Linux-Hardware Digest #563, Volume #10 Wed, 23 Jun 99 07:13:41 EDT
Contents:
Re: Windows easy to install? BULLSHIT! (Brian Hartman)
Movie Machine II? (Gonzalo Garcia)
Re: Windows easy to install? BULLSHIT! (Brian Hartman)
Re: Backpack CD with Install and Thinkpad (The Rev)
APM problem on MACOM Laptop (Bjoern Giesler)
Errors on IDE Zip drive. ("Bobby D. Bryant")
Re: CD-R and linux ("test")
Re: Can't write to IDE zip drive (John Blinka)
How to compile SMP driver( just -D__SMP__?) into RedHat 6.0 kernel? ("robert_c")
Re: Auto-init DMA on SB16? (Davin McCall)
ModemBlaster I56k PCI Card ("Christopher R. Mitchell")
IDE CDR Writer with RedHat 6.0 ("Steven Webb")
Re: SoundBlaster PCI 16 or Ensoniq AudioPCI (Steven Ponsford)
----------------------------------------------------------------------------
From: Brian Hartman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.linux.setup
Subject: Re: Windows easy to install? BULLSHIT!
Date: Wed, 23 Jun 1999 01:12:11 -0400
[EMAIL PROTECTED] wrote:
> On Mon, 21 Jun 1999 12:08:07 -0400, Brian Hartman <[EMAIL PROTECTED]> wrote:
> >[EMAIL PROTECTED] wrote:
> >
> >> On Sat, 19 Jun 1999 18:56:06 -0700, Jack Coates <[EMAIL PROTECTED]> wrote:
> >> >[deletia]
> [deletia]
> >> GOD DAMN! Are you telling us you have to be a bloody
> >> MCSE to deal with Windows problems. Arguement's done
> >> right there...
> >>
> >> Windows cabal conceeds.
> >>
> >> [deletia]
> >>
> [deletia]
> > If you look at sales, Windows outsells Linux by a wide margin. (I think we
> >can all concede that.) The fact that a specific company that sells Unix boxes
> >was mentioned lends credence to the point that you have to look pretty hard to
> >find an off the shelf Unix box. In most situations, if a user is going to be
>
> No. One must merely exercise some care when one is selecting
> hardware. This is much like the burden to avoid the like of
> ATI and their 'poor drivers'. The net effect is still the same.
> The so called Windows advantage can evaporate at any time if
> you're unlucky.
In the first place, you need to be a *lot* more unlucky to find a company with a bad
Windows driver. And for the record, I didn't claim that ATI made "poor drivers" in
general. Maybe just for that one device. Or maybe it was a bad disk. But to blame
Windows for a problem with a driver that ATI released is not rational.
>
>
> >installing a new OS, they're going to go from Windows 95 to something else. If
> >that "something else" is Unix, they've got a long road ahead of them. Drives
> >aren't even referred to the same way in Unix as in DOS. While it's a minor
> >point, it can add to a novice's confusion.
>
> This is an non-issue. Inflexible people will have trouble going
> from Program Manager to Explorer or from Explorer to Finder.
>
I wasn't talking about a change in technology. I was refering to a paradigm shift,
wich
is much more difficult for new users, who don't necessarily understand what's "under"
the OS (and shouldn't actually have to, if it's designed well). Going from a world
where A: is the floppy drive to one where /dev/fp0 is the floppy drive and hda is a
hard
drive, is a semantic shift which takes adjustment at best.
>
> > Another point I'd like to make: I upgraded my kernel last night, and it was
> >way more difficult than anything Windows ever put me through. The upgrade had
>
> Then don't upgrade your kernel by yourself. Such things are not
> rocket science.
And why should a user have to rely on outside help for what ought to be a simple task?
In the first place, this dismissive attitude is exactly why we're having such an
argument. Your response seems to always be, "Well, of course you shouldn't do that!
What are you, an idiot??" But if Linux was really the easy OS you claimed it to be,
what would be the harm in a new user upgrading his or her own kernel? An end user
would
have absolutely no problem installing a Service Pack (other than the obscene download
time). And yes, there are RPM's available for less recent kernels (or some number to
that effect) but not for the latest kernel, and installing an RPM could leave a user
without options in the kernel that they want or need. The point of an easy-to-use
operating system ought to be that it is as easy to use for the beginner as the expert.
Obviously, no OS does this perfectly, but Windows, for all its other faults, does this
better than Linux.
