Linux-Misc Digest #746, Volume #21 Fri, 10 Sep 99 02:13:08 EDT
Contents:
Re: bypassing fsck (Leslie Mikesell)
Re: Amiga, QNX, Linux and Revolution (Guy Macon)
X server problem (Charles Stroom)
Ret Hat 6.0 ---- Help Help Help .... ("dj")
Cannot find map file? ("Colin R. Day")
Re: rxvt scrolling (Juergen Heinzl)
Re: adaptec AVA1502 HELL! ("Bud Beckman")
Re: Amiga, QNX, Linux and Revolution (Jeffrey C. Dege)
Re: General Rant from a Linux Newbie (Jeff Gentry)
Re: General Rant from a Linux Newbie (Miles Bader)
Re: Amiga, QNX, Linux and Revolution (Tony Talarico)
Re: General Rant from a Linux Newbie (Christopher B. Browne)
Re: REAL PLAYER in LINUX. Which ver. (Eric Y. Chang)
Re: Q? - best combo of linux distrib and apps for 3rd world (B'ichela)
gnome-config: command not found... (Timothy)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: comp.os.linux.setup
Subject: Re: bypassing fsck
Date: 9 Sep 1999 23:07:33 -0500
In article <[EMAIL PROTECTED]>,
Bob Hauck <b o b h @ w a s a t c h . c o m> wrote:
>[EMAIL PROTECTED] writes:
>
>> I'm surprised how many people here assume that the problem of fsck
>> holding everything up for minutes has something to do with unclean
>> unmounting. The problem is of course that every n-th time I boot, fsck
>> decides to check anyway ("maximal mount count reached, check forced"),
>> which I want to disable.
>
>You can change the maximal mount count with tune2fs. If you have
>multiple disks/partitions, it might be helpful to set the count on each
>to a different prime number so you never have to wait through a full
>check.
>
>Or, if you're brave, set it to something huge like 8192.
And you can mount some partitions read-only to eliminate the
the chance of accidental corruption.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Guy Macon)
Crossposted-To: comp.os.qnx,comp.realtime
Subject: Re: Amiga, QNX, Linux and Revolution
Date: 09 Sep 1999 21:12:09 PDT
In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] (Igor
Kovalenko) wrote:
>
>Guy Macon wrote:
>>
>> I have two problems with what the Ballista project is doing here.
>>
>> [1] They fail to differentiate between an OS failing to handle one
>> of the test cases and an OS purposefully handling the test case
>> in a different way than their assumption of what the Only Valid
>>
>> Way To Handle Errors is.
>>
>> [2] They printed posters and put up web pages containing bar charts
>> showing "Robustness Failure Rate" for various Operating Systems,
>> despite the fact that what they are measuring is not robustness.
>>
>> I believe that the Ballista project should measure the ability of
>> each OS to exhibit whatever error-response behavior is documented
>> for that OS. That's what most programmers would expect of a test
>> of "operating system robustness".
>>
>> I also believe that a process failure which prevents the failing
>> process alone from completing it's task should be listed differently
>> from a process failure which prevents unrelated processes or the
>> operating system itself from completing their tasks. In the real
>> world of high reliability systems, the ability of a critical module
>> to continue despite the failure of a non critical module is an
>> important reliability issue.
>>
>> What the Ballista Project is testing for is very useful and I am glad
>> to see them do it. I just wish that they wouldn't call what they are
>> testing for "Robustness Failure Rate", and that they would realize that
>> they are not in charge of deciding what the proper response to a bad
>> input is for all operating systems.
>
>It looks like most people comment Ballista results reading mostly QSSL's
>comments and not the results ;) While in general I agree with QNX
>approach to kill an offending application, there is another issue as
>well. Greg & Dan talked about stray pointers a lot but look at test
>cases published by Ballista. Some of them (e.g., calloc() with negative
>arguments) have NOTHING to do with stray pointers and still get killed.
