Linux-Setup Digest #596, Volume #20 Fri, 9 Feb 01 11:13:07 EST
Contents:
newbie question: configuring keyboard ([EMAIL PROTECTED])
Need help:routing table for 1 eth and n ppp interfaces ("Wu, Simon
[WDLN2:2X38:EXCH]")
Re: Staroffice 5.1 and Linux on slow machine (Kevin Croxen)
can't get scsi tqc to work with rh6.2 (dan)
Write protect windows mount (Jim Lokken)
Re: install on old comp or new? Yo help need Linux ([EMAIL PROTECTED])
Re: Need help:routing table for 1 eth and n ppp interfaces (James Carlson)
Re: Write protect windows mount (Lew Pitcher)
USB printer trouble ([EMAIL PROTECTED])
Re: Is swap really needed? ("g.montgomery")
Re: Is swap really needed? (Paul Kimoto)
Re: Error when Netscape starts ([EMAIL PROTECTED])
Re: Is swap really needed? ("g.montgomery")
Re: backspace vs. delete (Paul Kimoto)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: newbie question: configuring keyboard
Date: Fri, 09 Feb 2001 14:03:35 GMT
Hi all, I've installed SuSE 6.3 and I don't find the file that specifies the
keyboard language to change the current settings from <?> to Italian. Could
you help me ? Thank you GmB
Sent via Deja.com
http://www.deja.com/
------------------------------
From: "Wu, Simon [WDLN2:2X38:EXCH]" <[EMAIL PROTECTED]>
Crossposted-To: comp.protocols.ppp
Subject: Need help:routing table for 1 eth and n ppp interfaces
Date: Fri, 09 Feb 2001 09:41:21 -0500
Hello,
In one of my special test, I need to make multiple ppp configurations on
one of my linux (Redhat 6.2) PC, which already have an eth0 connection
to the corporate LAN. I need to create a IP routing for ppp0, ppp1, ...
ppp32 in the same PC. Two questions:
1. Should I use "defaultroute" in the /etc/ppp/options file?
2. I require all datagram for each ppp# go through the ppp# interface,
not the eth0 interface which is up all the time. I tried a few routing
opions in /etc/ppp/ip-up shell script file. None worked. Which argument
should I use for /sbin/route for this situation?
I appreciated any help. Thanks a lot.
Simon
------------------------------
From: [EMAIL PROTECTED] (Kevin Croxen)
Subject: Re: Staroffice 5.1 and Linux on slow machine
Date: 9 Feb 2001 14:55:16 GMT
I use 5.2 on a similar machine from time to time, and while Star Office
takes forever to load, its response time once loaded is tolerable. You
need to conserve CPU cycles and memory by using a very minimal window
manager (nothing bulkier than, say, fvwm) and you should be more-or-less all
right.
But if these folk only word process, why not just load a word processor
instead of a bloated office suite? WordPerfect 8 is actually very zippy
in the environment described above, and so is Applix Words.
And, really, if all they do is word process with word 95, why change their
environment at all? The hardware as described certainly remains adequate
for that --they have the licenses for the software, they are all trained and
productive, why uproot all of that and invest the time and money retooling
when their current desktop tools meet their needs at least as well as
what will be substituted?
--Kevin
In article <960k04$lga$[EMAIL PROTECTED]>, TheMartian wrote:
>Hi
>
>I have several small clients of mine that are running old hardware, all P120
>or worse with 64MB, 2GB disks. Currently with win95 and word 95, thay want to
>extend the life of these things, and as they only ever do WP, no lan, no
>internet etc. I am thinking of installing a cut down linux with Star Office
>as the only app. Doing this is not the issue, I just have never seen Linux
>and Star office on a machine this small.
>
>I have done this already for severl people, all with P166 or better, the
>results are better than windoze (no supprise there) on the same hardware,
>several happy users saving $$$
>
>Any thoughts on this?
>Will it be fast enough for them?
