Linux-Misc Digest #800, Volume #24 Tue, 13 Jun 00 08:13:02 EDT
Contents:
Re: Decompressing corellinux-oc_1.1.iso.gz ("Yehuda Raveh")
No Disk space left !?! (Toni)
how to configure kernel to suitable video card? (jeffrey)
Re: how to configure kernel to suitable video card? (Allen Ashley)
Re: Missing characters in Netscape pages (Paul Colquhoun)
Inkjet printer ([EMAIL PROTECTED])
Mandrake Linux (Koen Van Baelen)
Re: hlt -instruction hangs boot proces (Thomas Zajic)
ext2 filesystem and duplicate blocks (Rui Pedro Mendes Salgueiro)
Re: Sun Sparc faster then intel pentium: is this true???? (Stephen E. Halpin)
Re: Linux helpdesk (H H Chau)
Re: installing soundblaster live! value driver (Mark Howson)
----------------------------------------------------------------------------
From: "Yehuda Raveh" <[EMAIL PROTECTED]>
Subject: Re: Decompressing corellinux-oc_1.1.iso.gz
Date: Tue, 13 Jun 2000 06:08:16 -0400
Winzip does not open iso images. In windows use winimage. in linux there is
a way to mount the file.
"Hafiz-ul-Islam" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Having downloaded the 279mb corellinux-oc_1.1.iso.gz archive from
> download.com three times, using the Download Accelerator program, I have
> been unable to decompress this archive into a 'iso' file for burning onto
a
> CDR. I have tried Winzip which manages to get a third of the way through
> the archive before reporting an 'unsupported format' error. I have also
> tried the Aladdin Decompression utility to no avail. Is it possible that
> all three downloads are corrupt? I would greatly appreciate any light you
> may be able to shed on this subject.
>
> Thank you
>
> Hafiz-ul-Islam
>
> --
> Posted via CNET Help.com
> http://www.help.com/
------------------------------
From: Toni <[EMAIL PROTECTED]>
Subject: No Disk space left !?!
Date: Tue, 13 Jun 2000 12:14:46 +0200
I was just installing some freeS/WAN software RPM on my SUSE 6.4 Linux
box as the console screamed to me that it had to allocate at least 250
kb
for my package on the / partition.
Ok, so I went into midnight commander and erased some 15 Megs in my /tmp
Directory that weren't cleaned up before.
(my tmp dir belongs to the / as well)
Then I started the rpm installation again - with no change !!!
Other rpms i wanted to install followed, and my system is permanently
bugging me with "Maximal mount count reached" messages and checks.
Although mc tells me that i have some 8 to 15 Megs spare on /
my box cannot allocate any space of that to a program...why ??
What Do i have to do (besides repartitioning my only 140 Megs /
partition ??
Thanks in advance,
Toni.
------------------------------
Date: Tue, 13 Jun 2000 18:18:43 -0700
From: jeffrey <[EMAIL PROTECTED]>
Subject: how to configure kernel to suitable video card?
how to configure my kernel to make my video card suitable for my linux?
i am using red hat 6.1 ,and the video card is not suitable
besides this,if install new kernel will solve this problem?
------------------------------
From: [EMAIL PROTECTED] (Allen Ashley)
Subject: Re: how to configure kernel to suitable video card?
Date: 13 Jun 2000 10:31:19 GMT
jeffrey <[EMAIL PROTECTED]> writes:
>how to configure my kernel to make my video card suitable for my linux?
>i am using red hat 6.1 ,and the video card is not suitable
>besides this,if install new kernel will solve this problem?
By "not suitable" do you mean that you cannot configure Xwindows?
If that is the case compiling a new kernel will not solve yoiur
problem. You need an upgrade of XFree86. I know that the latest
versions of Mandrake and SuSE support a number of recent AGP
video cards.
------------------------------
From: [EMAIL PROTECTED] (Paul Colquhoun)
Crossposted-To: comp.os.linux.x
Subject: Re: Missing characters in Netscape pages
Reply-To: <[EMAIL PROTECTED]>
Date: Tue, 13 Jun 2000 10:44:24 GMT
On Mon, 12 Jun 2000 19:50:16 GMT, Christopher Wong <[EMAIL PROTECTED]> wrote:
|In article <[EMAIL PROTECTED]>, Dances
|With Crows wrote:
|>On Mon, 12 Jun 2000 15:27:36 GMT, Christopher Wong
|><<[EMAIL PROTECTED]>> shouted forth into the ether:
|>>When browsing the web with Netscape or kfm, I often come across pages
|>>where some characters do not display in the browser. I would see a '?'
