Linux-Misc Digest #814, Volume #20 Sun, 27 Jun 99 21:13:08 EDT
Contents:
Re: Compiling kernel 2.2.10: file asm/mman.h missing (Cokey de Percin)
Re: Docbook? Linuxdoc? Re: Documentation issues. (Leslie Mikesell)
Re: Docbook? Linuxdoc? Re: Documentation issues. (Jochem Huhmann)
rvplayer spooky story (Paul S. Aspinwall)
Looking for Linux audio mixer Package ("Bunbi")
Linux Reading Partitions ("Jody Thigpen")
Problems with internal LS-120 drive (Aaron Ginn)
modem (Quoc Nguyen)
Re: linux and tapes (Robert Heller)
Re: Docbook? Linuxdoc? Re: Documentation issues. (Cameron L. Spitzer)
----------------------------------------------------------------------------
From: Cokey de Percin <[EMAIL PROTECTED]>
Subject: Re: Compiling kernel 2.2.10: file asm/mman.h missing
Date: Mon, 28 Jun 1999 00:09:59 +0000
Jozef Dodziuk wrote:
>
> I have been running kernels 2.2.x applying patches and compiling
> successfully up to 2.2.9. When I tried upgrading to 2.2.10, compilationmake
> bombed out with a complaint that the file asm/mman.h was missing.
> I downloaded the full 2.2.10 source, installed it and tried to recompile
> with the same result.
>
> What am I doing wrong? Any suggestions will be appreciated.
>
Not sure what your problem is, but I just downloaded 2.2.10 from us.kernel
and it compiles just fine... Make sure that the link from
/usr/src/linux/include/asm-i386 to asm exits.
Best
Cokey
--
==================================================================
Cokey de Percin, DBA Email:
Policy Management Systems Corp. Work - [EMAIL PROTECTED]
Columbia, South Carolina Home - [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: gnu.misc.discuss
Subject: Re: Docbook? Linuxdoc? Re: Documentation issues.
Date: 27 Jun 1999 17:58:57 -0500
In article <[EMAIL PROTECTED]>,
Cameron L. Spitzer <[EMAIL PROTECTED]> wrote:
>>
>>> Had it [locate] been absent instead of broken, I would have kept
>>> poking around with find(1), which is slower than locate(1), but
>>> works.
>>
>>*shrug* locate works for me and has done for years; I've no idea why
>>it doesn't work for you.
>
>"It works for me" does not mean "it works." It means it is possible to
>make it work under some unknown conditions.
Most likely problem is that you don't have enough disk space to
sort the index. Have you tried running the updatedb script
manually to see if you get an error message?
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Jochem Huhmann)
Crossposted-To: gnu.misc.discuss,comp.unix.bsd.misc,comp.text.sgml
Subject: Re: Docbook? Linuxdoc? Re: Documentation issues.
Date: 27 Jun 1999 23:48:33 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Christopher B. Browne) writes:
> The difference between SGML and XML is mainly that it's easier to write a
> parser.
>
> Once the document gets turned into a tree of nodes, what you do with it
> is the SAME whether you're using SGML or XML.
But there are a lot of lightweight parsers for XML. That *does* make a
difference. There are fine and simple XML-parsers for usage with
almost every programming language out there. If I have some XML and
would like to convert it to HTML or LaTeX, I can choose between doing
it in Java, C, perl, python, tcl - just what I'm most familiar with.
> The real point is that XML is not particularly "more accessible" than
> full SGML, and certainly not from the perspective of anyone that isn't
> writing a parser or developing using a parser.
>
> [Another "real difference" is that if you are writing an application that
> needs to work with structured data, it's easier to integrate in an XML
> parser. That is *almost* the same issue.]
Which makes a huge difference in reality. The available range of
parsers and applications is wider for XML as for SGML. And because
they are simpler to write, they are cheaper and more common.
To get back at sgmltools: It's funny enough, that it's entirely based
on jade and DSSSL. If someone wants to write some other backend (or
improve one), he has to know not only DocBook, but also jade, C++ and
DSSSL (and the output-format of course). This is a lot to know
about. Nobody is really satisfied with the print output of jadetex,
but nobody is inclined to dig into it. It's way too difficult to
attract developers. If I would have to produce high-quality printouts
from DocBook, I would use DocBook-XML and map it to high-level LaTeX
with some straight perl or tcl. Sgmltools used to be *the* tool for
the Linux Documentation Project, because it produced nice HTML, ASCII
and LaTex output. Nearly nobody switched to the new DocBook-based
Sgmltools, because it is hard to install, slow and produces output of
very limited quality. The Postscript produced by jadetex is a
typographic nightmare. A lot of people have switched back to writing
LaTeX in the first place and converting this to HTML, if they want
both printed and online docs.
Jochem
--
Hi! I'm a .signature virus! Copy me into your ~/.signature to help me spread!
------------------------------
From: [EMAIL PROTECTED] (Paul S. Aspinwall)
Subject: rvplayer spooky story
Date: 28 Jun 1999 00:18:59 GMT
rvplayer5.0 was working perfectly for me on my RedHat 5.1 system with
Ultrasound card.