>
>David
>--
>Sydney, Australia
>
>
>Sent via Deja.com
>http://www.deja.com/
------------------------------
From: dan <[EMAIL PROTECTED]>
Subject: can't get scsi tqc to work with rh6.2
Date: Fri, 09 Feb 2001 09:01:37 -0600
i can't get the parameters passed to linux
(as per /usr/src/linux/drivers/scsi/README.aic7xxx)
to work
i have an all-scsi system with LILO installed
and have tried to pass the parameters when booting the kernel
and also have specified them in lilo.conf as "append="
after booting i look at /proc/scsi/aic7xxx/0 and
all values are 255
Tag Queue Enable Flags: 0x0000
Ordered Queue Tag Flags: 0x0000
Default Tag Queue Depth: 8
Tagged Queue By Device array for aic7xxx host instance 0:
{255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
Actual queue depth per device for aic7xxx host instance 0:
{1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}
the help when building the kernel said this may
give significant performance increase for devices which support
tqc
the drives are all LW type devices
either U2W or U160 with an adaptec 2940U2W controller
thanks for any help,
dan
------------------------------
From: Jim Lokken <[EMAIL PROTECTED]>
Subject: Write protect windows mount
Date: Fri, 09 Feb 2001 09:03:15 -0600
I would like to make my Windows drive writable only by root. Any other
users should only have read privleges as I don't want someone learning
linux to accidentally mess something up on windows. If I need something
transferred over, I would do it as root. As root, I went to the /mnt
directory and did "chmod 755 windows". That command doesn't seem to
actually change the permissions. The only way I found to make it read
only was to go in linuxconf and set the read only flag. That solves the
first part I need by not letting joe user write to it, but it also makes
it so the root can't write to it. If I want to write to it, I have to
go back into linuxconf and change it. Something that takes longer than
I would like it to.
Can someone recommend the best way to do this for me?
JRL
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: install on old comp or new? Yo help need Linux
Date: Fri, 09 Feb 2001 15:04:57 GMT
Thanks Dan for your help. I was wondering. I saw that caldera allowed
the use of pci modems yet I read
somewhere(actually pete norton) that pci modems were a problem a lot of
times for Linux. Can I install a pci modem instead of an isa?
Anyway thanks for your help.
Pete
Sent via Deja.com
http://www.deja.com/
------------------------------
From: James Carlson <[EMAIL PROTECTED]>
Crossposted-To: comp.protocols.ppp
Subject: Re: Need help:routing table for 1 eth and n ppp interfaces
Date: 09 Feb 2001 09:53:09 -0500
"Wu, Simon [WDLN2:2X38:EXCH]" <[EMAIL PROTECTED]> writes:
> In one of my special test, I need to make multiple ppp configurations on
> one of my linux (Redhat 6.2) PC, which already have an eth0 connection
> to the corporate LAN. I need to create a IP routing for ppp0, ppp1, ...
> ppp32 in the same PC. Two questions:
> 1. Should I use "defaultroute" in the /etc/ppp/options file?
No. The "defaultroute" option means that PPP will install a default
route pointing over the PPP link. It's generally used when the PPP
link connects to an ISP, and should never be used if the peer is just
an end node.
Use "nodefaultroute" instead to make sure that no bogus routes are
added by these remote users.
> 2. I require all datagram for each ppp# go through the ppp# interface,
> not the eth0 interface which is up all the time. I tried a few routing
> opions in /etc/ppp/ip-up shell script file. None worked. Which argument
> should I use for /sbin/route for this situation?
For most configurations, you shouldn't need any commands to do this.
Please post more information about your configuration. There are lots
of options here, and it's impossible to tell what you're doing without
having detailed information.
The easiest way to configure simple remote nodes is to give each node
an unused remote address that's inside the same subnet as assigned to
the eth0 interface, and use the "proxyarp" option. Don't forget to
enable IP forwarding as well.
--
James Carlson, Internet Engineering <[EMAIL PROTECTED]>
SUN Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677
Second Edition now available - http://people.ne.mediaone.net/carlson/ppp
------------------------------
From: [EMAIL PROTECTED] (Lew Pitcher)
Subject: Re: Write protect windows mount
Reply-To: [EMAIL PROTECTED]
Date: Fri, 09 Feb 2001 15:17:30 GMT
On Fri, 09 Feb 2001 09:03:15 -0600, Jim Lokken <[EMAIL PROTECTED]>
wrote:
>I would like to make my Windows drive writable only by root. Any other
>users should only have read privleges as I don't want someone learning
>linux to accidentally mess something up on windows. If I need something
>transferred over, I would do it as root. As root, I went to the /mnt
>directory and did "chmod 755 windows". That command doesn't seem to
>actually change the permissions. The only way I found to make it read
>only was to go in linuxconf and set the read only flag. That solves the
>first part I need by not letting joe user write to it, but it also makes
>it so the root can't write to it. If I want to write to it, I have to
>go back into linuxconf and change it. Something that takes longer than
>I would like it to.