>That basically means QNX libc does not do a particularly good job of
>argument checking. That saves some overhead but indeed sacrifies
>robustness for some extent. You have to program range checking for all
>input in your applications instead of relying on libc. I ran into
>problems due to that many times, because some QNX functions may return
>undocumented values when you feed them with inappropriate data.
As for my point [1],
pointing out flaws in QNX is all well and good, but the fact remains
that the flaws you list are not what the Ballista project measured.
Instead, they called the documented output of well designed error
handling code "Robustness failures". I really wish that they had
measured real problems, but they didn't. They measured whatever
would make the OS's on there poster look as bad as possible.
As for my point [2],
I can give you a bunch of direct quotes from the Ballista project
website where they themselves say that they are not testing
robustness. Anyone who reads all of the docs for the project
(which I did) can see that what they test for is only one facet
of robustness. The posters and webpage graphics are misleading
to the point of being a reckless disreguard for truth.
------------------------------
From: [EMAIL PROTECTED] (Charles Stroom)
Subject: X server problem
Date: Fri, 10 Sep 1999 04:07:01 GMT
Reply-To: [EMAIL PROTECTED]
I use /opt/kde/bin/klock to lock my Linux when I am away (so, I don't
logout and keep the X up).
After some days my box is up (SuSE 6.1), klock refuses to lock the
screen with the error message:
_X11TransSocketUNIXConnect: Can't connect: errno = 111
kblankscrn.kss: cannot connect to X server :0
Everything else still running, but it will not lock, so I have to stop
and restart the X server. It is a 233 Mhz, 128 Mb, and I have normally
also Staroffice and vmware up (between 20-40 Mb swapped).
Anyone has had similar problems?
--
Charles Stroom
email: [EMAIL PROTECTED]
url: http://www.stroom-schreurs.demon.nl/
------------------------------
From: "dj" <[EMAIL PROTECTED]>
Subject: Ret Hat 6.0 ---- Help Help Help ....
Date: Fri, 10 Sep 1999 12:16:56 -0700
I am very new to Linux (usually a Windows user). Recently I bought RH
version 6 and now trying to install it. I am facing some problems during
installation. Need help desperately.... :-(
During the installation system always hangs when it comes to install
"Desktop back ground images" (10.198MB). Once it hanged during the
installation of another large program. I usually let it run for few hours,
but seems to have no progress and no activities on my PC. I am very keen to
install test Linux. Please help.
Here is my disk setup :
Total space 2.5 Gb
Ram 94 Mb
/ 100 Mb
/usr 1200 Mb
/home 1000 Mb
swap 50 Mb
------------------------------
From: "Colin R. Day" <[EMAIL PROTECTED]>
Subject: Cannot find map file?
Date: Fri, 10 Sep 1999 03:57:57 +0000
Hello,
In my log files, I find the message "Cannot find map file". I have
map=/boot/map in my lilo.conf file. The kernel boots, but could this
have subtle effects? I'm running kernel 2.2.10-4
--
Colin R. Day [EMAIL PROTECTED] alt.atheist #1500
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: rxvt scrolling
Date: Thu, 09 Sep 1999 21:49:02 GMT
In article <[EMAIL PROTECTED]>, aydincem wrote:
>hi,
>my question is about the rxvt x11 term emulator and its scrolling
>function.
>is there any possibility that i can set the scrolling to the keys
>shift-up or shift-pgup like in kvt,xterm... ? the option -sk doesn't
>works, is it this? (do i have to recompile again :(.
>if i have to, how do i compile with two options ?
>./configure -enable-sk -enable-pixmap <-- like this?
try shift+up/down lest the shell is interpreting these keys.
Ta',
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: "Bud Beckman" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.setup
Subject: Re: adaptec AVA1502 HELL!