For your reference, here's the URL to upgrading the kernel:
http://www.redhat.com/mirrors/LDP/HOWTO/Kernel-HOWTO-3.html
There's a more concise guide at:
http://www.thecomputergallery.com/redhat/kernel.shtml
At the page above, you can see that a complete kernel upgrade (in which you're actually
getting the benefits of upgrading the kernel that you wanted in the first place) takes
no less than 18 steps. The individual commands themselves are not rocket science,
obviously. But string them all together and factor in that if you miss any of the
steps
your kernel won't upgrade, and there you have a laborious procedure, at the very least.
> This is why the Linux distribution was invented.
>
> >at least 12 steps to it, none of which are handled programatically or even
> >cued. The whole process took about an hour and a half the first time, after
> >which I discovered I hadn't included PPP in the kernel, and therefore couldn't
> >get out on the net. So I tried again. And again. Finally I realized I was
> >missing a step and had to do it one more time. I started the process at around
> >12 and didn't finish until 4:30. How does that compare with installing a
> >Service Pack? (which kernel upgrades are roughly equivalent to) Windows has a
>
> Why is it even relevant? You choose the most difficult manner
> to 'fix' a system and then complain when it doesn't go very
> nicely. The problem is hardly due to Linux.
>
It's relevant because it's what you have to go through to get the features you want and
need out of the kernel. And it's relevant because there are numerous pieces of
hardware
which depend upon the kernel to be active at boot-up (a CD-ROM drive, for instance).
You can't make an ease of use argument and then say it's irrelevant when difficulties
are encountered. It's interesting that I point out all these ways in which
installing/upgrading Windows is easier, and all you come back with is, "that's
irrelevant". In any case, I suspect that the kernel upgrade problem would be easy
enough to fix in a later distribution. (A simple shell script to do all the commands
automatically, and to guide you step-by-step through your kernel configuration choices
would do it.) It's simply more of a burden than a user looking to upgrade the features
in their kernel should have to face.
>
> >big jump on Unix when it comes to usability, and the novice benefits from it.
>
> No, you just like spreading FUD. There are ease of use update
> systems in place just for people like you. Bother to use them.
>
As I mentioned, the ease of use systems in place are simply not up to the task. An
experienced Windows user migrating should not have to sacrifice features they want
simply because the upgrade procedure is complex. Yes, a user that didn't know what
they
wanted or care about what features the kernel supported could easily go for an RPM.
But
a beginner in Linux who *isn't* a beginner to computing would not want to sacrifice
such
features, and they shouldn't be made to suffer for it.
>
> >It's only when the user outgrows the novice stage that they realize what they
> >give up for that usability. But to say that Unix is just as easy to install as
> >Windows is silliness.
>
> This machine that I am typing on here (NTW/4) was a simple RH 5.2
> install. It was put together specifically to be an NT workstation,
> yet all it's hardware was autodetected and configured by the RH
> installer.
>
"Simple" is obviously a relative term. As I stated earlier (and referenced in the
HOWTOs) hardware is not necessarily the fly in the ointment. The main components of my
system (video display, keyboard, touchpad) were all autoinstalled and configured by
both
NT and Red Hat. It's the other points (repartitioning, fips, mount points) that make
RH a more difficult install. No, it doesn't take a CS degree to do an install on a
clean hard drive. But it does involve repartitioning (for the recommended swap drive,
at least) and giving up some of the simplicity built in to Windows.
You can dress your argument up all you like. What it boils down to is this: You want
stable and robust? Go with Linux. You want easy to use? Windows is a better bet.
>
> [deletia]
> >>
> >> The crux of my argument against Windows for the novice
> >> has always been that it's really no less complex for
> >> that class of user than anything else that has to deal
> >> with the kludge klone underneath.
> >>
> >> Want easy? Just buy that to begin with.
>
> BE the invisible hand; allow those vendors that care to
> cater to what you believe to be your real needs to
> actually flourish.
>
>
> --
>
> bash: the power to toast your registry in style... |||
> / | \
>
> Seeking sane PPP Docs? Try http://penguin.lvcm.com
------------------------------
From: Gonzalo Garcia <[EMAIL PROTECTED]>
Subject: Movie Machine II?
Date: Wed, 23 Jun 1999 11:25:31 +0200
Hi
I recently found an old Movie Machine II (video editing card, ISA, mpeg,
tvtuner, video mixing...) and thought... Is there anything at all for ir
under linux?
