Linux-Misc Digest #282, Volume #21 Wed, 4 Aug 99 04:13:11 EDT
Contents:
Abit BP6 + ultra ATA 66 + RIVA TNT2 + SBlive (zitang)
Re: chmod: 04000 and 02000 in directories? (M van Oosterhout)
Re: [Q] linux license? (Christopher Browne)
can't find module st, also 9Gig SCSI disks? OK? (Lyndon F. Bartels)
Re: IDE vs scsi? (William Burrow)
Re: Distributions ("Noah Roberts (jik-)")
Re: IP:Always Defragment (required fro masquerading) ([EMAIL PROTECTED])
Re: math.h problem (Steve Summit)
Re: Large HD Access Problem ("Lord Byron")
Re: Need good sites for unix/linux (Steen Suder)
Need good sites for unix/linux (Kelvin Tsang)
Re: Buy vs Download Difference: Documentation Only? (Philipp Maier)
ICQ and linux client ("Larry Clark")
Clustering & Emulation (Dionysus)
Re: Distributions (Philipp Maier)
Re: Kppp/pppd probs (Frederick Paul Suedmeyer)
Re: Java makes Netscape crash (David Frantz)
----------------------------------------------------------------------------
From: zitang <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware,comp.os.linux.x
Subject: Abit BP6 + ultra ATA 66 + RIVA TNT2 + SBlive
Date: 4 Aug 1999 03:30:35 GMT
anyone try to install linux on this board ?
my system ...
celeron 400
Micron 128MB
SBLive Value
PCI NE2000
Monitor 15' Mag DX15T
-> Creative RIVA TNT2 Value 16MB
-> IBM 13.5 GB 7200rpm ATA66
-> Fujisu 2.5 GB 5400rpm Mode 4
when i connect the IBM hardisk on the ATA66 IDE ...
when i try to install Redhat 6.0
i got Error in the FDISK
(Fujisu it detect - Cyc.. 1024 Head...128 Sec...63)
after i disconnect the IBM ... then it can detect it "fine"(better) ...
(Fujisu it detect - Cyc.. 4952 Head...16 Sec...63)
i can install it but i cant boot it using the LILO ...
only can boot by Floopy ...
finally all done ...
but the X window got problem ...
so i think i really need to install XFree86 3.3.4 maybe 3.9.15 ...
about the SBlive ...
is it only can work on Kernel 2.2.5-15 ?
if i patch the kernel to 2.3.12 then should it possible to work on it ?
coz i need to try kernel 2.3.12 to test the Ultra ATA66 to be work ...
so anyone got same question like me ?
zitang - ^Devil Evil^ -
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
Date: Wed, 04 Aug 1999 13:47:18 +1000
From: M van Oosterhout <[EMAIL PROTECTED]>
Subject: Re: chmod: 04000 and 02000 in directories?
Otavio Exel wrote:
>
> hello all,
>
> looks like a FAQ but I could not find it: what does the 04000 and 02000
> permission bits mean? all I read (including 'man 2 chmod') is that..
>
> S_ISUID 04000 set user ID on execution
> S_ISGID 02000 set group ID on execution
>
> ..but I assume this applies to executable files, not directories;
These bits do have an effect on directories. Namely, if the setuid
bit is set, when a file is created in that directory it will
automatically
become owned by the owner of the directory. Similarly for setgid.
HTH,
Martijn van Oosterhout
Australia
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Subject: Re: [Q] linux license?
Reply-To: [EMAIL PROTECTED]
Date: Wed, 04 Aug 1999 04:16:41 GMT
On 2 Aug 1999 21:56:21 GMT, student <[EMAIL PROTECTED]>
wrote:
>I have some administrative questions.
>As a end user, I've been wondering the question related to
>the license of linux.
>Specifically, for example, many linux OSes
>such as RedHat, Slackware, Debian, etc
>have 'free' or 'share' packages of gnu society
>for compilers, editors, xterms, etc, in their CDROM.
>In that case, the company of RedHat or Slackware or Debian
>need to pay some license fee to 'gnu' society?
>(Or to any person that made free/share packages called
>freeware/shareware).
