivosheev)
Re: Does this cd-rom burner work w/ Linux? (Grant Edwards)
Re: blindly accepted net advice--> blew away filesystem (Jean-David Beyer)
Re: Linux disk blocksize (Manfred Bartz)
Need MINIMAL Linux for a laptop dinosaur... (D'Arque Bishop)
xchat variable docs (Rick)
Ethernet & Redhat 5.0 (Mike Fontenot)
Re: Intel D815EEA sound/audio & Linux problem. ("machine22")
----------------------------------------------------------------------------
From: Emmanuel Beranger <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux.mandrake
Subject: Re: Can any1 tell me why I can't print postscript ?
Date: Mon, 18 Dec 2000 22:19:24 +0100
"Peter T. Breuer" a écrit :
>
> In comp.os.linux.misc Bill Unruh <[EMAIL PROTECTED]> wrote:
> > In <91il8n$280v$[EMAIL PROTECTED]> "Emmanuel Beranger"
><[EMAIL PROTECTED]> writes:
>
> >>While gs is installed, and I can print ASCII
> >>Am I missing something ?
> > How can anyone know? You have given close to zero information here. --
> > What OS, how you set up printing, etc.
>
> A naive person would suspect that he doesn't know how, or his printer
> doesn't know how [to print PS].
>
> ;-)
>
> Peter
All right, allright ...
mandrake 7.0 with 2.2.17 kernel
Using lpr, configured with printtool. My printed is an Epson stylus 820.
I know it works postscript, since I have already used it in the past
(before I screwed and reinstalled Linux) ...
Thing is : file never reach pool when I launch a print command from a
non ASCII origin (like, say "print" in Netscape ...) I suspect the
filter does not read okay, but i don't know which log file to look at
...
Also, my printcap has the filter that comes with Mandrake ...
Any further thoughts ?
------------------------------
From: [EMAIL PROTECTED] (Andrew Purugganan)
Subject: Network kits that work w/ Linux?
Date: 18 Dec 2000 21:13:15 GMT
I am no network guru and would like the simplest kit to install in my
home. I have 2 Win98 SE boxes and one Linux PC (Mandrake 6.0, kernel
2.2.13mdk). In fact, the second SE actually has Corel Linux 2.1, and if
were a wiz at this...
Man if the network runs over the home phone lines, that would be best, but
so far, all I've seen is everything works with Windows-only setups, even
that network-in-a-box dealie from Intel. Ho-hum.
--
jazz
Registered linux user no. 164098 +--+--+--+ Litestep user no. 386
Doesn't it bother you, that we have to search for intelligent life
--- OUT THERE??
------------------------------
From: "Dan Jacobson" <[EMAIL PROTECTED]>
Crossposted-To:
alt.comp.linux.isp,alt.netscape,comp.infosystems.www.browsers.misc,gnu.misc.discuss
Subject: Re: How is a GNU/Linux modem user supposed to do his offline www browsing?
Date: Tue, 19 Dec 2000 05:19:35 +0800
"The Real Bev" <[EMAIL PROTECTED]> ????? news:[EMAIL PROTECTED]...
> "Peter T. Breuer" wrote:
> >
> > In comp.os.linux.misc Dan Jacobson <[EMAIL PROTECTED]> wrote:
> > > How is a GNU/Linux modem user supposed to do his
> > > offline www browsing? [Don't tell me to use somethng that I can't see
Chinese
> >
> > What is the attraction in this? I don't get it. If you go to a mound of
> > pages that you don't read, you have to spend the time to download them
> > anyway! So what's the advantage in not reading them?
>
> See below.
Me Mr. modem user's strategy with Taiwan telecom charging by the minute:
Write on paper all the things to find on google.
Connect, look on those search results pages, click some references, make sure
each page fully loads into the browser... going up and down trees of webpages in
a flurry.
Go to local newspapers, find interesting stories click on them and make sure
they fully load... at the same time this is happening I'm 'synchronizing my
netnews reader...
Ok, maybe 30 minutes late I'm through and then I spend the next 6 hours
analyzing all the junk I hurriedly snarfed into the browser's cache.