------------------------------
From: Brian Hartman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.linux.setup
Subject: Re: Windows easy to install? BULLSHIT!
Date: Wed, 23 Jun 1999 01:16:21 -0400
Aaron Miller wrote:
> > Linux requires a great deal more initial planning
> > before you set it up, whereas most of the problems with Windows happen
> > *after* you install it.
>
> Please excuse my profanity, but what the hell are you trying to get
> across? Would you rather have to re-do a Linux install once or twice, or
> have your Windoesnot box crash a year from now and lose everything on
> it? I mean, are you an idiot, or *what*?
>
> ---- ----
> This has been a paid presentation amille emiss.
> for Aaron Miller by Channel 42. m r@ol edu
> ---- what was THAT? --- --- look out batman! ---
The name of the thread is "Windows easy to install?" Yes, I would rather
have to install Linux twice and know that all my data is safe (which is why
I'm running Linux now, BTW). But that has nothing to do with which is
easier to install. Didn't mean to drive you to profanity. I was just
staying on topic. ;)
------------------------------
From: The Rev <[EMAIL PROTECTED]>
Subject: Re: Backpack CD with Install and Thinkpad
Date: Wed, 23 Jun 1999 08:52:18 GMT
Cheers for that. I downloaded the new disk images from RH and despite
claiming that it would fix the Backpack CD problem it didn't.
I tried the 5.2 Boot floppy with the 6.0 CD and everything went fine
until it rebooted and lilo tried to launch the new kernel and BOOM! it
panicked claiming that it had been compiled for a Pentium+ and since
I've only got a 486 (DX2 50, on an ancient Thinkpad laptop), then that's
no good.
My questions are...
Are there more than one kernel on the RH install CD to cope with
different architectures or is the 486 no longer supported by RH?
Did this happen because I used the 5.2 boot floppy? It happily went on
to "second stage install", which I thought meant it was doing everything
off the CD.
Ideas?
TIA
Craig
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> On Mon, 21 Jun 1999 12:03:22 GMT, The Rev <[EMAIL PROTECTED]>
wrote:
>
> >when I installed RH 5.2 on my Thinkpad 360CSE I simply told it that I
> >wanted it to use my Backpack CDROM and it did so.
> >
> >With RH6.0 (choosing local CDROM, Backpack, autoprobe) the installer
> >says: "I can't find the device anywhere on your system!".
>
> The backpack driver on the 6.0 boot disk is bad. RedHat are aware of
the
> problem and are planning to release a new boot floppy. Check their
> web site.
>
>
==========================================================================
> Grant R. Guenther
[EMAIL PROTECTED]
>
==========================================================================
>
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: Bjoern Giesler <[EMAIL PROTECTED]>
Subject: APM problem on MACOM Laptop
Date: 23 Jun 1999 08:58:06 GMT
Hi,
I'm having a couple of problems using AGP on my new MACOM laptop. First and
foremost, when I switch from battery to AC or vice versa, often the thing
completely locks up, so I have to switch it off. Sometimes it unlocks after
a few seconds with "hda: lost interrupt". Anyone have any idea on how to fix
this?
Second (but this may be a fundamental problem): I've set suspend mode to
"Save to disk". 'apm -s' works like a charm, but on return from suspend
mode, my pcmcia network card is lost; I have to re-start cardmgr to make it
work again. Is there a way to change this?
TIA,
--Bjoern
--
================================/\==One OS to rule them all===Windows NT=====
Bjoern Giesler / \ One OS to find them
<[EMAIL PROTECTED]> / <> \ One OS to bring them all
=============================/______\==And in the Darkness bind them=========
Thank you for your time, worship the Antichrist, and have a nice day.
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Subject: Errors on IDE Zip drive.
Date: Tue, 22 Jun 1999 23:34:35 -0500
Cliff Bergman wrote:
> I have a brand-new pretty standard PC with and ATAPI zip drive. I have
> no trouble reading from the drive (in either ext2 or vfat format), but I
> get numerous write errors when I try to write to the drive. (Sample
> errors below.) I'm running Redhat 6.0 right out of the box.
> ...
> Here is a sample of the error messeges in /var/log/messages:
>
> Jun 22 20:28:41 bergman kernel: ide-floppy: hdd: I/O error, pc = 2a, key
> = 4, asc = 47, ascq = 0
> Jun 22 20:28:41 bergman kernel: end_request: I/O error, dev 16:41 (hdd),
> sector 754
I brought up this group tonight just to ask the same question. I've been
seeing the same errors with a new Zip drive in my newly built system. I
see them on various disks, I saw them when I moved the drive to my old
system, and I still see them after having the dealer replace the drive.