>
>Can someone recommend the best way to do this for me?
>
>JRL
man mount, and look for the -o flag
umount /mnt/windows
chmod 0755 /mnt/windows
mount -t vfat -o umask=033 /dev/hd.of.windows.drive /mnt/windows
Lew Pitcher
Information Technology Consultant
Toronto Dominion Bank Financial Group
([EMAIL PROTECTED])
(Opinions expressed are my own, not my employer's.)
------------------------------
Crossposted-To: comp.os.linux.hardware,comp.os.linux.misc,alt.os.linux
Subject: USB printer trouble
From: [EMAIL PROTECTED]
Date: 09 Feb 2001 10:20:16 -0500
Hi,
I'm running a RedHat7 fully updated, kernel 2.4.1. and I own a Cannon
BJC-2110 printer which I'm trying to use through the USB port. I have
support in my kernel for USB and when I boot up I can the USB-UHCI (I
think) is being used. when I try to configure my printer using
printtool it does not find any local printers, but I tell it that the
printer is under /dev/usb/lp0 and everything seems fine. the I restart
the lpd daemon and again everything seems fine. then I send a file to
be printed using lpr, and I monitor what's happening using lpq, and
what's happening is that it says the job is being "processed" and
that's it the printer never prints. I also tried to manually load some
modules like printer.o usbcore.o and usb-utci.o. some of them were
already loaded and it didn't help at all.
any help would be appreciated.
p.s. the printer works under windows.
--
#############################################
# # # # # # # #
# # # ALWAYS BE TRUE TO ONESELF # # #
###########################################
------------------------------
From: "g.montgomery" <[EMAIL PROTECTED]>
Subject: Re: Is swap really needed?
Date: Fri, 09 Feb 2001 15:24:16 GMT
Thanks for the comment, Paul. I will note the buff and cached
entries in the future to see how they play into the overall picture.
However, I still haven't given complete voice to my concern.
It isn't that applications (via the operating system) claim all
the turf in sight, it's that when the process is killed, the memory
claimed is not released back. If that were to be done by a
specific application, I think it might be called a memory leak -
memory "malloc'd" and not "free'd".
I just re-booted RH7.0 system. Before rebooting, the only
"application" I was running, aside from the many daemons,
X, and the stuff that gets run when you log onto the system,
is Netscape Communicator (Navigator + Messenger). Top's
"mem:" showed the usual 508976K ( 512M less the 8 Meg
for the AGP). The "used" was within 4 Meg of that amount,
implying to me that I had a set of whopper applications in
mrmory. But none showed up on the %MEM column,
save for the usual X, netscape-communicator, etc. Then,
I re-booted and after logging on and starting X, I showed
only 84 Meg used. I started Netscape, the used number
climbed, and then Messenger, still a little more climbing.
I am now sitting at 135 Meg used, and the only apps I have
brought in are Communicator (Navigator + Messenger).
I would expect in a system which is operating properly,
that when I shut down Netscape (Navigator and Communicator)
and return to the status just after I had logged on,
started X and invoked top in an xterm, that the system
would return to the 84 Meg of memory used, as
opposed to the now 136 Meg it claims is used. To
my way of thinking, there is dead wood out there, if
I have no application that is reserving the space. The
application which called for the space is gone, so the
dibs on the memory should also be gone if the OS
and it's children are working correctly!
Am I wrong?
Gene.
Paul Kimoto wrote:
> In article <[EMAIL PROTECTED]>, g.montgomery wrote:
> > 2. Now, my box sits here and (if I can believe the Mem: numbers
> > on "top"), linux 2.2 soaks up almost all the available memory
> > in sight, even with only a web browser running on the screen.
> > I know, there are a lot of processes using up Ram, but I bet
> > if top were to have a total for the %MEM column, I would only
> > be using maybe 10 % in reality.
> >
> > 3. While a small number of processes are managing to
> > fool me into saying they are using 505596K out of 508976K
> > available, the Swap: says that only 5980K is used out of
> > the 273088K available.
> >
> > What is going on here? Is this the way it is supposed to work?
>
> Yes. Look at the figures for "buff" and "cached". The system uses "spare"
> memory to improve performance. The memory is there; why shouldn't it use
> it? Most of it can be made available for processes when needed.