Date: Thu, 9 Sep 1999 22:13:49 -0600
Dave Nejdl wrote in message
>I just got an internal scsi zip drive with the adaptec AVA-1502 scsi
>controler(zip zoom). I didn't get any manuals or anything with it. I
>reconfigured my kernel to include scsi, scsi disk support and the aha152x
>driver. Then I went into windows nt to try and find out the irq and I/O
>port. It always says the I/O range is 140-015f (0x140 I assume) and the IRQ
>is 11. SO I said "great" and "that was so easy." I entered the boot
>parameters "aha152x=0x140,11,7,1", then I got the message:
>detected 1 controller(s)
>aha152x0: vital data: PORTBASE=0x140, IRQ=11, SCSI ID=7,
>DAVE
Just let the boot handle it. Sometimes it will find the IRQ and may change
it.
GL
------------------------------
From: [EMAIL PROTECTED] (Jeffrey C. Dege)
Crossposted-To: comp.os.qnx,comp.sys.amiga.misc
Subject: Re: Amiga, QNX, Linux and Revolution
Date: Fri, 10 Sep 1999 05:24:13 GMT
On Thu, 09 Sep 1999 18:46:01 -0400, Tony Talarico <[EMAIL PROTECTED]> wrote:
>
>"Mainframes" - i.e. IBM, Hitachi, Amdahl, Fujitsu - don't (normally) use
>directories. Except for IBM's AIX/390, which is a Unix clone.
>
>IIRC, "Directories" came about with the advent of the minicomputer (I
>don't know if DEC's OS or AT&T's Unix was first). It is my belief that
>the reason for directories was so the system administrator could easily
>separate the users' files from other users and from the system files.
I thought Unix inherited its directory structure from Multics.
In any case, directories only become really critical in timesharing
systems. The older batch-oriented systems could function with
much less sophisticated file systems.
Unix and VMS were the earlier widely used timesharing OSes.
Does anybody know if ITS had directories?
--
You'd think that after all this time
I would have dreamed up a really clever .sig!
------------------------------
From: [EMAIL PROTECTED] (Jeff Gentry)
Crossposted-To: comp.os.linux.advocacy
Subject: Re: General Rant from a Linux Newbie
Date: 9 Sep 1999 18:05:25 GMT
K. Bjarnason ([EMAIL PROTECTED]) wrote:
: Silly me; I thought Linux was being touted (by the hypsters, at least)
: as the Windows Killer - the latest and greatest desktop OS for the
: masses, the one that'll wipe MS out (or at least give them some
: competition).
So the way to be better than something is to be exactly the same then?
Is that what you are saying?
That you thought that Linux was more Windows than Windows?
Ever consider that perhaps there is a better way of doing things?
And maybe *that* is why Linux is better than Windows?
--
Jeff Gentry [EMAIL PROTECTED] [EMAIL PROTECTED]
"You're one of those condescending UNIX users! ...."
"Here's a nickel kid ... get yourself a real computer."
------------------------------
From: Miles Bader <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.advocacy
Subject: Re: General Rant from a Linux Newbie
Date: 10 Sep 1999 12:45:17 +0900
K. Bjarnason <[EMAIL PROTECTED]> writes:
> In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] says...
> > You figure it's ok to accuse someone of "spewing" as long as you hang a
> > smiley on the end?
>
> No, I figure it's okay to accuse someone of spewing when they're
> obviously spewing. The smiley just indicates I don't take their spewing
> too seriously.
Ok: You're spewing.
Please stop.
-Miles
--
Love is a snowmobile racing across the tundra. Suddenly it flips over,
pinning you underneath. At night the ice weasels come. --Nietzsche
------------------------------
From: Tony Talarico <[EMAIL PROTECTED]>
Crossposted-To: comp.os.qnx,comp.sys.amiga.misc
Subject: Re: Amiga, QNX, Linux and Revolution
Date: Thu, 09 Sep 1999 18:46:01 -0400
Guy Macon wrote:
< snip >
>
> Directories exist so that software, including the OS can find files.
> Directories were designed the way that they are so that software can
> present a one-to-one representation of the directories (which humans
> are unable to see) on keyboards and display devices. This was an
> important thing to do when 64 users were sharing a 1Mhz mainframe.
> >
"Mainframes" - i.e. IBM, Hitachi, Amdahl, Fujitsu - don't (normally) use
directories. Except for IBM's AIX/390, which is a Unix clone.