>Or they don't need to pay to them at all.
>
>If they do not need to pay at all, can I(as an end user)
>sell my own CDROM that has such freeware/shareware packages
>without violating any law?
>Any replies would be appreciated.
The Linux kernel, as well as a vast quantity of the other stuff
included with "Linux systems," represents Free Software.
A particular aspect of the Freeness of the software is that it is free
of licensing charges and other similar committments.
When a group got together and came up with the (sometimes
controversial) term "Open Source" as a moniker to apply to such
software, one of the alternatives that was rejected was that of
"Freely Redistributable Software."
That phrase is nicely descriptive of what is true of Linux and similar
software.
--
To iterate is human; to recurse, divine.
[EMAIL PROTECTED] <http://www.ntlug.org/~cbbrowne/lsf.html>
------------------------------
From: [EMAIL PROTECTED] (Lyndon F. Bartels)
Crossposted-To: comp.os.linux.setup,comp.os.linux.hardware
Subject: can't find module st, also 9Gig SCSI disks? OK?
Date: 4 Aug 1999 04:14:42 GMT
Hello,
I have the following hardware/software setup.
Intel P-II 350 CPU.
192 Meg Ram.
S3 Virge video card
3C509 ethernet card
Adaptec 2940UW controller.
2 4.5G hard drives.
Adaptec 2910 SCSI controller
SCSI CD-ROM
Exabyte 8200 tape drive.
RedHat 6.0 Linux. As yet, with no updates.
2.2-5-15 kernel.
When I boot the system, I get the following error message.
Enabling Swap Space [OK]
can't locate module st
INIT: Entering run level 3 yadda yadda yadda.
I'm assuming this has something to do with the tape drive? Any obviously
simple answers?
Also, I'm planning on adding 2 9G drives to this system for data storage.
I have the OS and httpd server setup on the existing 4.5s. I'm merely adding
the 9Giggers. I read somewhere that either Adaptec doesn't like drives larger
that 8Gig, or that Linux doesn't. What's the skinny? Am I doomed?
Thanks in advance,
Lyndon
------------------------------
From: [EMAIL PROTECTED] (William Burrow)
Subject: Re: IDE vs scsi?
Date: 4 Aug 1999 04:58:05 GMT
Reply-To: [EMAIL PROTECTED]
On Tue, 03 Aug 1999 17:28:35 GMT,
Rod Smith <[EMAIL PROTECTED]> wrote:
>> 35 sectors/track * 512 b/cyl * 10,000 RPM = 179,200,000 bytes/minute/track
>
>Since the vast majority of hard drives can only read from one track at a
>time (per head, anyhow), the two statements are identical in practice. I
>just didn't bother to make this explicit because it seemed so obvious.
>As I stated in my original post, adding heads (platters) is the real
>issue that'll get higher transfer rates.
Yes, and drives can come with multiple platters within the unit. In
that case, a cylinder is more than one track.
Q: How can you tell if a unit has more than one platter? I don't seem
to see this datum being tossed around -- the CHS numbers are totally
fake, so are meaningless for determining how many platters in the
unit. Can one assume that the larger drives must have more than one
platter?
--
William Burrow -- New Brunswick, Canada o
Copyright 1999 William Burrow ~ /\
~ ()>()
------------------------------
Subject: Re: Distributions
From: "Noah Roberts (jik-)" <[EMAIL PROTECTED]>
Date: 03 Aug 1999 21:16:21 -0700
Latenar <[EMAIL PROTECTED]> writes:
> On Wed, 4 Aug 1999, eze wrote:
> > which is the most easy to install linux's distribution?
>
> I installed Slackware (4.0) on a laptop with pretty standard hardware
> (toshiba) without a hitch...took me 4 hours from when i put the cd in to
> when i had a running, functional installed system on my hardrive...add
> another 2 hours for X :-p. note that the figure above includes
> reformatting my entire (1gig) harddrive and also i picked the install
> method that asked me about all non-critical packages (you could cut it
> down to an hour or so otherwise).
Takes me 20 minutes to install slackware just the way I want it.