>
> --
> Paul Kimoto
> This message was originally posted on Usenet in plain text. Any images,
> hyperlinks, or the like shown here have been added without my consent,
> and may be a violation of international copyright law.
------------------------------
From: [EMAIL PROTECTED] (Paul Kimoto)
Subject: Re: Is swap really needed?
Date: 9 Feb 2001 10:44:56 -0500
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>, g.montgomery wrote:
> Thanks for the comment, Paul. I will note the buff and cached
> entries in the future to see how they play into the overall picture.
> However, I still haven't given complete voice to my concern.
> It isn't that applications (via the operating system) claim all
> the turf in sight, it's that when the process is killed, the memory
> claimed is not released back. If that were to be done by a
> specific application, I think it might be called a memory leak -
> memory "malloc'd" and not "free'd".
>
> I just re-booted RH7.0 system. Before rebooting, the only
> "application" I was running, aside from the many daemons,
> X, and the stuff that gets run when you log onto the system,
> is Netscape Communicator (Navigator + Messenger). Top's
> "mem:" showed the usual 508976K ( 512M less the 8 Meg
> for the AGP). The "used" was within 4 Meg of that amount,
> implying to me that I had a set of whopper applications in
> memory. But none showed up on the %MEM column,
> save for the usual X, netscape-communicator, etc. Then,
> I re-booted and after logging on and starting X, I showed
> only 84 Meg used. I started Netscape, the used number
> climbed, and then Messenger, still a little more climbing.
> I am now sitting at 135 Meg used, and the only apps I have
> brought in are Communicator (Navigator + Messenger).
>
> I would expect in a system which is operating properly,
> that when I shut down Netscape (Navigator and Communicator)
> and return to the status just after I had logged on,
> started X and invoked top in an xterm, that the system
> would return to the 84 Meg of memory used, as
> opposed to the now 136 Meg it claims is used. To
> my way of thinking, there is dead wood out there, if
> I have no application that is reserving the space. The
> application which called for the space is gone, so the
> dibs on the memory should also be gone if the OS
> and it's children are working correctly!
>
> Am I wrong?
Yes, you are wrong.
The amount top(1) shows are "used" _includes_ the amounts under "buff" and
"cached". Look at the output of free(1), which shows the same quantities
but also subtracts out "buff" and "cached" from "used" for your
convenience.
Example: Even after you shut down netscape, the images of the executable,
its dynamically linked libraries, and other parts of the filesystem (e.g.,
your .netscape/cache) remain cached (or maybe that's "buffered", I can
never remember which) in memory until the system finds some better use for
that memory. (Why should it not?) You may find that large programs (e.g.,
netscape) start up noticably slower the first time. That's because they
may already be (partially) mapped into memory on subsequent invocations.
--
Paul Kimoto
This message was originally posted on Usenet in plain text. Any images,
hyperlinks, or the like shown here have been added without my consent,
and may be a violation of international copyright law.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Error when Netscape starts
Date: Fri, 09 Feb 2001 15:39:17 GMT
Thanks for the reply & the correct anwser that was the problems.
Running great now!
Sent via Deja.com
http://www.deja.com/
------------------------------
From: "g.montgomery" <[EMAIL PROTECTED]>
Subject: Re: Is swap really needed?
Date: Fri, 09 Feb 2001 15:50:20 GMT
I just re-started Netscape so as to "reply" to this thread.
To recap: I re-booted my RH7 box, logged on as a user,
then started a terminal and top. The Mem was at 84 Meg.
Then I started Netscape Communicator (Navigator,
and then Messenger to get to the Newsgroups). The
Mem went up to something over 100 Meg, and crept up
to end up at 136 Meg just before I sent the previous post.
Current: After terminating Netscape all the way, the Mem
went down to 104 Meg. I would have expected it to
go down to 84 Meg again, since the program was not
in execution. So, I think I lost 20 Meg of dead wood right
there, one time. I would have to go through this cycle
multiple time to see if it is cumulative. But I suspect
that there is just something that I don't understand
about the memory "used" tally. If the combination
of Netscape and Linux had a "leak" as bad as 20 Meg
for a session, someone would have brought it up
before, I think. And, I am not sure that other apps
don't cause the same sort of "used" indications. In fact,
I'm not sure I should even be calling this a leak - maybe
my problem is in interpretation of what I am seeing.
Gene.
"g.montgomery" wrote:
> Thanks for the comment, Paul. I will note the buff and cached
> entries in the future to see how they play into the overall picture.
