Linux-Misc Digest #278, Volume #21 Tue, 3 Aug 99 19:13:12 EDT
Contents:
Re: GDP (Richard Kulisz)
Re: IDE vs scsi? (Rod Smith)
Does Linux support the TR-4 Tape drive? ("Matt McDevitt")
Re: diagramming tool (Bob Tennent)
Installing ftpsearch on Linux? (Steen Suder)
printer problem (Ricardo Carvalho)
chmod: 04000 and 02000 in directories? (Otavio Exel)
Re: IDE vs scsi? (Johan Kullstam)
make bin (Artem Ponomarev)
Re: VMWare (Windows on Linux)? (Matthew Bafford)
Re: Installing Netscape 4.61 ("loatherier")
Re: Massive video flickering with Mandrake 6.0 (Leonard Evens)
Re: Can't log in to Red Hat system. (Leonard Evens)
Re: bash question: changing path within script? (Jan Just Keijser)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Richard Kulisz)
Crossposted-To: comp.os.linux.advocacy,gnu.misc.discuss
Subject: Re: GDP
Date: 3 Aug 1999 21:17:43 GMT
In article <[EMAIL PROTECTED]>,
Donovan Rebbechi <[EMAIL PROTECTED]> wrote:
>On 1 Aug 1999 22:37:11 GMT, Richard Kulisz wrote:
>[ snip: claim: GDP is relatively useless ]
>
>Your comments are interesting. I'd agree that GDP is not extremely
>useful for measuring living standards. However, it's not completely
>useless.
The GDP is not useful to measure the *economy*.
>>GDP measures only how much people slave away,
>
>IOW, trends in the GDP correlate with trends in unemployment.
IOW, if your economy is based on slavery then the GDP is useful.
But the GDP doesn't tell you anything more than that people are
busily building empty boxes and then shipping them into the sea.
>Leisure, health and the environment are very hard to measure.
This is no excuse. Assigning them a value of *ZERO* is vastly
more inaccurate than any half-assed estimate the lowest of idiots
can make. Besides, the facts are that they aren't hard enough to
measure that many people haven't measured them by now (to come up
with the GPI for example).
>This is why an isolated "number" will never measure progress
>on these issues.
>
>The mistake that the conservatives ( or at least the "economic
>rationalists" ) seem guilty of is a tendency to simply ignore
>these things and concentrate on getting "good numbers".
Yes. And what "numbers" are you referring to if not the GDP? It's
simpler to just say the GDP is useless.
------------------------------
From: [EMAIL PROTECTED] (Rod Smith)
Subject: Re: IDE vs scsi?
Date: Tue, 03 Aug 1999 20:54:46 GMT
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>,
"Art S. Kagel" <[EMAIL PROTECTED]> writes:
> Rod Smith wrote:
>>
>> [Posted and mailed]
>>
>> In article <[EMAIL PROTECTED]>,
>> "Art S. Kagel" <[EMAIL PROTECTED]> writes:
>> >
>> > So the point is still made, many of the best SCSI drives out there
>> > claim 40-80MB/sec sustained throughput which is, alone, sufficient to
>> > saturate a SCSI bus
>>
>> Again, this isn't true! 40-80Mbps (and higher), yes, but not 40-80MB/s.
>
> OK, actual specs. Disk to cache rate is the drives sustainable transfer
> rating:
>
> Drive Reported internal transfer rate disk->cache
> --------------------- --------------------------------------------
> Quantum Fireball Plus KA 235 Mb/Sec. or 29.4 MB/sec.
> Seagate Barracuda 50 264 Mb/Sec. or 33.0 MB/sec.
> Quantum Atlas 10K 315 Mb/Sec. or 39.4 MB/sec.
> Seagate Cheeta 18LP 308 Mb/Sec. or 38.5 MB/sec.
>
> These are some 30 and 40MB/sec class drives. Apparently my memory was
> not correct about having seen specs on 80MB/sec drives. Again the
> original point is not lost.