Using the novice install procedure is a lengthy process, if you know
what type of software you want, then the menu system is a lot faster
then any of the other distros when it comes to a custom setup.
------------------------------
Subject: Re: IP:Always Defragment (required fro masquerading)
From: [EMAIL PROTECTED]
Date: Wed, 04 Aug 1999 05:23:14 GMT
In article <[EMAIL PROTECTED]>, mrmanny <[EMAIL PROTECTED]> wrote:
>When I was configuring my 2.2.10 kernel I came upon this option.
>I sometimes use my machine as a ip masquerade proxy for my cable modem
>and
>sometimes use it as a normal host. While it says this is required for ip
>masq
>The help also says Never say Y for a normal host.
>
>Do I need to say yes or no? I selected no but have not tried the masq
>feature yet since
>the kernel re-compile.
I just checked the source documentation and what it actually says is:
> If you say Y here, then all incoming fragments (parts of IP packets
> that arose when some host between origin and destination decided
> that the packets were too large and cut them into pieces) will be
> reassembled (defragmented) before being processed, even if they are
> about to be forwarded.
>
> You must say Y here if you want to enable "IP: masquerading" or "IP:
> transparent proxying".
>
> When using "IP: firewalling" support, you might also want to say Y
> here, to have a more reliable firewall (otherwise second and further
> fragments must be dealt with by the firewall, which can be tricky).
>
> Only say Y here if running either a firewall that is the sole link
> to your network or a transparent proxy; never ever say Y here for a
^^^^^^^^^^^^^^^^^^^^^^^^^^^
> normal router or host.
^^^^^^^^^^^^^^^^^^^^^^
I think the statement about never saying Y for a host is overkill. I
believe what they are trying to prevent is a having a Linux based
router not forward fragments to a final destination when it doesn't
have all of the fragments. The problem is that if there are multiple
paths to the final destination the fragments might split between
them. Thus they would never be sent. If the machine in question is
an endpoint (host), I don't think it matters. Or at least this is my
best guess. Hopefully someone will correct me if I'm wrong.
Bill Bogstad
------------------------------
From: [EMAIL PROTECTED] (Steve Summit)
Crossposted-To: comp.lang.c,gnu.gcc.help
Subject: Re: math.h problem
Date: 2 Aug 1999 23:54:47 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Kaz Kylheku) wrote:
> On 27 Jul 1999 17:03:34 GMT, Steve Summit <[EMAIL PROTECTED]> wrote:
>> It's time for my semiannual plea to the GNU C and Linux
>> development communities: *PLEASE* bite the bullet and put the
>> math functions into libc, already!
>
> I disagree. The GNU C and Linux communities didn't invent this convention.
It's not a "convention". Kindly put it's a "historical quirk";
today I'd call it a bug, pure and simple.
> Whereas it well within the glibc maintainers' grasps to trivially
> fix this problem, I don't think that they should.
It may not be so trivial, after all. Sure, it's a 1- or 2-line
change to a Makefile somewhere, but first we have to agree that
it's the right thing to do, and based on the way this thread has
proceeded I'm beginning to fear that agreement will be impossible.
> The -lm kludge is just another way in which Linux is compatible
> with other UNIX operating systems.
Bug-for-bug compatibility is the worst kind, usually practiced by
the likes of IBM and Microsoft. Unix is, or used to be, a fluid
system, unafraid to experiment with new ideas or to abandon old
ones that didn't work or weren't needed any more. (How many
people remember multiplexed files?) It seems, though, that some
of its aficionados revere it a bit too highly (I'm not saying you
do this, Kaz), and insist on a state of preservation (of every
feature and misfeature) that's dangerously close to ossification.
A good number of the arguments in favor of preserving the
relegation of libm.a sound a bit to me -- and I suspect to many
of the beginners around here -- too much like "I had to walk ten
miles barefoot in the snow to get the math functions when I was
learning Unix, and you pampered youngsters don't deserve to have
it any easier." It seems as if the strongest reason to leave
the math functions stranded in libm.a may be nothing more than
"because we've always done it that way", and the reason not to
promote them to libc.a is "because we've never done it that way."
> The -lm means that Makefile rules from other platforms work
> on Linux.
