Linux-Development-Sys Digest #196, Volume #6 Thu, 31 Dec 98 22:14:14 EST
Contents:
Linux directions (Jim O'Day)
Driver for Ali 1541/1543 IDE chipset? (Bill Gribble)
help with trakker QIC-80 tape drive connected at printer port ("guenter weissenseel")
Re: parport/ppa problem with 2.2.0pre1 (Michael Powe)
[Q] Bug Report: Kernel 2.0.36/AIC7xxx problem (Michael Peterson)
'make xconfig' problem in 2.2.0-pre2 (N1ho)
Re: Registry for Linux - Why? (Christopher B. Browne)
interrupt handling in device drivers (nexus)
Re: Something Simple? ("David D. Gitchell")
Help: problems with 'w', 'who' and 'last' (Ted Sariyski)
Re: Registry for Linux - Why? (George MacDonald)
Re: parport/ppa problem with 2.2.0pre1 (Griffin Caprio)
Re: How to run Windows Applications on Linux (Eric Miller)
Re: rewrittable CD (David Philippi)
----------------------------------------------------------------------------
From: Jim O'Day <[EMAIL PROTECTED]>
Subject: Linux directions
Date: Thu, 31 Dec 1998 13:08:29 -0500
Anyone know of a good source for a roadmap of Linux in the future if
there is such a thing?
------------------------------
From: Bill Gribble <[EMAIL PROTECTED]>
Subject: Driver for Ali 1541/1543 IDE chipset?
Date: 31 Dec 1998 17:09:06 -0600
Are there drivers in development for this chipset? I noticed an Ali
14xx driver in very recent kernels; what's the difference between
these two chipsets? It's in a new Gigabyte socket 7/AGP motherboard.
Thanks,
Bill Gribble
------------------------------
From: "guenter weissenseel" <[EMAIL PROTECTED]>
Subject: help with trakker QIC-80 tape drive connected at printer port
Date: Thu, 31 Dec 1998 18:10:08 -0500
Hi
I use Caldera 1.3 Kernel version 2.0.35
Is there a module for the trakker tape backup available?
It is an external device, whichoperates through the printer port and uses
standard QIC-80 tapes.
Thanks for any help
Guenter
------------------------------
From: Michael Powe <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.hardware
Subject: Re: parport/ppa problem with 2.2.0pre1
Date: 31 Dec 1998 15:14:13 -0800
=====BEGIN PGP SIGNED MESSAGE=====
Hash: SHA1
[posted and mailed]
>>>>> "Griffin" == Griffin Caprio <[EMAIL PROTECTED]> writes:
Griffin> I can't even get my printer to work under 2.2.0pre1. It
Griffin> works fine in 2.0.36, but not in the newer kernels. I
Griffin> tried to switch from lp1 to lp0 and it still doesn't
Griffin> work. Any tips?
Have you read parport.txt in /usr/src/linux/Documentation? Might be
some necessary information in there.
To get my printer to work on my PC I added this to my lilo startup
line:
parport=0x378,7 lp=parport0
Then I set /dev/lp -> /dev/lp0 and in /etc/printcap, changed the
device to /dev/lp0 as well.
When you boot, if you look at your boot messages, you should see
something like
parport0: PC-style at 0x378, irq 7 [SPP,ECP,ECPPS2]
[ ... ]
lp0: using parport0 (interrupt-driven).
Looking at these messages may give you a clue where the error occurs.