|>>or some other generic placeholder depending on the font I am
|>>using. The undisplayed characters are generally punctuation characters
|>>such as quotes. I usually use Microsoft's TrueType fonts when
|>>browsing. This is probably a well-known problem, so I would appreciate
|>>it if someone could point me to a document explaining the problem and
|>>work-arounds. Thanks in advance.
|>
|>The problem is boneheaded Web designers who use direct character codes
|>like — in their HTML. (ASCII code 151, shows as an em-dash under
|>Lose9x.) You'll find that all the characters that show up as "?" are in
|>the range 128-159, which is not defined in the ISO 8859-1 standard.
|>There probably isn't a workaround for Netscape, but you have the source to
|>kfm and someone could probably kludge something in to display " for the
|>"Stupid Quotes" and -- for the dashes. Submit a feature request to the
|>Mozilla team if no one's done so already.
|
|Ah, so that's it. Still, it would be hard to argue that the designers
|are being "boneheaded": they could just as easily argue that this is a
|deficiency in Linux/X. The HTML standard (I am looking at the HTML
Is that the same HTML spec that defines specific, non-numeric, codes
for these characters ?
Section 24.4.1 of the spec you quote below defines — ( &mdash;
for those with an HTML aware newsreader ) as the code for the em dash
character.
|4.01 spec on w3c.org) specifies ISO10646 aka Unicode as the HTML
|standard for the numerical character references. And of course, the
|ISO10646 standard defines a lot more characters than the ISO8859-1
By-the-way, the ASCII code 151 is not an em dash under ISO10646.
The ISO10646 code for an em dash is — ( again, &#8212; )
Microsoft's "Smart Quotes" *should* use lsquo, rsquo, ldquo, & rdquo
for left/right single/double quotes, or the hex/numeric version for
maximum compatibility.
The hex equivalents are:
Left single : #8216
Right single : #8217
Left double : #8220
Right double : #8221
|standard. Web designers could argue that Linux web browsers are not
|standards compliant. Is there a better argument to present to
|webmasters than "you bonehead, you are not following the standard"?
What is wrong with following the published standards?
The above hex codes work correctly on my Linux system under Netscape,
the text versions don't work.
Both representations work under the (text mode) Lynx browser.
Anybody want to test other browsers?
|I suppose another way to fix this problem is to have a web proxy fix
|up the HTML stream on the fly. This has the effect of modifying the
|content, so I am not sure what the legal issues there are behind
|this. It may be asking too much to have Junkbuster handle this, but it
|would be nice.
Do a web search for "demoronizer" ;-)
--
Reverend Paul Colquhoun, [EMAIL PROTECTED]
Universal Life Church http://andor.dropbear.id.au/~paulcol
-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
xenaphobia: The fear of being beaten to a pulp by
a leather-clad, New Zealand woman.
------------------------------
From: [EMAIL PROTECTED]
Subject: Inkjet printer
Date: Tue, 13 Jun 2000 10:52:28 GMT
I am about to buy an inkjet printer for my Linux system.
Can some user please comment on the following printers and
how well they are supported by current Linux/Ghostscript
versions?
Epson Stylus Color 670
Epson Stylus Color 760
Canon BJC-3000
HP Deskjet 610C
HP Deskjet 840C
HP Deskjet 930C
Thanks for any help
Bernd
--
Bernd Schandl
Tired of spam - use spamcop.net to complain!
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Koen Van Baelen <[EMAIL PROTECTED]>
Subject: Mandrake Linux
Date: Tue, 13 Jun 2000 11:08:02 GMT
Hi guys,
is Mandrake Linux any good? I've heard it's very easy to install and
it's the perfect distribution for newbies.
--
Koen Van Baelen
[EMAIL PROTECTED]
http://users.pandora.be/koen.van.baelen
--
------------------------------
From: [EMAIL PROTECTED] (Thomas Zajic)
Subject: Re: hlt -instruction hangs boot proces
Reply-To: [EMAIL PROTECTED] (Thomas Zajic)
Date: Tue, 13 Jun 2000 11:23:28 GMT
On Tue, 13 Jun 2000 09:15:14 GMT, [och] wrote:
> When booting Linux from disk I get (about) the folllowing msg:
> (left out some dots here and there)
>
> [ ... ]
> Checking 'hlt' instruction...
>
> at this point the computer hangs and I have to reboot
> the machine. ctrl-alt-del does not invoke a reboot.
> [ ... ]
> What are my options?
Try adding the "no-hlt" option at the boot prompt.
HTH,
Thomas
--
=-------------------------------------------------------------------------=
- Thomas "ZlatkO" Zajic <[EMAIL PROTECTED]> Linux-2.2.16/slrn-0.9.6.2+ -
- "It is not easy to cut through a human head with a hacksaw." (M. C.) -
=-------------------------------------------------------------------------=
------------------------------
From: Rui Pedro Mendes Salgueiro <[EMAIL PROTECTED]>
Subject: ext2 filesystem and duplicate blocks
Date: 13 Jun 2000 12:45:54 +0100
Hi.