1) Unless I'm mistaken, those speeds are all under 40MB/s -- some not
by much, I admit, but under, nonetheless.
2) The speeds you've quoted appear to be for the MAXIMUM internal transfer
rates. For instance, the fastest drive there, the Quantum Atlas 10K,
actually claims rates of 215-315Mb/s (26.9-39.4MB/s).
3) These drives are available in high-end SCSI busses that far exceed
40MB/s capacity -- the Atlas 10K comes in (among others, according to
the spec sheet) Ultra160 (though Ultra160 is the only interface listed
at buycomp.com as being actually available), which does 160MB/s, so
even at something approaching 40MB/s maximum, the drive only chews 1/4
the bus's capacity.
Since your "original point" was that internal transfer rates exceed the
attached SCSI bus speeds, IMHO you *HAVE* lost that point. Now, if you
want to mix and match interfaces, and say that, for instance, current
drives exceed the capacity of the original SCSI-1, or even more recent
interfaces, then fine, I'm quite willing to concede that point (and I
never claimed otherwise anyhow). Most SCSI drives you buy today won't
chew more than half their bus's capacity, though some might, particularly
high-end drives (like the Atlas 10K) attached to (relatively)
low-performance busses.
> Rod you said:
> I also doubt the 40MB/s figure somebody else quoted. Those
> are both limits for particular implementations of SCSI busses.
> Real-world hard disk transfer speeds are almost always
> substantially lower than the hard disk bus speed.
>
> These drives are transferring data into their cache at almost EXACTLY
> their bus speeds and can transfer from cache to bus at substantially
> higher rates for burst transfers!
I wasn't aware of any 40MB/s designs when I wrote that. I now am (or at
least, rates very close to 40MB/s), but they're still *VERY* high-end
drives, and if you're going to be buying one of those for $767.95 (the
current price at buycomp.com), you can afford a high-end SCSI interface
for it, too.
Note also that I said I *DOUBTED* that figure. I never stated that
40MB/s internal transfer definitely rates didn't exist. I did deny your
claim of 120MB/s rates and of somebody else's claim of 80MB/s rates.
> I do apologize for overstating the case. Above are the actual figures
> the case still stands.
As I say, IMHO it doesn't. Unless you're attaching a very high-performance
drive to a host adapter that's distinctly beneath it, the SCSI bus itself
isn't a limiting factor in the drive's speed.
--
Rod Smith
[EMAIL PROTECTED]
http://www.channel1.com/users/rodsmith
NOTE: Remove the "uce" word from my address to mail me
Author of _Special Edition Using Corel WordPerfect 8 for Linux_, from Que
------------------------------
From: "Matt McDevitt" <[EMAIL PROTECTED]>
Crossposted-To:
alt.linux,alt.os.linux,alt.os.linux.redhat,comp.os.linux.hardware,comp.os.linux.help,comp.os.linux.setup,linux.redhat.misc
Subject: Does Linux support the TR-4 Tape drive?
Date: Tue, 3 Aug 1999 11:36:50 -0500
I'm using RedHat 6.0 -
I can't seem to get the TR-4 tape drive to be recognized.
Is there a driver out there for it?
Matt McDevitt
[EMAIL PROTECTED]
http://www.maktos.com
------------------------------
From: [EMAIL PROTECTED] (Bob Tennent)
Subject: Re: diagramming tool
Date: 3 Aug 1999 21:05:28 GMT
Reply-To: rdt(a)cs.queensu.ca
On 03 Aug 1999 13:08:53 -0700, Lee Doolan wrote:
>
>
>is anyone aware of an open/free graphical diagramming
>tool (a la visio or vthought) for linux??
>
Try dia:
Summary : Dia is an advanced program for making technical diagrams
Description :
Dia is a program for creating diagrams of all kinds. The current
version can do UML class diagrams, Entity-Relationship modeling and
network diagrams. The engine is very flexible and dynamically loads
diagram-types from disk. It can currently export to postscript and
load/save in an xml format.