As each day's search goals differ, I don't see to much chance to write a script
to do it without me finding I overlooked some info that I would have spotted has
I been on-line.
Ok, I am able to do this currently on M$ with OE & IE. I have great hopes for
GNUS on Linux and can already see Chinese in it. However I must have a decent
browser to use so I don't double my phone bill by downloading news on Linux and
browsing on win98...
Sure I could theoretically use gnus on win98, but I thought I was supposed to
move to Linux because that's where the in crowd is, and though using
gnus-agentize mode
to download onto the vfat disk and then move to Linux to read.... never mind
so I went to mozilla's site and it seems they haven't made any fresh ChineseBig5
updates lately, etc.
> > If you are trying to save the (minuscule) time it takes to read a page
then I'm a slow reader
> > of info, just set Netscape to read once per session. Or are you saying
huh, what's that? I don't see such an option in 4.75 for Linux...
What I'm saying is that IE has this nice ordered history tree list,
indeed, you select the off line mode, and then easily browse today's previous
snarfings.
Apparently, I never expected not to find an equivalent feature in Netscape. I
mean weren't it a participant in the 'browser wars'? How could it participate
with this missing arm and leg, unless its audience is already 24 hours connected
and wouldn't cherish such a feature.
Ok, I'm glad there's competition. One should get one's browser up to par before
hawking one's operating system eh?
> > that even then netscape checks the net first? It shouldn't. If it does,
> > report it as a bug.
>
> Yeah, like anybody cares! 4.75 checks anyway, especially from the BACK
> button. Moreover, it ignores maximum cache size and occasionally destroys
> the history.dat file, which is -- to somebody's eternal shame --
> unsearchable unless you write it to a file and then search that, which is a
> nuisance.
>
> It would be really handy to go back and look at the stuff that's stored in
> the 19 meg and growing cache file without being on line, although
> cablemodem makes it less essential. Just being able to FIND stuff there
> would be a good thing. I hate having a mess of saved files hanging around
> just because I think I might want to look at them again, especially since
> they're already saved in the cache.
--
www.geocities.com/jidanni E-mail: restore ".com." ???
Tel:+886-4-5854780; starting in year 2001: +886-4-25854780
------------------------------
From: Jean-David Beyer <[EMAIL PROTECTED]>
Subject: Re: Do Linux ext2 partition need defrag?
Date: Mon, 18 Dec 2000 17:07:23 -0500
[EMAIL PROTECTED] wrote:
>
> > > As title, do I need to defrag it to maintain performance just like
> > > Windows platform?
>
> > Fragmentation is not always a detrement to performance. In certain
> > classes of system (of which Linux is one), fragmentation isn't the
> > performance impact that it is in MSWindows (sometimes it can even
> > enhance performance).
>
> I disagree. Fragmentation is not that important as long as all the data
> can be accessed through the filesystem cache. That's because there won't
> be any harddisk seeking at all when all data is already in RAM. But as
> far as I examined running Linux systems, this is rarely the case.
> Fragmentation considerations only apply when talking about actual disk
> accesses.
Does anyone know the difference between the memory used for BUFFERS and
the memory used for CACHE?
>
> BTW: due to a bug in the 2.2.x cache management (at least I suspect it
> to be a bug), by just reading in a number of directories, the file
> system cache may be emptied almost entirely. All subsequently requested
> data has to be read in again. You should be able to watch this via 'top'
> or 'free' when your cache is well filled by just performing a 'du -kcs
> /', for example.
>
> And don't forget the ATime-attribute of Inodes which is usually set and
> causes every read operation in a directory (maybe from cache) to be
> followed by a write operation (definitely to the disk). In this case
> even lazy write buffering doesn't help, when the particular Inodes are
> scattered wildly across the disk (Inode groups). Defragmenting and
> trying to put all contents of a directory close together, does.