mp
8<---------------how-easy-is-it-to-demunge-an-address?------------------->8
#! /usr/bin/perl # if you are [EMAIL PROTECTED] (Another Luser):
while ($line = <>){ if ($line =~ m/^\s*$/ ){ last; }
if ($line =~ m/^From: (\S+) \(([^()]*)\)/){ $from_address = $1; } }
if ($from_address =~ m/\S+NOSPAM\S+/){ $x = index($from_address, NOSPAM);
substr($from_address, $x, 6+1) = ""; printf("The real address is %s\n",
$from_address);}else { printf("No munge, just plain %s\n",$from_address);}
printf("\nBrought to you by the Truth In Mail Headers Foundation\n");
8<-----------------------here's-one-example------------------------------>8
- --
Michael Powe
[EMAIL PROTECTED] http://www.trollope.org
Portland, Oregon USA
=====BEGIN PGP SIGNATURE=====
Version: PGP for Personal Privacy 5.0
Charset: noconv
iQA/AwUBNowE6rajuNi/6Js3EQLoPgCgoTE+UrkEnu30NaDLQhz7NDx0nBsAoI5p
47L5adQ/9PpslyXbLblapPLQ
=A3o8
=====END PGP SIGNATURE=====
------------------------------
From: Michael Peterson <[EMAIL PROTECTED]>
Subject: [Q] Bug Report: Kernel 2.0.36/AIC7xxx problem
Date: Thu, 31 Dec 1998 16:12:06 -0800
My 2.0.36 kernel (Redhat 5.2) recompiled to use the AIC7xxx driver does not
see/install my SCSI tape device. However, a 2.1.132 kernel, compiled as before,
correctly installs the tape device.
My adapter is a 2940U/W.
Is this a known problem? Is an updated driver available?
I'd be happy to send all the details (logs, detailed descriptions, etc.,) if
anyone is interested in pursuing this and/or helping me get my tape device
installed on 2.0.36.
Thanks,
Michael
------------------------------
From: [EMAIL PROTECTED] (N1ho)
Subject: 'make xconfig' problem in 2.2.0-pre2
Date: 1 Jan 1999 00:25:01 GMT
While attempting to build a kernel under 2.2.0-pre2, the 'make xconfig' failed
with an error. I read Linus' and Alan's comments about this (for 'pre1') and I
hope I can add some constructive observations.
The error that I got was:
wish -f scripts/kconfig.tk
Error in startup script: invalid command name "clear_choices"
while executing
"clear_choices"
(procedure "read_config" line 3)
invoked from within
"read_config .config"
(file "scripts/kconfig.tk" line 485)
Well, I took a look at kconfig.tk and compared it to the one that was used
in 2.1.131, and while they are similar, it appears it was extensively rewritten
for 2.2.0-pre*. The function "clear_choices" (and "update_choices" and an
unknown number of others) are called from within another function in the
script, but the '_choices' functions are NOT defined, at least not in the same
fashion that they were in .131. (There may be some sort of C-like header file
or Ksh-like function file inclusion mechanism that was used instead, which
may be broken. I am not a Tk expert, and I only took a cursory look). I never
got around to testing .132, so I don't know when the changes were introduced.
Good news is that I was able to use 'make menuconfig' to build my kernel, and
there appear to be no regressions from .131. It even now calls my AWE64Gold
an EMU8000 now. (I still had to tweak drivers/sound/lowlevel/awe_config.h).
======
If you send me spam (except for Hormel products), be aware that I'll simply
forward
it to [EMAIL PROTECTED] and let AOL's legal team handle it as they see fit.
------------------------------
From: [EMAIL PROTECTED] (Christopher B. Browne)
Subject: Re: Registry for Linux - Why?
Date: 31 Dec 1998 19:31:31 GMT
Reply-To: [EMAIL PROTECTED]
On Thu, 31 Dec 1998 17:40:49 GMT, Timothy J. Lee
<[EMAIL PROTECTED]> posted:
>Thornton Prime <[EMAIL PROTECTED]> writes:
>|1. Centralized administration:
>| Having all the configuration settings in one place makes backup and
>|management easier.
>
>Not necessarily. Some programs are best installed on the local disk
>at the time of OS install, while others are often installed on file
>servers and used by client computers directly from the file server.
>The latter type of programs can include browsers, editors, compilers,
>etc.. If installing a new browser required installing it on every one
>of the client computers (because each computer requires that its
>centralized "registry" be updated), rather than a single install on a
>file server, the amount of system management work would be significantly
>increased.
This still doesn't disagree with the notion that configuration information
could reasonably be "centralized." The key is that different parts of system
configuration are "centred" in different places, thus suggesting that there
needs to be a multiplicity of "configuration centres."
- Thus, overall config for a LAN probably should stay in a "central"
location *on the LAN,* over all hosts and all users. (Which is largely how
a DNS name server works...)
- Hardware config for a particular host should surely be "centralized" some
place on that host.