I'm starting to think it's something to do with RH6, because I've ruled out
the media, the drive, and even the system the drive is in.
I've had the errors show up with the drive connected as hdb and hdd, so
that doesn't seem to be an influence either. I see them whether I use the
default kernel or my custom kernel.
I'm getting the same numbers in the message as you did, except that the
sector number varies with each message. I just copied c. 60Mb of files to
the drive, and it looks like it generated almost 1000 messages. The 'diff'
utility reports that some of the files on the drive are in fact different
from the originals.
Another thing I've noticed is that some of my media which always mounted on
my old system will not even mount on the new system. Finally, the mount
command will sometimes complain that the device is reporting two different
block counts.
Any further troubleshooting hints will be greatly appreciated.
Bobby Bryant
Austin, Texas
------------------------------
From: "test" <[EMAIL PROTECTED]>
Subject: Re: CD-R and linux
Date: Wed, 23 Jun 1999 08:50:45 GMT
http://www.freshmeat.net/appindex/1998/03/13/889803174.html
Tero Niemi <[EMAIL PROTECTED]> wrote in article <[EMAIL PROTECTED]>...
> How to setup a burning... eh.. writing cd-rom in RH6.0?
>
> It works just fine if I try to read but not when I want to write. I've
> tried to change it through fstab writable,
> but that of course didn't help. I have mitsumi 8x/4x
>
>
------------------------------
From: John Blinka <[EMAIL PROTECTED]>
Subject: Re: Can't write to IDE zip drive
Date: Wed, 23 Jun 1999 06:24:01 -0400
Cliff Bergman wrote:
>
> Hi,
> I have a brand-new pretty standard PC with and ATAPI zip drive. I have
> no trouble reading from the drive (in either ext2 or vfat format), but I
>
> get numerous write errors when I try to write to the drive. (Sample
> errors below.) I'm running Redhat 6.0 right out of the box.
>
> I've tried more than one disk in the drive, with the same result, and in
>
> fact, the same thing happens on a second, identical computer. Also,
> both reading and writing work fine in Windows 98. So I don't think
> either the disk or the drive is defective.
>
> The zip drive appears as hdd in linux. It is on the same ide cable as
> the cdrom drive (which works fine---at least when I read from it. I
> haven't tried to write to it. :-) Both devices are configured (by
> jumper) to "cable select". I tried setting the cdrom to master and the
> zip to slave. The result: it still didn't work under linux, and it
> stopped working under Win98. In fact, Win98 couldn't find the cdrom,
> and it thought the zip was an ordinary floppy drive.
>
> Does anybody have any suggestions? Needless to say, the Dell computer
> people weren't much help. And the RedHat people can't seem to solve it
> either. If possible, please email me directly. ([EMAIL PROTECTED])
>
> Thanks in advance.
>
I had exactly this problem with a 100 mb atapi zip
drive and solved it by hard coding the zip drive
geometry as 96/64/32 C/H/S in the bios instead of
letting the bios autodetect the drive. End of problem.
I don't know why this works, or even if 96/64/32 is
"correct", but the drive now works fine under linux
and win95. At boot, I always get the message:
kernel: hdd: The drive reports both 100663296 and 1006
46912 bytes as its capacity
but this doesn't seem to have any adverse effect on using
the drive. (This message appeared before I made the BIOS
change.)
I've got my zip drive configured via jumper as a slave
to a hard disk, configured via jumper to master. I'm
running Slackware 3.6 with a customized 2.0.36 kernel.
-- John Blinka
------------------------------
From: "robert_c" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: How to compile SMP driver( just -D__SMP__?) into RedHat 6.0 kernel?
Date: 22 Jun 1999 04:24:29 GMT
Could someone tell me about the following:
How to compile SMP driver ( just -D__SMP__ or need more options or __SMP__
can be igored?) into RedHat 6.0 kernel?
------------------------------
From: [EMAIL PROTECTED] (Davin McCall)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Auto-init DMA on SB16?
Date: Wed, 23 Jun 1999 10:18:38 GMT
Actually, it's not as bad as you seem to think. Firstly, there are
lots of things you can do to control things like buffer sizes using
IOCTL calls. You can use map the DMA buffer directly into your
application's address space, and find out how much of the buffer has
actually played at any one time.
Sadly, I don't think it's very easy to detect iterrupts in linux.