> However, I still haven't given complete voice to my concern.
> It isn't that applications (via the operating system) claim all
> the turf in sight, it's that when the process is killed, the memory
> claimed is not released back. If that were to be done by a
> specific application, I think it might be called a memory leak -
> memory "malloc'd" and not "free'd".
>
> I just re-booted RH7.0 system. Before rebooting, the only
> "application" I was running, aside from the many daemons,
> X, and the stuff that gets run when you log onto the system,
> is Netscape Communicator (Navigator + Messenger). Top's
> "mem:" showed the usual 508976K ( 512M less the 8 Meg
> for the AGP). The "used" was within 4 Meg of that amount,
> implying to me that I had a set of whopper applications in
> mrmory. But none showed up on the %MEM column,
> save for the usual X, netscape-communicator, etc. Then,
> I re-booted and after logging on and starting X, I showed
> only 84 Meg used. I started Netscape, the used number
> climbed, and then Messenger, still a little more climbing.
> I am now sitting at 135 Meg used, and the only apps I have
> brought in are Communicator (Navigator + Messenger).
>
> I would expect in a system which is operating properly,
> that when I shut down Netscape (Navigator and Communicator)
> and return to the status just after I had logged on,
> started X and invoked top in an xterm, that the system
> would return to the 84 Meg of memory used, as
> opposed to the now 136 Meg it claims is used. To
> my way of thinking, there is dead wood out there, if
> I have no application that is reserving the space. The
> application which called for the space is gone, so the
> dibs on the memory should also be gone if the OS
> and it's children are working correctly!
>
> Am I wrong?
>
> Gene.
>
> Paul Kimoto wrote:
>
> > In article <[EMAIL PROTECTED]>, g.montgomery wrote:
> > > 2. Now, my box sits here and (if I can believe the Mem: numbers
> > > on "top"), linux 2.2 soaks up almost all the available memory
> > > in sight, even with only a web browser running on the screen.
> > > I know, there are a lot of processes using up Ram, but I bet
> > > if top were to have a total for the %MEM column, I would only
> > > be using maybe 10 % in reality.
> > >
> > > 3. While a small number of processes are managing to
> > > fool me into saying they are using 505596K out of 508976K
> > > available, the Swap: says that only 5980K is used out of
> > > the 273088K available.
> > >
> > > What is going on here? Is this the way it is supposed to work?
> >
> > Yes. Look at the figures for "buff" and "cached". The system uses "spare"
> > memory to improve performance. The memory is there; why shouldn't it use
> > it? Most of it can be made available for processes when needed.
> >
> > --
> > Paul Kimoto
> > This message was originally posted on Usenet in plain text. Any images,
> > hyperlinks, or the like shown here have been added without my consent,
> > and may be a violation of international copyright law.
------------------------------
From: [EMAIL PROTECTED] (Paul Kimoto)
Subject: Re: backspace vs. delete
Date: 9 Feb 2001 10:54:52 -0500
Reply-To: [EMAIL PROTECTED]
In article <95v472$[EMAIL PROTECTED]>, John Doner wrote:
> I run Linux on a PC at home, and MacX 1.5 on a Macintosh at
> work. I work a lot on a Unix system, via MacX. I can also
> have windows from my Linux system open at work. Emacs likes
> to use 'Delete' (octal 177) for its erase character, so I set
> my xterm sessions the same way, using stty. That works for
> the Unix system, but not for Linux. Regardless of what I do
> with stty, or the -tm option of xterm, I can't get it to
> recognize Delete as an erase character; it remains Backspace
> (octal 10). Interestingly, Emacs is normal; an Emacs session
> with a window on my Mac responds to Delete as expected. The
> xterm session just flashes the menu bar (I suppose it would
> beep if I made some other change), and doesn't erase.
>
> Obviously, there's something I'm missing. Any ideas, folks?
Backspace and delete make my head hurt.
Have you looked at any of these?
http://www.linuxdoc.org/HOWTO/Keyboard-and-Console-HOWTO-5.html
http://www.linuxdoc.org/HOWTO/mini/BackspaceDelete/
http://www.ibb.net/~anne/keyboard.html
--
Paul Kimoto
This message was originally posted on Usenet in plain text. Any images,
hyperlinks, or the like shown here have been added without my consent,
and may be a violation of international copyright law.
------------------------------
** 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.setup.
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-Setup Digest
******************************