Bob T.
------------------------------
From: Steen Suder <[EMAIL PROTECTED]>
Subject: Installing ftpsearch on Linux?
Date: Tue, 03 Aug 1999 20:29:04 +0000
I'd like to set up an ftpsearch server
(http://ftpsearch.lycos.com/search-info/software.html) running on Linux.
I'd like to hear from someone who has already been there and perhaps
succeeded.
I know it's in the FreeBSD ports collection.
Can it be done?
--
Best regards / Mvh.,
Steen Suder
sysadm kollegie6400.dk
GNU - makes me feel better! Ehhh, Linux is GNU, right...?
------------------------------
Date: Tue, 03 Aug 1999 22:18:07 +0100
From: Ricardo Carvalho <[EMAIL PROTECTED]>
Subject: printer problem
hello
does anyone knows why my HP Laser jet IIIP prints 2 thirds in the first
page a the rest in another page
the layout its ok it simple uses two pages
changing the print driver doent work
i prints ok in windows
TiA
Ricardo Carvalho
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Otavio Exel)
Subject: chmod: 04000 and 02000 in directories?
Date: 3 Aug 1999 21:14:42 GMT
Reply-To: [EMAIL PROTECTED]
hello all,
looks like a FAQ but I could not find it: what does the 04000 and 02000
permission bits mean? all I read (including 'man 2 chmod') is that..
S_ISUID 04000 set user ID on execution
S_ISGID 02000 set group ID on execution
..but I assume this applies to executable files, not directories;
TIA
--
Otavio Exel . /<\oo/>\ . [EMAIL PROTECTED]
------------------------------
Subject: Re: IDE vs scsi?
From: Johan Kullstam <[EMAIL PROTECTED]>
Date: 03 Aug 1999 18:00:39 -0400
"Art S. Kagel" <[EMAIL PROTECTED]> writes:
> Johan Kullstam wrote:
> >
> > "Art S. Kagel" <[EMAIL PROTECTED]> writes:
> >
> > > On a single user workstation? Rarely. On an Internet Server with a Database
> > > server providing data? Absolutely you can tell the difference.
> >
> > > Certainly for a single user system SCSI only makes sense if you want
> > > to take advantage of the high speed peripherals and external
> > > connections that SCSI will provide over EIDE. There are no EIDE
> > > scanners or external drive cabinets, and SCSI tape drives still give
> > > better throughput. But is that enough to warrant the extra expense,
> > > even in the US where the penalty is more like 40% than double as it
> > > seems to be in Europe? You have to decide. I went SCSI, most do
> > > not for a workstation. BTW my company uses SCSI in all of the WinNT
> > > workstations here, and I have nothing to do with that, FWIW.
> >
> > nothing says you cannot have an EIDE hard disk drive with a cheap
> > narrow/slow SCSI for the external devices like tape and scanner.
>
> Nothing except giving up another slot and interrupt and perhaps even more
> cables.
yes, you need a hard drive, it'll want a cable of some sort whether
ide or scsi. if you have an external device like tape drive or
scanner it will also need cabling. good luck getting a non-scsi
scanner to work without using a card or interrupt. i am not sure if
the tape needs a card and interrupt or not.
the choice comes down to eide hard disk and scsi periferals or scsi
all over. unless scsi is built in to your mobo, you will need a card
slot. not doing ide frees an interrupt.
given the cost of uw and u2w scsi hard disks and pci adapters, one
could still get a cheap eida hard drive and isa scsi adapter, save
money and have a reasonable system. sure, all scsi is nicer. no one
is disputing that. we're not all montebanks either.
--
J o h a n K u l l s t a m
[[EMAIL PROTECTED]]
Don't Fear the Penguin!