- "Global" application defaults should be centralized in some place
associated with the application, possibly "over all hosts and all users."
- Individual users' application and environmental configuration should be
centralized in some place associated with the user. (Arguably "over all
hosts," which probably means having $HOME sit on NFS some place to be
exported to wherever the user is, or perhaps residing in some other form of
"database" that can be accessed from any host...)
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
[EMAIL PROTECTED] - "What have you contributed to Linux today?..."
------------------------------
From: [EMAIL PROTECTED] (nexus)
Subject: interrupt handling in device drivers
Date: 31 Dec 1998 20:25:19 GMT
i understand the irq_request() function, but say i wanted to hook
some code up to the timer interrupt or the keyboard interrupt...in DOS
its simple- i just change the vectors in the IDT to point to my function
when the hard/software interrupt is called via some easy to use bios
function calls....but in linux it seems that i have to request a unique
IRQ for my device or else share one w/ another device- but the keyboard
and timer IRQ's are unsharable....is there any way I can hook my code up
to an interrupt vector number (like say 0x#)?......
thanx for any help- im really confused......
------------------------------
From: "David D. Gitchell" <[EMAIL PROTECTED]>
Subject: Re: Something Simple?
Date: Thu, 31 Dec 1998 14:21:45 -0600
Yes, the problem is something simple: Not all of the elements of the
stat struct are of size integer, yet you used the "%d" (integer) format
specification for all of the parameters to printf. This should put you
out of alignment with the remaining parameters after you hit the first
long in the parameter string.
ncc1701d wrote:
>
> I'm using the
> following code for my printf statement in my program:
>
> printf("File Statistics:\n");
> printf(" Device: %d\n"
> " Protection: %d\n"
> " Hard Link Count: %d\n"
> " Owner's ID: %d\n"
> " Group's ID: %d\n"
> " Device Type: %d\n"
> " File Size: %d\n"
> "FileSysBlocksize: %d\n"
> "Allocated blocks: %d\n"
> "Last Accessed on: %d\n"
> "Last Modified on: %d\n"
> " Last Changed on: %d\n"
> "Inode:
> %d\n",st.st_dev,st.st_mode,st.st_nlink,st.st_uid,st.st_gid,st.st_rdev,st.st_
> size,st.st_blksize,st.st_blocks,st.st_atime,st.st_mtime,st.st_ctime,st.st_in
> o);
Here's a simple program to display the sizes of each of the elements of
the stat structure, followed by the results of its execution:
#include <sys/stat.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
/* ********************************************************* */
void main() {
printf("Sizes of 'struct stat' elements:\n");
printf("st_dev: \t%d\n",sizeof(dev_t));
printf("st_ino: \t%d\n",sizeof(ino_t));
printf("st_mode:\t%d\n",sizeof(mode_t));
printf("st_nlink:\t%d\n",sizeof(nlink_t));
printf("st_uid: \t%d\n",sizeof(uid_t));
printf("st_gid: \t%d\n",sizeof(gid_t));
printf("st_rdev:\t%d\n",sizeof(dev_t));
printf("st_size:\t%d\n",sizeof(off_t));
printf("st_blksize:\t%d\n",sizeof(unsigned long));
printf("st_blocks:\t%d\n",sizeof(unsigned long));
printf("st_atime:\t%d\n",sizeof(time_t));
printf("st_mtime:\t%d\n",sizeof(time_t));
printf("st_ctime:\t%d\n",sizeof(time_t));
}
/* ********************************************************* */
Sizes of 'struct stat' elements:
st_dev: 8
st_ino: 4
st_mode: 4
st_nlink: 4
st_uid: 4
st_gid: 4
st_rdev: 8
st_size: 4
st_blksize: 4
st_blocks: 4
st_atime: 4
st_mtime: 4
st_ctime: 4
Note that st_dev is EIGHT bytes long. Since you passed the %d format
specification to printf, printf will take the first four bytes off the
stack as an integer parameter, and print that result. The next four
bytes of that same parameter will be used for the NEXT %d format
specification, and each of your entries will be displaced by one entry
for the rest of the printf statement.
Note also that you will have the same problem when you hit st_rdev.
You have to give the correct format specifications to printf; it can't
second-guess what you're passing it.