About the next best thing is using threads and select() calls.
The best source of info is at:
http://www.opensound.com/pguide/
On Tue, 22 Jun 1999 15:38:26 +0100, "Stephen Jacob"
<[EMAIL PROTECTED]> wrote:
>Hi folks,
>
>I and another guy developed an application 2 years ago (to play back seismic
>data as audio) that uses the SoundBlaster16 under DOS. We require very
>precise timing (to know exactly what the sound card is actually playing at
>the moment when somebody presses a key), and the way in which we implemented
>this was to set the SB16 up in Auto-initialise DMA mode (see the SBHWPG on
>www.soundblaster.com).
>
>In auto-init DMA mode, a DMA buffer is used, the card reading through the
>buffer constantly (looping through it), and sending an interrupt each time
>it finishes half of the buffer. What you do when it finishes reading the
>'bottom' half is you refill the 'bottom' half while it continues ... playing
>the 'top' half... and when it sends the next interrupt, it means it's
>finished the 'top' half, so you refill that while it plays the 'bottom'
>half. This works well because all that is required is to count the
>interrupts and to make the buffer sufficiently small to suit our resolution
>requirements.
>
>It has been suggested that I port the software to Linux (well, to Windows
>was suggested before, but that's a horrible idea). This sounds nice, until I
>read up on sound programming in Linux... As a user-level program, it looks
>like there is no access to such low-level stuff. /dev/dsp just supports you
>writing data to it and reading data from it. I assume that what happens is
>that the /dev/dsp driver keeps a large buffer of what you've written to it,
>and uses something like Auto-init DMA (or one of the other modes) to feed it
>to the soundcard. The problem is that this buffering will mean that I won't
>be able to be certain of what is playing _right now_.
__________________________________________________________
*** davmac - sharkin'!! [EMAIL PROTECTED] ***
my programming page: http://yoyo.cc.monash.edu.au/~davmac/
------------------------------
From: "Christopher R. Mitchell" <[EMAIL PROTECTED]>
Subject: ModemBlaster I56k PCI Card
Date: Mon, 21 Jun 1999 19:49:00 -0400
Boy, ever since I install RH5.1 I been meeting making new friend. Let's see
who is next. Ok, I have an ModemBlaster i56k pci card on my linux machine.
I would to use it if possible... Yes it and how do I set it up...
------------------------------
From: "Steven Webb" <[EMAIL PROTECTED]>
Subject: IDE CDR Writer with RedHat 6.0
Date: Tue, 22 Jun 1999 01:09:07 +0100
I've got a CD-ROM and a Philips 3610 writer in my PC which works fine with
Windows for writing CD's but I would like to use them with linux (redhat
6.0). With the normal installation I can use the CD-ROM but I need to use
scsi emulation for writing the CD's. I've tried rebuilding the kernal with
scsi emulation but when I did I couldn't figure out how to mount my cd's. I
tried using scd0 ..... but got know where. What do I need to put in the
/etc/fstab file and are there any other things I have to do.
Any help would be gratefully received.
------------------------------
From: Steven Ponsford <[EMAIL PROTECTED]>
Subject: Re: SoundBlaster PCI 16 or Ensoniq AudioPCI
Date: Mon, 21 Jun 1999 12:59:18 -0700
"Kevin C. Weissman" wrote:
>
> Which is supported better? A Creative Labs SoundBlaster 16 PCI or and
> Ensoniq AudioPCI. Can the onboard midi sequencer/port be used on either of
> these? Thanks.
> --
> /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
> Kevin C. Weissman (KW) |ACORN "techie" and proud UNIX advocate.
> [EMAIL PROTECTED] |Warning: failed.
Hi Kevin,
According to my /proc/pci
....
Bus 0, device 16, function 0:
Multimedia audio controller: Ensoniq Unknown device (rev 6).
Vendor id=1274. Device id=1371.
Slow devsel. IRQ 10. Master Capable. Latency=64. Min
Gnt=12.Max Lat=128.
I/O at 0xdf00 [0xdf01].
....
my SB16PCI is identified as an Ensoniq 1371 so you should be able
to use the builtin Creative/Ensoniq AudioPCI 1371 support in Linux.
Just run sndconfig and it should set it up.
Midi support isn't there yet but you can use TiMidity compiled with
esound support(softmidi). TiMidity/SB16PCI sounds good with the
GUS patches :)
------------------------------
** 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.hardware) 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-Hardware Digest
******************************