IIRC, "Directories" came about with the advent of the minicomputer (I
don't know if DEC's OS or AT&T's Unix was first). It is my belief that
the reason for directories was so the system administrator could easily
separate the users' files from other users and from the system files.
None of the early microcomputer disk operating systems had directories,
and they had no problem finding files. This includes Apple DOS, Atari
Dos, CP/M, HDos (Heathkit :-) ), TRS-Dos, and others. Of course, what
need is there for directories when storage is only 100k and all of it is
removable? ;-)
The first micro OSs that had directories were variations on Unix (Xenix,
Chromix, and whatever Alpha Micro called theirs) which, of course, had
directories.
Again, no specifically "mainframe" OS uses directories. OS/390 doesn't
seem to have any problems finding files on disks that are now
approaching (or may have already reached) a Terabyte in size. That's
1000 Gigabytes.
< more snip >
> Did I ever say that having no human interface was desirable? No. I said
> that the human interface does not have to match the internal details.
> Then I gave an example that you yourself use (newsgroups).
This is absolutely, positively true. It can also be very desirable if
done properly. If done wrong, it can be Windows 3.x (-shudder-).
< snip automotive analogy (sp?) food analogy, and all the rest >
To agree with you, directories are not _needed_. However, to agree with
Joe, under current OSs, they DO exist. Users really should be taught
about them. It amazes me when our support staff at work is called
because a user can't figure out where their document went when they
exited the program.
Of course, even a super computer whiz like me can _occaisionally_ forget
where he left his data, but at least I know enough about the machine to
go look for it without handholding.
--
Tony T
To contact me, take away the SVO
(OVER MY COLD LIFELESS CORPSE!!) ;->
------------------------------
From: [EMAIL PROTECTED] (Christopher B. Browne)
Crossposted-To: comp.os.linux.advocacy
Subject: Re: General Rant from a Linux Newbie
Reply-To: [EMAIL PROTECTED]
Date: Fri, 10 Sep 1999 04:57:16 GMT
On Fri, 10 Sep 1999 05:26:05 GMT, Albert Ulmer <[EMAIL PROTECTED]> posted:
>> >>Ooh! That sounds hard! All I have to do is run dselect and hit "+"
>next to
>> >>stuff I want. It automatically takes lets me know about
>dependencies, so I can
>> >>see that it requires a 40Mb browser before I even think of
>downloading it. The
>> >>I hit install and it even downloads it for me, It then installs and
>configures
>> >>and I don't even have to think. Wouldn't it be nice if M$ Windows
>had
>> >>something as easy to use as that? You would think it would,
>especially as they
>> >>say "ease of use so often".
>
>> > What is dselect?
>> > Sounds like something I would be interested in.
>
>> Isn't it the Debian package manager?
>
>Indeed it is!
Careful!
One might make the mistake of mistaking dselect and rpm as being
"basically the same thing."
They're not.
The view of the world is more like:
Red Hat Debian
===============================================
Package Installer RPM dpkg
Package selector autorpm apt-get
Interactive package
selector install util dselect
or Glint
This is still not really fair to dselect; Glint is a largely
content-free piece of software, whereas dselect is "smart enough"
that it knows how to join together lists of packages from multiple
sources (e.g. - CDROM, local disk, remote filesystem, floppy, ftp,
http, with some options to do further multiplexing...), and does
itself analyze what packages are required by other packages,
so that if, for instance, you select "Netscape Navigator," it will
urge you to install various packages that the Netscape Navigator
package either forcibly requires or recommends.
That being said, dpkg and apt-get both have some ability to do some
of the same things, albeit with a less "interactive" user interface.
--
"Unless you're on the Forbes' richest 100 list, you're not a market,
just another photon in the rainbow." -- Monty Brandenberg
<[EMAIL PROTECTED]>
[EMAIL PROTECTED] <http://www.hex.net/~cbbrowne/lsf.html>
------------------------------
From: [EMAIL PROTECTED] (Eric Y. Chang)
Crossposted-To: comp.os.linux.x,alt.os.linux.caldera,comp.os.linux.setup
Subject: Re: REAL PLAYER in LINUX. Which ver.