>From this you should be able to figure out on your own why it works when
you print them individually, but doesn't when you stack them into one
printf statement.
Enjoy!
-- Dave
============================================================
#include <std_disclaimer.h> /* I speak for myself, only. */|
============================================================
LCDR David D. Gitchell, USN (Retired), Hutchinson, KS, USA
[EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED]
------------------------------
From: Ted Sariyski <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Help: problems with 'w', 'who' and 'last'
Date: Wed, 30 Dec 1998 10:24:34 -0800
Recently I upgraded our machines from RedHat5.1 to
RedHat5.2. Everything is OK, but now three machines behave
strange with w, who and last. last returns trash:
[ted@sor ted]$ last
***@* **** dabdub Wed Dec 31 16:00
still logged in
p1 21.pv.reshsg Wed Dec 31 16:00 -
16:00 (00:00)
dialin33 p1 Wed Dec 31 16:00 -
16:00 (00:00)
...
w and who give only the top line and empty body:
[ted@turina1 ted]$ w
10:03am up 7 days, 19:52, 0 users, load average: 0.00,
0.00, 0.00
USER TTY FROM LOGIN@ IDLE JCPU
PCPU WHAT
I copied these executables from machines where they behave
well along with /var/log/wtmp file and for a while this
fixed the problem, but after a couple of minutes the problem
appeared again. It seems that there is a reference data set
that is updated periodically ... I really have no idea what
is going on.
Is there an easier way to fix this problem than reinstalling
the OS?
Thanks. Ted
------------------------------
From: George MacDonald <[EMAIL PROTECTED]>
Subject: Re: Registry for Linux - Why?
Date: Thu, 31 Dec 1998 23:41:54 GMT
Christopher B. Browne wrote:
>
> On Thu, 31 Dec 1998 17:40:49 GMT, Timothy J. Lee
> <[EMAIL PROTECTED]> posted:
> >Thornton Prime <[EMAIL PROTECTED]> writes:
> >|1. Centralized administration:
> >| Having all the configuration settings in one place makes backup and
> >|management easier.
> >
> >Not necessarily. Some programs are best installed on the local disk
> >at the time of OS install, while others are often installed on file
> >servers and used by client computers directly from the file server.
> >The latter type of programs can include browsers, editors, compilers,
> >etc.. If installing a new browser required installing it on every one
> >of the client computers (because each computer requires that its
> >centralized "registry" be updated), rather than a single install on a
> >file server, the amount of system management work would be significantly
> >increased.
>
> This still doesn't disagree with the notion that configuration information
> could reasonably be "centralized." The key is that different parts of system
> configuration are "centred" in different places, thus suggesting that there
> needs to be a multiplicity of "configuration centres."
>
> - Thus, overall config for a LAN probably should stay in a "central"
> location *on the LAN,* over all hosts and all users. (Which is largely how
> a DNS name server works...)
>
> - Hardware config for a particular host should surely be "centralized" some
> place on that host.
>
> - "Global" application defaults should be centralized in some place
> associated with the application, possibly "over all hosts and all users."
>
> - Individual users' application and environmental configuration should be
> centralized in some place associated with the user. (Arguably "over all
> hosts," which probably means having $HOME sit on NFS some place to be
> exported to wherever the user is, or perhaps residing in some other form of
> "database" that can be accessed from any host...)
>
That's a nice summary! I think it's a little more complicated. i.e.
Applications can be customized/configured differently for users/groups/
sites/departments/... Such configurations are not flat, i.e. there
could be company wide config(i.e. company name, address ...) then
Department config(Dept code) then group( groupname ) and finally
user(username). Perhaps the department overrides the address if it
has a different mailing address, perhaps not. I can envisage complex
hierachies for application configuration. This could be handled with
an arbitrarily deep path which specify the trees that are projected
on top of one another to form a "current config" view specific
to the current frame of reference.
Also simply centralizing user config does not solve all the problems,
i.e. they may want different settings for apps based on the machine
they are running on. For example different color settings for
workstations that have different color depth. So again allowing
arbitrary levels of a hierarchy would allow both centralized
configs with localized overrides. This would allow users to
roam to other systems carrying as much config as they want,
and being able to tune the app to the current platform if
desired.