I have here a number of PCs running Linux and with a batch of them
I have two problems:
- they freeze without an obvious reason (and without error messages);
- when they freeze they frequently lose/damage files, which means that
shortly I will have to reinstall the whole lot :(
About the first problem I don't expect much help because I can't give
much information: the computers are Athlons 550-600 MHz, with FIC SD-11
boards, Riva TNT2 graphic cards, and Netgear FA-310Tx network cards.
The software is Suse 6.3 (kernel 2.2.13), XFree86 Version 3.3.5 (with
glx) and a netgear-specific driver (included in their floppy disk).
Curiously the crashes seem to happen more frequently if they are
running seti@home.
I suspect the memory, because most of them have 2 different DIMMs
and Athlon boards are known to be picky about which memory they
run with. But Memtest-86 didn't find any problem.
About the second problem, I have read that the way Linux (I assume
it is meant ext2) handles disk buffers/cache is a bit more fragile
than the Berkeley FFS (which I am more used to), because more info
(metadata) is kept in memory. But anyway I find the following
problem a bit strange: sometimes the e2fsck complains that there
are duplicate blocks between the metadata and some file, and so it
corrupts or deletes the file to solve the problem.
The strange part is in some cases, the file is an old file which
I know has not been touched (in the sense of written) for months.
What I find strange is how the filesystem allocated that file's
blocks to the metadata. It seems to me that a "normal" crash could
leave duplicated blocks between a deleted file and a newly-created
file like this:
- a file is deleted, but that info is not yet written to the disk;
- a new file is created using the same blocks. This information
is written to the disk;
- then the system crashes leaving the block allocated to both files.
but I don't see how something similar can happen with "cold" files
(in one of the cases, it is not even likely that some process was
trying to read the file which would force an updated of the access
time in the inode).
So my question is: is this filesystem problem usual without an
hardware problem causing the corruption ? Because if it is not
usual, then the possibility that the memory is not working 100%
reliably gets more likely.
Thanks in advance
--
http://www.mat.uc.pt/~rps/f1/ an half-tifoso until Canada 2000
Mark Sandman - Morphine, RIP (1952-1999/07/03, Italy)
.pt is Portugal| `Whom the gods love die young'-Menander (342-292 BC)
Europe | Villeneuve 50-82, Toivonen 56-86, Senna 60-94
------------------------------
From: [EMAIL PROTECTED] (Stephen E. Halpin)
Crossposted-To: comp.sys.sun.hardware,comp.sys.sun.misc
Subject: Re: Sun Sparc faster then intel pentium: is this true????
Date: Tue, 13 Jun 2000 11:51:43 GMT
On Mon, 12 Jun 2000 22:24:10 -0500, Dave Schanen <[EMAIL PROTECTED]> wrote:
>"Stephen E. Halpin" wrote:
>
>> Clue: You can buy Intels tools TODAY through Intels web site. The C/C++
>> compiler (version 4.5, as used in the SPECxxx2000 tests) set are $299
>> list, and the VTune analysis package is $429. Add in Microsofts
>> Visual C++ environment for a list price of $549, and you have a full
>> suite of tools TODAY for $1,277, which is quite a bit cheaper than the
>> $1,995 you will have to pay to get Suns Forte C++ Personal Edition 6.
>So once you buy this compiler from Intel, what are you going to do,
>re-compile nt? Obviously Microsoft didn't use intel's latest compiler
>to build their bloated OS, and it will be at least 6 months before we
>even see any of these features in DirectX. Most applications don't even
>take advantage of MMX, much less the newest thing from intel. (Find an
>application which claims to utilize the v4.5 intel compiler) No matter
>how great the compiler optimization used on windows is, poorly written
>code is still poorly witten code.
The topic is SPARC vs Intel, and its been demonstrated that for a
specific set of tasks that the Intel chips can easily be more than
twice as fast as the UltraSPARCs when running the same code. Just
as there is code compiled nonoptimally for various Intel platforms,
there is also code which has not been optimized for V9 of the SPARC
architecture, and which does not take advantage of the VIS instruction
set. Plenty of folks use GCC when it doesnt necessarily produce code
that is competative with the various RISC vendors compilers. If you
are deploying a custom system where your hardware costs for an Intel
solution is $10M and your equivalent Sun hardware still costs $20M
after discount, you can afford to do a lot of code optimization.
Also, when Im writing CPU bound applications, the time is spent in
my code, not the kernel, so what little NT does in the background
isnt of much concern.