Date: 10 Sep 1999 00:02:31 GMT
Reply-To: [EMAIL PROTECTED]
Hi CH. No error message or dialog box. Just no sound. The little GUI
window pops up. The way I discovered it was by clicking on the "Real"
emblem in the GUI. That played sound and a movie, and I saw a URL.
Then I made a wild guess and it worked. Actually, I should have known
better. These little interprocess messages just seem to have a knack
of getting lost, especially under Unix. It's a symptom of careless
progtamming/configuration.
Eric
CHamel ([EMAIL PROTECTED]) wrote:
: >if you shift-1 click on the link, then open it inside RealPlayer as a
: >file, it works just fine. Great sound. But that is a hassle. GUI's
: >should be (but aren't always) point and click.
: I'm curious: what sort of error, if any, did you get when you attempted to
: play a file *before* you discovered SHIFT + 1-click? Would it have been
: anything like Error# 38...?
: ...CH
------------------------------
From: [EMAIL PROTECTED] (B'ichela)
Crossposted-To:
comp.os.linux.admin,comp.os.linux.advocacy,comp.os.linux.hardware,comp.os.linux.help,comp.os.linux.questions
Subject: Re: Q? - best combo of linux distrib and apps for 3rd world
Date: Thu, 9 Sep 1999 14:42:10 -0400
Reply-To: [EMAIL PROTECTED]
On Thu, 09 Sep 1999 07:30:14 GMT, David M. Cook <[EMAIL PROTECTED]> wrote:
>On Wed, 08 Sep 1999 22:00:17 GMT, Allan L. Brown <[EMAIL PROTECTED]> wrote:
>
>>- Reasonable performance on 486-33, 16 MB Ram, <400 MB HD.
>>- All software free.
>>- apps appropriate for NON-Technical end users need to do wordprocessing,
>
>WordPerfect 8 might run OK on that machine with judicious choice of window
>manager (no GNOME or KDE). Note that the free version of WP8 is not the
>full version.
Assuming that the person has a VGA or sVGA monitor and
compatible Video card. at first my 486 used a CGA monitor (it worked
ok for text use.
>
>But frankly I don't think this is realistic hardware (I've never used X on
>anything less than a 486/66 with 32MB RAM.). The above machine would
>be a good X terminal, though.
I have used The Xwindows System on a 486 Dx2/66mhz machine
with 8mb of ram. need I say, it could not run netscape very well and
it crawled! I have Xwindows on my same machine with 20mb of ram and it
still crawls a little.
The X windows system in my situation is not my main interface
however, standard ASCII is.
>
>Linux runs well on 486s, but not, unfortunately, when they are encumbered by
>lots of fat GUI office software.
Thats true. Doesn't anyone have a recomendation for a Word
processor that can be used with ncurses or slang? on a standard
console or a remote terminal? He would need something like that. I
myself use vim and groff so I am not a definitive source for info on
fancy word processors.
>The other issue is HD space. I think you're going to have a hard time
>squeezing what you need into 400MB.
Not if you are careful in choosing the bare basics. Slackware
is good for this. I used and still have my original ver 3.1 disks
here. If bare basics is all you want just install the A,AP and N disk
sets, (if one does not need tcpip. skip the N set.
--
A pearl of wisdom from the y2K newsgroups:
=========================================================================
Y2K appears to be the Baby Boomers mid-life crisis, and it has the
potential to be a dandy.
-- Anonymnous --
==========================================================================
B'ichela
N O T E
---------------------
If [EMAIL PROTECTED] don't work try [EMAIL PROTECTED]
------------------------------
From: Timothy <[EMAIL PROTECTED]>
Subject: gnome-config: command not found...
Date: Fri, 10 Sep 1999 14:55:00 +1000
Hi
I was trying to './configure' pan-0.4.9 (newsreader) when I got this
message. I'm running RH6.... Any ideas on how to fix it? Thanks......
Tim
------------------------------
** 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
******************************