It's trivially easy to arrange that -lm continue to work, and I
was never suggesting breaking it.
> And conversely, Linux programmers are forced to adopt a convention
> that will work on many other UNIX-like operating systems, so that
> it's one less detail to worry about when porting.
That sounds like a nice (or nicely hardnosed) sentiment at first,
but there's no basis for it. The GNU and Linux communities have
(to their credit) already fixed countless traditional bugs in
Unix, some far more significant (in terms of "compatibility")
that libc's prejudice against the math functions. Part of the
GNU Coding Guidelines is the recommendation to fix bugs, remove
limitations and add extensions when reimplementing standard
utilities. If we were truly concerned about avoiding any nasty
surprises when porting GNU or Linux programs back to Unix, we
couldn't fix grep to handle lines of more than 512 characters,
or make it so -print is implicit in find(1), or add any
extensions to the language accepted by gcc.
> Look on the bright side; at least it's not as bad as Microsoft.
> Want sockets? Okay, that will be winsock.lib. No wait, wsock32.lib.
> Or should that be ws2_32.lib? Does this program need gdi32.lib?
> Or advapi32.lib?
That's one commodity I don't think Microsoft has a monopoly on.
Solaris (and who knows how many other "modern" Unix variants)
require you to use -lsocket and -lnsl (nsl? what's nsl?) if you
want to do simple networking, while those had always been in
libc in BSD Unix. A cow-orker has a Makefile with nearly a page
of -l directives in order to link some Oracle-using code
correctly; it took him days to get that part of the Makefile
right.
Steve Summit
[EMAIL PROTECTED]
------------------------------
From: "Lord Byron" <[EMAIL PROTECTED]>
Subject: Re: Large HD Access Problem
Date: Wed, 4 Aug 1999 01:31:09 -0500
Thanks to an e-mail from someone, I was able to track the problem down to a
kernel issue. Basically, the hard drive quit working when I recompiled my
kernel without frame buffer support, and some other vital thing. I still
had my frame buffer enabled kernel, so I tried that one, and was able to
access the drive with no problems again. When I recompiled the kernel, I
must've switched something else too, but I don't have a clue what it was,
and I lost the .config file that works. My question now is: Is there any
way to make a .config file with options from an existing kernel? Is that
what 'make oldconfig' does? I'd try it right now, but I'm not at my
computer.
--
Chad
------------------------------
From: Steen Suder <[EMAIL PROTECTED]>
Subject: Re: Need good sites for unix/linux
Date: Wed, 04 Aug 1999 06:53:23 +0000
Kelvin Tsang wrote:
>
> Hi,
>
> Which sites are the best sources for Unix/Linux
> software ?
>
> Thanks,
> Kelvin
http://freshmeat.net is widely used.
--
Best regards / Mvh.,
Steen Suder
sysadm kollegie6400.dk
GNU - makes me feel better! Ehhh, Linux is GNU, right...?
------------------------------
From: Kelvin Tsang <[EMAIL PROTECTED]>
Subject: Need good sites for unix/linux
Date: Wed, 04 Aug 1999 14:35:32 +0800
Hi,
Which sites are the best sources for Unix/Linux
software ?
Thanks,
Kelvin
------------------------------
From: Philipp Maier <[EMAIL PROTECTED]>
Subject: Re: Buy vs Download Difference: Documentation Only?
Date: Wed, 04 Aug 1999 08:59:14 +0000
Reply-To: [EMAIL PROTECTED]
Stewart Honsberger wrote:
>
> On Mon, 02 Aug 1999 11:32:11 +0000, Philipp Maier wrote:
> >- also commercial software such as StarOffice,
>
> I'm glad they didn't include it! Imagine adding 70 to the already 600 meg
> download.
>
> >- and, in general, the CD's you buy work much better. Don't ask why, but
> >this is my experience.
>
> Bad burn maybe? The CD we downloaded + burned works perfectly. Sure it
> doesn't have a pretty logo on the top, but it still works.
I was comparing the 'evaluation version', 'try-out versions' etc. with
the full product. My experience with the smaller versions was negative:
The software on the CD worked well, but as soon as I installed other
SuSE software from their server, I ran into difficulties. Certainly a
more experienced Linux-user would find its way out, but I got quite
frustrated...