------------------------------
From: Artem Ponomarev <[EMAIL PROTECTED]>
Subject: make bin
Date: Tue, 03 Aug 1999 14:42:54 -0700
==============490FA78EC001C19C61382520
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
My /usr/bin/make bin does not work, it reports "Segmentation Fault". for
some
reason I could not install amy "make" utilities from the sun.unc site.
.Where
to get "make" bin for Linux 2.0.34?
Thanks, Artem.
--
AtB, Artem
==============490FA78EC001C19C61382520
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<HTML>
My /usr/bin/make bin does not work, it reports "Segmentation Fault". for
some
<BR>reason I could not install amy "make" utilities from the sun.unc site.
.Where
<BR>to get "make" bin for Linux 2.0.34?
<BR>Thanks, Artem.
<PRE>--
AtB, Artem</PRE>
</HTML>
==============490FA78EC001C19C61382520==
------------------------------
From: *@dragons.duesouth.net (Matthew Bafford)
Subject: Re: VMWare (Windows on Linux)?
Reply-To: *@dragons.duesouth.net
Date: Tue, 03 Aug 1999 22:40:48 GMT
On Tue, 3 Aug 1999 11:26:07 -0500, Boisy G. Pitre" <[EMAIL PROTECTED]>
cut a telephone line, and tapped the following to comp.os.linux.misc
using only his tongue:
: The one application that concerns me is DirectCD from Adaptec, as well as
: Easy CD creator. I use a HP 8100 CD Writer on my PC, and am wondering if
: VMWare will allow Windows 98 to access this unit as a burner, and also if
: DirectCD will work as well.
In that case, you have nothing to worry about. The HP 8100 is, AFAIK,
supported by cdrecord. Do a search on <URL: http://www.freshmeat.net>
for it.
: Any comments would be appreciated.
HTH,
--Matthew
------------------------------
From: "loatherier" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: Installing Netscape 4.61
Date: Tue, 3 Aug 1999 02:06:39 -0400
Rado Faletic <[EMAIL PROTECTED]> wrote in message =
news:[EMAIL PROTECTED]...
>=20
> > I use IE5.
>=20
> Me too. I find it MUCH more stable than any of the Netscape builds. =
Pity
> MS doesn't get into the Linux buzz, I'm sure many people would still =
use
> MS-Office under Linux, and IE.
>=20
> Not that open source is no good... but MS do have some reasonable
> products and I think we could all benefit from some of them (like =
IE5).
MS doesn't need to get into the Linux buzz. What for? NT is the most =
stable operating system I ever used, even more so than Linux, in my =
humble opinion.
Perhaps, some day, Linux may earn its place among the big boys. Until =
then, I won't hold my breath. MS products are good. Love IE5 -- the =
best browser ever --and MS-Office. Those who can't see the quality of =
the above mentioned, are in the bottom of a deep damp cave struggling to =
get out while blaming Bill Gates for their misfortune.=20
Incidentally, Mac users make the same mistake. =20
------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Subject: Re: Massive video flickering with Mandrake 6.0
Date: Tue, 03 Aug 1999 10:04:21 -0500
Lee Graba wrote:
>
> I just installed Mandrake 6.0, and under both KDE and Gnome, I get
> massive flickering whenever I do just about anything, such as typing,
> using the mouse, etc. It looks like all windows are being drawn about 1
> inch to the right of the actual window for a brief instant, and then
> being drawn in the correct place. I previously had Redhat 5.1 and KDE
> 1.1 installed on the system, and did not have this problem. I
> speculated that the XF86Config file was the culprit, but I had saved the
> previous one and was able to use it, but this did not solve the problem.
>
> I am using a S3 Trio64-based video card. My system is a Pentium Pro 150
> Mhz, 64 Mb of RAM.
>
> Has anyone else had this problem?
>
> --
> ===============================
> Lee Graba
> [EMAIL PROTECTED]
> [EMAIL PROTECTED]
> ===============================
We had a vaguely similar problem when installing RH6.0 on
a Dell Inspiron laptop. We fixed it by choosing a lower
vertical refresh rate in the configuration file. We did this by
commenting out the line that was active and uncommenting
what seemed like an appropriate line.