Then I moved house and plugged my computer back in.
Now when rvplayer is lauched from netscape it gets stuck at the the
"Buffering 16.0 Kbps clip" point and refuses to go any further. (It will
play the welcome file directly though.)
Anyone care to explain why moving house should have this effect???
--
Paul S. Aspinwall
Duke University
------------------------------
From: "Bunbi" <[EMAIL PROTECTED]>
Subject: Looking for Linux audio mixer Package
Date: Sun, 27 Jun 1999 23:46:08 -0400
Reply-To: "Bunbi" <[EMAIL PROTECTED]>
I would like to setup my Linux machine to do audio mixer then burn it into
the CDROM. What I am looking for is a professional level audio mixer
software package available freely under Linux. I've checked with the Linux
software repository site:
http://www.linux.org/apps/applications.html
and found a broken link to the "SoundStudio" site. Then, I went to metalab
site and found a huge audio mixer under
/pub/Linux/apps/sound/mixers/slab-2.20.1.tar.gz. Being a newbie in this
audio mixer field, I was amazed by the software package's size and am
interested to know if this audio mixer package is better than the Sound
Studio.
If you know of other better audio mixer that is freely available, I sure
would like to hear from your comment and pointer.
--
[EMAIL PROTECTED]
------------------------------
From: "Jody Thigpen" <[EMAIL PROTECTED]>
Subject: Linux Reading Partitions
Date: Sun, 27 Jun 1999 19:06:14 -0500
Reply-To: "Jody Thigpen" <[EMAIL PROTECTED]>
I'm running a multi boot system between Win 95, NT and Linux Slackware 3.4.
The trouble I'm having is that my Linux partitions are in the middle of my
disk, and Linux reads them at the end. As a result, Partition Magic won't
read any of the drive. Does anyone know how I can get Linux to read the
drives in the order they are, or am I going to have to fdisk and re-install?
Thanks,
Jody Thigpen
[EMAIL PROTECTED]
------------------------------
From: Aaron Ginn <[EMAIL PROTECTED]>
Subject: Problems with internal LS-120 drive
Date: 28 Jun 1999 02:21:02 +0200
I'm having real problems getting this drive to work at all under Linux. I hope someone
out there can give me some direction about where I should go next. I posted about a
month
ago saying that I couldn't get this drive to work under Linux, and I got several
suggestions
including the following:
1) Change the jumper settings on the drive so that it became the master on the
secondary drive,
2) Check the settings in my fstab file,
3) Update my BIOS.
I finally got around to changing the jumper settings to make the CD-ROM the slave and
the
LS-120 the master, and that didn't solve the problem. Here is the message I get during
boot for this drive:
=====================================
hdc: �, ATAPI UNKNOWN (type 31) drive
=====================================
I've checked the settings in my fstab, and I'm certain they are correct. Finally,
someone had
written that they had seen a similar boot-time message for this drive, and the
solution was a
BIOS upgrade. My question is that Windows 98 is able to see and use this drive fine.
If it
wa a BIOS issue, wouldn't Win98 have the same problem as Linux? I just bought the
computer
about 4 months ago, so the BIOS can't be that old, can they?
When I try to mount a VFAT formatted disk, here is my error message:
============================================
mount: /dev/hdc is not a valid block device
============================================
If I try to fdisk an unformatted disk, here's my error message:
=======================
Unable to open /dev/hdc
=======================
I hope someone out there can help me. I've got sound, an internet connection, and
X-windows all working; this has been by far the biggest headache. This really
shouldn't
be that difficult.
Thanks a lot!
Aaron Ginn
------------------------------
From: [EMAIL PROTECTED] (Quoc Nguyen)
Subject: modem
Date: 28 Jun 1999 00:33:42 GMT
Hi everyone,
It seems that everyone in this newsgroup has problem with there modems. I
am one myself too. Can anyone show me how to config my modem. I have a
33.6 diamond multimedia, 14.4 usb, 56k flex. my slackware 3.4 kernel
2.0.30 detects my modems okay (dev/modem -> /dev/cua1) but they don't
seem to work when I used ppp-go to dial. there wasn't any dial tone or
anything. Also, I'm trying to learn how to write Motif Application on my
Xwindows I don't know how to compile my source code. Do I need a c89
compiler? where can I find that? thanks in advance.
--
Regard,
Quoc Nguyen
Computer Science
******************************************************************
* This method is to define as the member of a class the class of *
* all classes similar to the given class. *
* Bertrand Russell *
******************************************************************
------------------------------
From: Robert Heller <[EMAIL PROTECTED]>
Subject: Re: linux and tapes
Date: Mon, 28 Jun 1999 00:05:23 GMT
Ralf Bechtel <[EMAIL PROTECTED]>,
In a message on Sun, 27 Jun 1999 19:14:11 +0200, wrote :
RB> Hi everybody !
RB>
RB> At the moment I'm doing some development for a backup application under
RB> Unix.