--
Ever wondered how you could feel during your first marathon?
www.crosswinds.net/~pmaier
------------------------------
From: "Larry Clark" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.help,comp.os.linux.questions
Subject: ICQ and linux client
Date: Sun, 1 Aug 1999 16:52:25 -0700
man I tell you I have spent the WHOLE day triing to get a client for ICQ
that will work...not a dam***** thing...nothing,.....what a drag...think I
will sleep now....help!!!!!!!
------------------------------
From: [EMAIL PROTECTED] (Dionysus)
Subject: Clustering & Emulation
Date: Wed, 04 Aug 1999 04:21:51 GMT
Has anyone considered combining clustering software (whether a full
clustering distribution such as Mosix or a Program such as Beowulf)
with Windows95 Emulation, thus gaining the power of a large cluster of
machines whilest being able to access the huge amount of software
available for the win95 platform? If so, i would be very interested in
hearing of your experiences.
------------------------------
From: Philipp Maier <[EMAIL PROTECTED]>
Subject: Re: Distributions
Date: Wed, 04 Aug 1999 09:07:21 +0000
Reply-To: [EMAIL PROTECTED]
eze wrote:
>
> which is the most easy to install linux's distribution?
I am a fan of SuSE Linux. It takes less than an hours to install, and
the installation works smoothly. Also it has a very good manual, which I
consider quite important for beginners...
Philipp Maier
--
Information about Psion S5mx, SuSE Linux, Sylt and Maerklin mini-club:
www.crosswinds.net/~pmaier
------------------------------
From: Frederick Paul Suedmeyer <[EMAIL PROTECTED]>
Crossposted-To:
alt.os.linux,ch.comp.os.linux,comp.os.linux.setup,comp.windows.x.kde,linux.redhat.install,linux.redhat.misc,linux.redhat.ppp
Subject: Re: Kppp/pppd probs
Date: Wed, 04 Aug 1999 07:13:26 +0000
Al Kooz wrote:
> I'm trying to use kppp to connect to my ISP and it doesn't work. first it
> dials then the usual noise, and then it says that pppd unexpectingly died
> and I get disconnected. I don't know what to do, can anybody help me?
>
> when I try to use the network configuration of Redhat, and activate the ppp
> interface there, it dials the number to my ISP it connects, but then it just
> sits there and waits and gets disconnected too after about 5 secs. in the
> predefined script there it says:
>
> Login: 10306
> word: *********
> Timeout: 5 sec
> ~--
>
> Please can anybody help me on either of those 2 probs I'd be happy to browse
> the net on my box, no matter how !!
>
> thx Al
I had the same problem when I started out with KPPP. I found that my problem
was not with KPPP but my US Robotics 33.3 settings which were not able to do
the speed. I kept setting it to 115200 and 57600, but I set it to 38400 by
accident late one evening and KPPP worked like a charm. Then I started
tinkering with the modem settings and when I set it with ATZ2 it ran at 115200
- now that modem screems. Also, I discovered once I had that problem resolved
it was just as easy to write a script or use linuxconf.
That's what worked for me, good luck!
--
[EMAIL PROTECTED]
ICQ #112187
------------------------------
From: David Frantz <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: Java makes Netscape crash
Date: Tue, 03 Aug 1999 20:29:32 -0400
Reply-To: [EMAIL PROTECTED]
jamie wrote:
> s.c.park <[EMAIL PROTECTED]> wrote:
> >I am a newcomer to Linux/Unix. But, I solved the problem by installing
> >kaffe which is a Java virtual machine.
>
> I'm not familiar with kaffe. How do you make it supersede Netscape's
> own java? I ask because someone suggested a JDK, which did nothing.
>
> --
> jamie ([EMAIL PROTECTED])
>
> "There's a seeker born every minute."
Hmm;
I have the JDK installed also; Blackdown 1.2 in this case.
I'm wondering if there is a simple fix like a symbolic link or a shell
variable that could be set to clear this up.
Dave
------------------------------
** 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
******************************