>
> Especially with respect to the ATime feature of Linux I would even state
> that the fragmentation impact in Windows is less than in Linux! Windows
> usually uses to access more files in a more frequent manner than Linux
> does. Additionally, the cache management of NT4 is a kind of broken,
> which reduces harddisk access savings also. Even on a defragmented file
> system. So I think that Linux would benefit more from a properly
> defragmented filesystem.
Have you ever measured it? Have you run a reasonable benchmark (a
heavily IO-dependent program that runs for a long time compared to
getting it into memory at startup) both before and after defragmentation
of the file system where its data reside? You should be able to
defragment easily enough by copying the entire file system to a backup
file system, tape, or whatever, unmounting the file system, doing a mkfs
on the partition, and then restoring the file system from the backup.
>
> Besides, I wonder what you mean by "sometimes [fragmentation] can even
> enhance performance"? Probably not that of your running system but that
> of your harddisk manufacturer's sales? ;-)
>
> Sent via Deja.com
> http://www.deja.com/
--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 73926.
/( )\ Shrewsbury, New Jersey
^^-^^ 5:00pm up 14 days, 1:47, 3 users, load average: 3.78, 3.65, 3.46
------------------------------
From: Oleg Krivosheev <[EMAIL PROTECTED]>
Subject: Memorex CDRW ATIP/calibration problem?
Date: 18 Dec 2000 16:11:12 -0600
Hi, All
have 2*2*24 memorex CDRW (?which is JVC W2080 ?)
in dual boot linux/win98 computer.
looks like under the linux
cdrecord is unable to read ATIP information,
get media info and calibrate the laser.
i feel it leads to less reliable writing to
phtalocyanine media i use.
if i reboot to win98, then both CD id program
and cloneCD were able to get media info
and calibrate laser.
Linux is debian unstable with cdrecord 1.9,
kernel 2.2.18
same blank CD from the same spindle were
used for tests
any ideas/info would be greatly appreciated...
regards
OK
------------------------------
From: Oleg Krivosheev <[EMAIL PROTECTED]>
Subject: Re: Do DMA mode relate to SCSI HDD?
Date: 18 Dec 2000 16:13:52 -0600
[EMAIL PROTECTED] (Dances With Crows) writes:
> On Tue, 05 Dec 2000 11:10:12 +0800, Carfield Yim staggered into the
> Black Sun and said:
> >I know that if I have IDE HDD, I can use hdparm -d 1 /dev/hda to
> >increase the speed of my HDD, but is it relate to scsi HDD?
>
> No. When power is applied to the SCSI device and its bus, the device
> and the card negotiate to find out the highest speed both are capable of
> handling, and they use that speed. You can speed up your SCSI devices
> by:
> buying a better SCSI card (PCI > ISA)
> upgrading the SCSI bus (Ultra > Wide > Narrow)
> buying faster SCSI devices (10000 RPM > 7200 RPM)
> hacking the SCSI drivers in an intelligent way (good luck!)
umm...
related question - how to set DMA for
ATAPI (==SCSI over IDE) CDRW which used
under ide-scsi driver ?
rgds
OK
------------------------------
From: [EMAIL PROTECTED] (Grant Edwards)
Subject: Re: Does this cd-rom burner work w/ Linux?
Date: Mon, 18 Dec 2000 22:16:13 GMT
In article <91eifj$s7$[EMAIL PROTECTED]>, Shawn Smith wrote:
>I am just wondering if anyone has this working w/ RH 7. I am just
>checking before I buy one.
>
>Hewlett-Packard 12x8x32 SCSI Internal CD-RW Drive 9600i
I've got a 9200e, and it works fine w/ RH6.2.
I've never found a SCSI CDROM that didn't work with Linux --
though those old Sun branded ones with the goofy block size
were pretty slow.