I would bet it is simply a matter of tuning the configuration.
You might try running Xconfigurator and having it figure out
what to do. Your card should be old enough so XFree should
be able to do it.
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Subject: Re: Can't log in to Red Hat system.
Date: Tue, 03 Aug 1999 09:37:24 -0500
Daniel Doreika wrote:
>
> I have a Red Hat 5.2 system which has just developed a little problem.
> Seems I can no longer log in to it now. Whenever I enter a log in
> name, and hit Enter, instead of asking me for a password, the screen
> clears quickly and goes back to asking for a username. I have
> rebooted the system several times (using Cntrl-Alt-Delete, not by
> hitting reset), but I still can't log in. Nor can I telnet in either,
> all I get is this:
>
> Red Hat Linux release 5.2 (Apollo)
> Kernel 2.0.36 on an i586
>
> ... after which nothing happens. What's weird is that everything else
> with the system appears to be running. It has a web server running,
> which I can get to (this system is on a network, so I can access its
> services from other computers). I can also FTP into the system, but I
> can not get it to allow me to log in, which is of course a major
> problem. If anyone has some suggestions on what I can do to fix this,
> I would be most appreciative. Thanks for any help!
I would suggest logging in single user by using
linux single
at the LILO prompt when booting. Then look at /etc/passwd
and /etc/shadow if you are using shadow passwords. Also look
at their permissions. You might also make copies of some
programs like login from another machine with the same version
of the OS. Use a floppy to transfer the (renamed) copies to
the affected machine and do a cmp to see if they are the same.
It is also possible that someone intruded on your machine and
corrupted your system. Or you may have a hardware problem.
Perhaps someone else will recognize this as a specific problem
and it will be easy to fix. Otherwise if you can't figure out
what caused the problem, you may have to reinstall the system.
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: [EMAIL PROTECTED] (Jan Just Keijser)
Subject: Re: bash question: changing path within script?
Date: Tue, 03 Aug 1999 17:44:12 GMT
In article <[EMAIL PROTECTED]>, "G. Pollack"
<[EMAIL PROTECTED]> wrote:
>I'd like to be able to change the path from within a bash script. The
>script, called set_path (with execute permission set, of course)
>consists of a single line:
>
>PATH=$PATH:/some/new/path
>
>Executing set_path results in no error messages, but inspecting $PATH
>from the command line, shows that it hasn't changed. The above command
>works fine, however, when typed in directly at the command prompt.
>
hmmm, did you come from an MS-DOS environment?
On UNIX and Linux, when you call a script, a child process is created and the
script is run as a child of the existing shell process (the "parent"). This is
similar to MS-DOS, BTW.
However, unlike MS-DOS, UNIX/Linux will not let you change the environment of
your parent process (and rightly so, since the parent process might be owned
by another user, e.g. root).
So your script changes the path, and for the duration of the child script the
PATH has been altered, but as soon as the script ends, the child process is
killed, including your modified PATH and you're back at your old PATH
statement
What you need to do is to run the script in the current shell (AKA "sourcing
the script'). You can do this by typing
. <your_script>
in bash/ksh, which is similar to the
source <your_script>
statement in csh/tcsh
HTH,
JJ
==========================================================
*NOTE*
My Email return address is not correct
in order to avoid mass mailings...
These are the correct addresses
(but with dashes between all letters):
Jan Just (JJ) Keijser
Unix Support Engineer / Configuration Manager
Logica Inc. - Lexington MA
SMTP: [EMAIL PROTECTED]
Just to confuse some of those junkmailers:
[EMAIL PROTECTED]
Your mouse has moved. Windows must be restarted for
the change to take effect. Reboot now? [OK]
My views are my own...
flames > /dev/null 2>&1
==========================================================
------------------------------
** 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
******************************