>> As for the SPEC numbers, there are reporting rules are quite clear
>> about not publishing numbers with research only compilers, and about
>> reporting availability, with the intent that anyone should be able
>> to buy the test bed as described and reproduce the benchmark results
>> within small tolerances. If anything, the greater difference in base
>> and non-base numbers on RISC parts is partly due to the allowance of
>> unsafe optimizations in the non-base tests, indicating that the RISC
>> vendors require more specialized, platform specific and unsafe
>> features for their compilers to produce fast code.
>Yes, yes software benchmarks are all very fair, thats why the company
>with the most money always wins.
You obviously werent watching DECs stock drop from $199 to $17 during
the development and initial releases of the Alpha AXP processors and
systems. Even with far better performance, they are having trouble
in the market because they dont have the application support they need.
>> Its all there on Intels, Microsofts and SPECs web sites for you to read.
>> If you dispute the numbers, produce the results you come up with using
>> the tools and hardware listed in the SPEC reports. Otherwise, the only
>> "bullshit" is coming from you.
>Don't get all huffy, it's just usenet.
To requote you:
>This is getting getting ridiculous... Less than a month ago 'Boris' was
>in here quoting the same bullshit benchmarks you were.
You have yet to offer any evidence that the benchmarks dont give any
indication of the CPU performance that can be extracted from these
processors by a developer who takes the time to properly optimize
their code.
>Dave-o
-Steve
------------------------------
Crossposted-To: alt.os.linux,alt.linux
From: [EMAIL PROTECTED] (H H Chau)
Subject: Re: Linux helpdesk
Date: Tue, 13 Jun 2000 12:20:25 +0100 (BST)
In alt.os.linux Steven Feil <[EMAIL PROTECTED]> wrote:
> I'm thinking about starting up a pay-per-call Linux help desk using a
> 1-900 number. I would use a web-page and e-mail to make the initial
> contact with potential customers. I would use the e-mail to evaluate
[...]
> Since I am not an expert on everything Linux, I would start out with
I assume you will specify what you can and cannot help on your web
pages. Will you charge for e-mail evaluation too?
> There are many advantage of a helpdesk over traditional ways people
> get Linux help. Problem could be fixed in a timely manner instead of
People who think time is important [for their business/work] should have
hired someone in the first place, if they can't handle themselves.
> the question and answer session drag on for days in Usenet. A helpdesk
> would be better than IRC, because with IRC you tend to get pot-luck,
> and then your competing with others for attention. When someone calls
> my helpdesk they will be getting my undivided attention. When getting
But here at aol, we have a wide range of expertise in additional to
NFS/NIS.
> IRC help it would be unlikely that you would find someone who is
> patient enough to go through the process of setting up an NFS server
^^^^^^^^^^^^^^
Everyone who are paid to do something are patient, like it or not.
> and client step. In IRC the best they could expect is a few tips and
> a pointer to the NFS HOW-TO.
The only help I can think of unable to get from HOWTOs is a patch
for bug fix. You offer this too?
> I guess it all comes down to a matter of price, and how much people
> are willing to pay. I'm thinking I could to a simple NFS or NIS
> installation for about $40-$60 over the phone.
People who rely on NFS/NIS to their network probably have quite a
large number of computers, and have a local expert (if not a dedicated
system administrator). Couples of ten is nothing to them. However,
NFS/NIS is just one of the many issues of system administration, I
wonder how to fit yourselves in. To me, I rather have fun to play
around it, provided large number of good books at different levels,
and HOWTOs.
> Do you think this is a good idea?
I wish you good luck.
> Do you think if I were to make announcements to this news group that I
> was spamming?
comp.os.linux.announce might reach more potential customers.
Cheers
Hau Hing
------------------------------
From: Mark Howson <[EMAIL PROTECTED]>
Subject: Re: installing soundblaster live! value driver
Date: Tue, 13 Jun 2000 13:00:08 +0100
Bbitzer wrote:
> my "modutils" is installed and should work. below is a copy of the script
> durring my last installation attempt. if anyone can see where I went wrong
> or has a better idea, I really, really would appreciate it before I throw
> this thing out the window.Again, please note that I have installed Red Hat
> Linux 6.2 deluxe, kernel 2.2.14-5, and am trying to get my SoundBlaster
>
> [root@localhost bbitzer]# find / -name depmod -print
>
> find: /proc/6/fd: Permission denied
>
> /sbin/depmod
Heh. You've just answered your own question...
BTW, it wasn't really necessary to include /all/ the information you
did. More information is always good, but you could safely have cut out
the download and compilation stages (which seemed to work fine) :)
> [root@localhost misc]# cp /home/bbitzer/sblive/emu10k1/emu10k1.o ./
> cp: overwrite `./emu10k1.o'? y
> [root@localhost misc]# depmod -a
> bash: depmod: command not found
Try /sbin/depmod here.
Mark
------------------------------
** 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
******************************