--
Grant Edwards grante Yow! ... I see TOILET
at SEATS...
visi.com
------------------------------
From: Jean-David Beyer <[EMAIL PROTECTED]>
Subject: Re: blindly accepted net advice--> blew away filesystem
Date: Mon, 18 Dec 2000 17:16:34 -0500
"(Wm. Randolph Franklin)" wrote:
>
> \begin{oldfartmode}
>
> This isn't new to Linux. On the old IBM OS/360 with punchcards,
> you could delete one file (called a dataset) by having the
> following input card to the IEHPROGM program, IIRC:
>
> SCRATCH VTOC=volname,DSNAME=filename comments go here
>
> Alternatively, you could delete everything this:
>
> SCRATCH VTOC=volname comments go here
>
> Note that a space separated the command from the comments. Guess
> what would happen if you changed one comma to a space in my first
> example above:
>
> scratch vtoc=volname dsname=filename comments go here
>
> Also, at this time, there was no security; any user could do
> anything to the disk.
I do not remember that there was no security. When a friend of mine
accidently deleted the Fortran H compiler and all that went with it (by
putting OLD,DELETE where it should not have been), the throughput of our
system went way up, or seemed to. They then set the permissions so that
only a SYSADMIN could do anything but read and execute the compilers,
link-editors, libraries, and loaders. This was probably in 1965. So
there was enough security to prevent things like that from happening,
once the sys-admins learned how to apply it.
> This survived into the early 1970s because
> the users were either ignorant or benevolent.
As I recall, there was almost as much ignorance then as now. But there
was no malevolence then. People, at least in the computing profession,
acted in a more civilized and professional way then than most do now.
Perhaps because they were more highly educated, but also because they
had higher ethical standards. People would not normally try to crack a
machine (other than to find out things about the software). The
Internet, once it started to get used, had little security implemented
in it, not because they did not know how, but because they could not
afford the overhead when all the nodes on the network where Honeywell
DDP-516's. (If you have never used one of these, be glad.) They omitted
the security because it was really not needed in those days. It was a
community of cooperating people who would no more harm another as they
would harm themselves, because they were all in it together.
We are still all in it together, but the barbarians do not appreciate
it.
>
> IBM's first step towards security was to delete this command from
> IEHPROGM. However, they left the corresponding system subroutine
> call available for awhile longer. The system call documentation
> was cleverly hidden in the publicly available SYSTEM CALLS manual.
>
> ----
> (Wm. Randolph Franklin) <[EMAIL PROTECTED]>
--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 73926.
/( )\ Shrewsbury, New Jersey
^^-^^ 5:10pm up 14 days, 1:57, 3 users, load average: 3.22, 3.45, 3.48
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.setup,comp.os.linux.networking
Subject: Re: Linux disk blocksize
From: Manfred Bartz <[EMAIL PROTECTED]>
Date: Mon, 18 Dec 2000 22:25:26 GMT
[EMAIL PROTECTED] writes:
> Linux someservername 2.2.14-VA.5.1smp #1 SMP
> Tue Sep 12 21:26:21 PDT 2000 i686 unknown
>
> On Solaris, I can find the blocksize used on the disks with:
> df -g
>
> What is the command to show the same thing in Linux?
dumpe2fs -h /dev/hdxx
--
Manfred
===============================================================
ipchainsLogAnalyzer, NetCalc, whois at: <http://logi.cc/linuc/>
------------------------------
From: [EMAIL PROTECTED] (D'Arque Bishop)
Crossposted-To: alt.os.linux
Subject: Need MINIMAL Linux for a laptop dinosaur...
Date: Mon, 18 Dec 2000 22:18:32 GMT
Hey there,
I've got a bit of a problem here, and I was hoping someone out there might
be able to help me. I've got an NEC Versa V/50 laptop w/ a 486/SX CPU, 4
MB of RAM, a 500 MB hard drive, and a 14.4k PCMCIA modem. What I'm wanting
to do is put a very minimal Linux on this laptop. Basically, ALL it would
be required to do is dial up a remote PC using minicom. In fact, if it
wasn't for the fact that the boot/root disks for Slackware require a minimum
of 8 MB of RAM, I'd just use the a1 disk series of Slack. Unfortunately,
the earliest distro available on their website is 3.3. :( Does anyone have
any recommendations for a distro and config of Linux that could be used to
make this laptop into a simple dial-up terminal?
Thanks in advance...
--
==============================================================================
"Do you see the smile in my words, sad and evil? Sad because
I am utterly alone. Evil because I am dead and yet I live.