RB> Now I encountered the problem to deal with tapes under Unix. Since I
RB> never
RB> used a tape under linux I've no experience at all how to do that.
Tapes under Linux are just like tapes under UNIX in general.
RB>
RB> What I need is the following:
RB>
RB> (1) How can I read/write ordinary text files from/to the tape? Do I have
RB> to use the device name (/dev/tape) or what?
RB>
RB> (2) How Do I know the filename of the next file following on tape?
RB>
RB> (3) What will happen, if I read beyond the eof of a file? Will I arrive
RB> automatically at the start of the next file?
RB>
RB> Would be very happy if there's somebody out who can help me. Thanks.
RB>
RB> Ralf
UNIX does not impose a 'file system' on 'block sequential' devices
(commonly known as tapes). That is, most common tape I/O operations
treat the tape as a single un-named file (but see below). Programs
like tar and cpio create an 'archive' file (in much the same way as
PKZIP does). This 'archive' file implements named chunks which
correspond to the original files.
So, to copy a file to tape, one can do:
% dd if=some.file.name of=/dev/tape bs=1024
This will write the contents of the file named 'some.file.name' to the
tape. There will be no record of the original name, protections, dates,
etc.
Yes, you are saying, 'Yuck'. This is not really very useful as a
general backup methodology.
Tapes are actually 'structured' in a very primitive way. Data is
written to tapes in 'blocks'. Old reel-to-reel tapes use 512 or 1024
byte blocks. Modern high-capacity streaming tapes using larger blocking
factors. In addition to data blocks, it is possible to record 'record
marks' and 'file marks'. These are special 'blank' areas and are used
to separate chunks of data. It thus possible to record multiple 'files'
on a tape. This brings us to the next level of tape hackery: ANSI
Labeled Tapes.
Back in the days of the dinosaur computers (like the ones people like
Dr. No and the like used to control their doomsday machines (which 007
eventually destroys, just before skipping out with the blonde)), banks,
insurance companies, and the government/military would store information
on big 1/2" reel-to-reel tapes. These were 12" in diameter, with 1/2"
tape, about 2400' and they would use a massive 1600bpi, which yields a
raw tape capacity of 43meg (yep, about 1/2 of a 100 meg Zip cart, or a
40pack of floppies). Actually, with a blocking factor of 1024bytes and
3/4" of tape between blocks, the actual amount of data was somewhat
less. Anyway, there needed to be a way of keeping track of what the
data was. So they invented this simple labeling technology.
ANSI Tape Labels are little 80 byte 'records' and 'files' on the tape.
These little data blocks are used to contain various bits of
information: name of the tape, names, owner, type, protection, blocking
factor, dates, etc. of files, whether the data spans multiple tapes and
so on.
The full scoop on ANSI Tape Labels is described in the ANSI documents:
o American National Standard Institute magnetic tape
labels and file structure for information interchange
ANSI X3.27-1978
o The document/working paper: Draft Proposed Revision
to ANSI X3.27-1978, Public Review Comment on ANSI
X3L5/83-28T 15-Oct-84 (describes the version 4 ANSI
standard)
(Snarfed from man pages for ltf on a DEC UNIX system. Don't know if
these docs are available on the web anywhere.)
So, you have three options:
Write an archive-like system (like tar or cpio).
Have lots of fun with ANSI Tape Labels. No, Linux does not have
any libraries or utilities for this. DEC's Ultrix and DEC UNIX systems
have a program named 'ltf' which gives a simple interface to labeled
tape. *I* wrote a simple program which implemented ANSI Labeled Tapes.
Invent your own hackery.
The *simplest* and most portable is to write a program that works like
tar or cpio -- a program that can write a serialized (flat file) archive.
--
\/
Robert Heller ||InterNet: [EMAIL PROTECTED]
http://vis-www.cs.umass.edu/~heller || [EMAIL PROTECTED]
http://www.deepsoft.com /\FidoNet: 1:321/153
------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Crossposted-To: gnu.misc.discuss
Subject: Re: Docbook? Linuxdoc? Re: Documentation issues.
Date: 28 Jun 1999 00:51:09 GMT
In article <7l6abh$si$[EMAIL PROTECTED]>, Leslie Mikesell wrote:
>In article <[EMAIL PROTECTED]>,
>Cameron L. Spitzer <[EMAIL PROTECTED]> wrote:
[locate(1) finds nothing]
>>make it work under some unknown conditions.
>
>Most likely problem is that you don't have enough disk space to
>sort the index. Have you tried running the updatedb script
>manually to see if you get an error message?
>
> Les Mikesell
> [EMAIL PROTECTED]
It returns 0 and prints nothing, quickly. The file
/var/lib/locate/locatedb it creates contains " LOCATE02 " with no '\n',
ten bytes. The /etc/updatedb.conf contains comments and this:
PRUNEFS="NFS nfs afs proc smbfs autofs auto"
export PRUNEFS
PRUNEPATHS="/tmp /usr/tmp /var/tmp /afs /amd /alex"
export PRUNEPATHS
NETPATHS=""
export NETPATHS
Cameron
------------------------------
** 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
******************************