If we change the "database" to multiple configBases which
can be cached on each client then I think we get all the
flexabilty needed. It would also be useful to allow
app managers to push config changes from a central source.
This would allow application updates to occur without
requiring users to exit/restart an app. Obviously this
is not required for single user systems, but in a big
network it would be useful to push such changes. For
example lets say a company has to change area codes
due to an area code split. Then it would be really
nice if this could be pushed to each app that uses
the old area code. The ultimate would be if such
a change could be done with a running app via
a notification! Someone mentioned such an implementation
for a single user system, doing it on a network basis
would make it one killer config system.
I could envisage implementing such a system using CORBA
as it provides much of the framework needed. Building
such a system without CORBA would be quite a bit more
work but it's doable. We seem to have come a long way
from a config library and a single system/user Registry.
--
We stand on the shoulders of those giants who coded before.
Build a good layer, stand strong, and prepare for the next wave.
Guide those who come after you, give them your shoulder, lend them your code.
Code well and live! - [EMAIL PROTECTED] (7th Coding Battalion)
------------------------------
From: Griffin Caprio <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.hardware
Subject: Re: parport/ppa problem with 2.2.0pre1
Date: Thu, 31 Dec 1998 15:21:42 -0600
Allin Cottrell wrote:
> I have been using a parallel-port zipdrive with kernel 2.0.36
> with no problem. Now I'm trying to use it with 2.2.0-pre1,
> using parport to get access to both the zipdrive and the printer.
> The printer is working (after shifting it from lp1 to lp0, in
> printcap), but the zipdrive is not.
>
>
I can't even get my printer to work under 2.2.0pre1. It works fine in
2.0.36, but not in the newer kernels. I tried to switch from lp1 to lp0
and it still doesn't work. Any tips?
------------------------------
From: Eric Miller <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.development,comp.os.linux.development.apps,linux.dev.c-programming
Subject: Re: How to run Windows Applications on Linux
Date: Thu, 31 Dec 1998 14:44:47 -0800
David Parrish wrote:
>
> In linux.dev.kernel George Bonser <[EMAIL PROTECTED]> wrote:
> > In article <[EMAIL PROTECTED]>,
> > Sandras Damien <[EMAIL PROTECTED]> writes:
> > >
> > > There is a project to run Win applications on Linux.
> > >
> > > It's called wine.
> > > Let's have a look at http://www.winehq.org (I'm not sure).
> > >
>
> > Except it doesn't work very well. Most applications will not run. If one
>
> Wine is improving very quickly! On www.winehq.org there is a list of
> applications that are proven to run.. Even recent games like StarCraft
> have been running under Wine (I haven't, but supposedly other people
> have).
>
> --
> --------------------------------------------------------------------
> David Parrish [EMAIL PROTECTED]
> Zeta Internet Technical Manager http://www.zeta.org.au/
> --------------------------------------------------------------------
> Mein Luftkissenfahrzueg ist mit Aalen besetzt.
> --------------------------------------------------------------------
I have great interest in being able to use apps written for WINNT under
Linux. Specifically, I use ESRI's ArcView GIS and I'd like not to have
to reboot to Winnt to run it and then back to Linux to run all of the
other stuff. I suspect ArcView would not run well under Wine as it would
be very demanding. Too bad those ESRI folks have jumped on the Windoze
bandwagon in such a big way.
--
Eric Miller: [EMAIL PROTECTED] http://www.tidepool.com/~egm2
There once was a fellow named Pope,
Who plugged into an oscilloscope.
The cyclical trace
Of their carnal embrace
Had a damn near infinite slope.
-- Thomas Pynchon
------------------------------
From: [EMAIL PROTECTED] (David Philippi)
Subject: Re: rewrittable CD
Date: 31 Dec 1998 21:48:51 GMT
On Mon, 28 Dec 1998 11:47:46 +0100, Philippe Le Foll
<[EMAIL PROTECTED]> wrote:
> Can it work under Linux ? if then where could I found some FAQ ?
You can delete it using cdrecord (don't know about xcdroast) but you can't
mount it read/write - yet.
Bye David
------------------------------
** 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.development.system) 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-Development-System Digest
******************************