Can you hear me? Listen. A dead man visits you."
--James O'Barr, The Crow
D'Arque Bishop -- [EMAIL PROTECTED]
http://www.ravenloft.net/~drkbish
"For a dark man shall come unto the House of God, and the
darkness shall be upon him, yea, even within him."
-- from Noctropolis: Night Vision
==============================================================================
------------------------------
From: Rick <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.powerpc,alt.os.linux.mandrake
Subject: xchat variable docs
Date: Mon, 18 Dec 2000 17:38:19 -0500
I am trying to write user commands, popups and the like, but I cant find
any docs for the variables. Does anyone know where I can find these?
Any and all help appreciated.
--
Rick
* To email me remove theobvious from my address *
------------------------------
From: Mike Fontenot <[EMAIL PROTECTED]>
Subject: Ethernet & Redhat 5.0
Date: Mon, 18 Dec 2000 22:40:23 GMT
I haven't been able to get my redhat 5.0 intel system to
use my ethernet card. I didn't have an ethernet card
when I originally installed rh5.0 several years ago.
Recently, I added SuSE 6.4, so that I could boot my
computer using either distribution. SuSE uses the
ethernet card fine, using the tulip driver. After booting
redhat, there is nothing about eth0 in /var/log/messages or
in the output from dmesg.
Under rh5.0, the conf.modules file was empty. This
indicates (I think) that, if there IS currently any
ethernet driver, it must be built into the kernel
that rh5.0 supplied...it isn't a module. An "lsmod"
shows no modules loaded. Redhat 5.0 does have a tulip.o
under /lib/modules/2.0.32/net.
I added the line "alias eth0 tulip" to the conf.modules
file, and tried doing a "modprobe tulip". The response
to the command is:
"/lib/modules/2.0.32/net/tulip.o: init_module:
Device or resource busy."
Does anyone know what that response means?
I conjectured that maybe it means that there is already a
driver for servicing eth0 built into the kernel, and which
can't handle my card. So I changed the conf.modules file
to say "alias eth1 tulip", and did another modprobe, but
I got the same response as before. I've also tried booting
the redhat system with the LILO command:
"LILO: redhat ether=9,0xfc00,eth0"
hoping that, if there IS a driver built into the kernel,
maybe I could get it to work with my card if I gave it the
irq and i/o address. That didn't seem to make any difference.
Any advice much appreciated.
Mike Fontenot
[EMAIL PROTECTED]
------------------------------
From: "machine22" <[EMAIL PROTECTED]>
Crossposted-To: intel.motherboards.pentium_ii,comp.os.linux.hardware
Subject: Re: Intel D815EEA sound/audio & Linux problem.
Date: Tue, 19 Dec 2000 00:44:30 +0200
And i can't play using XMMS
With freeamp all is ok
<[EMAIL PROTECTED]> wrote in message news:8tmvrj$1ue$[EMAIL PROTECTED]...
> Intel D815EEA sound/audio & Linux problem.
>
> I have Intel D815EEA motherboard, which has integrated sound/audio,
> video, & LAN.
> I'm running RedHat Linux 7.0, which has kernel 2.2.16.
> When I try to play an MP3 file using XMMS, it seems to play, but I
> don't hear anything.
> When I switch the OS to Win98SE (I have dual boot), I'm able to play
> and hear the music without problems. When I switch back to Linux, XMMS
> seems to play the MP3 file, but I don't hear anything out of the
> speakers.
> I tried sndconfig from the root account.
> It seems to identify the sound card as,
> Model: Intel Corporation|82801AB AC'97 Audio
> It tries to play a sample sound as a test, but I don't hear anything.
> Then it presents me with a list of soundcards.
> I choose Intel i810 AC97 Audio, because it's already selected, and it
> seems like the closest model. It tests the sound driver again and
> fails.
> I contacted Intel tech support, but they couldn' help me because they
> don't officially support Linux.
> Any help would be appreciated.
>
> ---
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
------------------------------
** 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 by posting to comp.os.linux.misc.
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
******************************