Linux-Hardware Digest #35, Volume #10 Thu, 15 Apr 99 14:13:30 EDT
Contents:
Re: All the current OSes are idiotic (was Re: Is Windows for idiots?) (GAZZA)
I/O error on SCSI disk ("Fr�d�ric Hoerni")
Re: Mac SCSI Scanner on a Linux _PC_ SCSI system? (Markus Wandel)
Creative Ensoniq AudioPci64 ("Alan W. Jurgensen")
Ditto 2GB help ("Billy B. Booker")
Re: All the current OSes are idiotic (was Re: Is Windows for idiots?) (westprog)
Multia/UDB Linux Installation (Mike Kuht)
Re: All the current OSes are idiotic (was Re: Is Windows for idiots?) ("Osvaldo
Pinali Doederlein")
Re: XServer for NVidia RIVA 128ZX (morgan)
Re: 2.2.5 Kernel is dead ("David Murray")
lost mixer with 2.2.5 upgrade (Bob van der Poel)
ITK Colombus World ISDN adapter ("Ole Eirik Dalbye")
Re: Travan TR4 / Seagate Hornet Tape Dirve ("Kenneth H. Cox")
Re: Support for CD changers? (Piniek aka Piotr Ingling)
----------------------------------------------------------------------------
From: GAZZA <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.java.advocacy,comp.os.linux.advocacy,comp.os.ms-windows.advocacy
Subject: Re: All the current OSes are idiotic (was Re: Is Windows for idiots?)
Date: Thu, 15 Apr 1999 11:39:15 +0800
westprog wrote:
>
> In article <[EMAIL PROTECTED]>,
> GAZZA <[EMAIL PROTECTED]> wrote:
> >
> > To say nothing of the fact that it prohibits useful mnemonics.
> > For example, a coding standard that states all class names should
> > have an initial uppercase letter, and all variable identifiers
> > should have an initial lowercase letter. This means I can do:
>
> > Object object
>
> > and instantly create a nice hint to the reader about exactly
> > what class object is, and so forth.
>
> This could be useful, indeed. The trouble is that now you have
> "Object" and "object", which mean different things - if you fail to
> press the shift key at the right time, you get a valid symbol in the
> wrong place.
No, you don't. You get an INvalid symbol in the wrong place. This
distinction is very important; the compiler (or a good GUI) will
spot this INSTANTLY.
> This leads to confusion if you are discussing the code with
> someone - "Object" sounds quite similar to "object", and it looks
> very alike when written on a whiteboard.
I don't tend to write up code on whiteboards, just designs. When I
want to discuss code, I get a printout. Is my methodolody in this
respect so very unusual?
> And what if you find that you need a second instance? Ah, I know -
>
> Object object, oBject; // sorry, being facetious
Depends on the situation of course. Maybe:
Object object1, object2;
or
Object objects[2];
The point being that in many methods you only need 1 temporary
object of a given type. Calling it by the same name as the class,
with just the case changed to conform to your coding standard for
a variable, is as good a method as any.
> There are plenty of other conventions which could be used - for
> example, the Borland libraries use
>
> TObject Object;
... which is more keystrokes; weren't you just objecting that I
might forget to press Shift somewhere? Isn't it just as likely
(at least) that I'll type T somewhere I shouldn't?
> > In a case sensitive language, such sloppiness is prevented. Isn't
> > this a good thing? Doesn't it mean that the compiler is able to
> > force more "natural looking" code?
>
> I don't think that it is up to the compiler to force "natural
> looking" code; that is appropriate for an editor.
Well, we disagree. Fair enough.
> > Probably my_filename, actually, from most C coding standards I've
> > seen. That's more of a cultural thing than a limitation of case
> > sensitivity, though. The Java coding standard uses case effectively,
> > and so do (eg) most C++ programs.
>
> The problem I have is that case is used, as it should be, to make
> the text representation of a program more readable. This has, in
> case-sensitive languages (which AFAIK are only the C family, in the
> broad sense) a semantic effect.
Saying that the case-sensitive languages are "only the C family" is
like saying influenza only infects humans with "2 arms and 2 legs".
It's true, more or less, but since the C family comprises the
majority of programs, it's also misleading.
Here's my basic issue: I think it's wrong to call something
"var1" in one place and refer to it as "VAR1" in another. Most
coding standards I've seen - for case insenstive languages such
as Ada - agree with me in this respect. They all enforce some
rule about what case to use for variables, class names, and so
forth.
If you're going to enforce such things ANYWAY, why not let the
compiler check it for you? Why not make it an error to NOT do
it, since (by the coding standard) it SHOULD be an error.
> > It doesn't, automatically. Chris was wondering - as I am - if your
> > problem is with filenames specifically, or case sensitivity in
> > general. It seems - from your above comments on Pascal, and below
> > on passwords - that it is indeed case sensitivity that you object
> > to rather than only case sensitive file names.
>
> I believe that case should not have semantic significance, because
> that is not its use in normal text.
So what?
Programs are not normal text. It is difficult to see why they should
therefore be judged by such a standard.
> I believe that case-sensitive systems - generally - reflect
> engineering concerns and the internal representation of symbols
> rather than their appropriate representation.
Considering that programs are written to address engineering
concerns, I think that using case sensitivity is NOT an inappropriate
representation.
If one is enamoured of the Natural Language paradigm, one can always
use some automatic converter. But when it comes to debugging source
code, I for one prefer the semantics that can be embedded into a
case sensitive language.
If Roedy ever gets his SCIDs, perhaps my concerns will be mute - but
at that point, so will yours.
> > A matter of opinion. It's a tautology that the more different
> > possible characters there are for passwords, the more difficult
> > they are to crack. Making passwords case-insensitive reduces the
> > number of different characters for passwords. Ergo, they are
> > less safe. Yes, you can have longer passwords, but (speaking
> > personally) I'd rather type in mYpAssw0rD than
> > mynoncasesensitive33pass9word simply because typing in passwords
> > is a common thing to do and its a PITA to have to type in a long
> > one. The shorter the better, as long as it's still reasonably safe.
>
> Here is a little test. Look quickly - about five seconds - at the
> above passwords. Then turn away and write them both down. I suspect
> that the longer password is more likely to be remembered correctly,
> and that you can write it quicker as well.
You were wrong, actually. I remembered both, but I had more trouble
with the latter than the former because there are more numbers in it.
Here was my mnemonic: "Look for exceptions". That is, I remember that
the basic phrase is "mypassword", and that it's mostly lowercase.
Then I remember that the Y, A, and D are uppercase, and that the o
is a zero. So I have 6 things to remember including each word of the
phrase.
With the longer one, I have 5 words to start with. I then have to
remember that there's a 33 between sensitive and password, and that
there's a 9 in the middle of password. Which is 7 things to remember.
In all fairness, though, you chose a particularly bad example. This
doesn't really say much one way or the other in terms of whether
mixed case passwords are better.
> The reasons for this are fundamental to human being's use of
> written language - we learn words in a case-insensitive way.
> Elephant, elephant - we don't learn the spelling seperately for the
> start of a sentence, we apply the case according to context.
> It is the job of a programming language or operating sytem to take
> human-readable text and graphics and hand movements and to
> translate them into a machine-comprehensible format.
Maybe in the far future this will be possible. Maybe in the 24th
century we'll be walking along the corridors of the Enterprise and
say, "Computer, where is Commander Riker?"
At which point case will indeed be irrelevant, but only because
text itself will be.
For the moment, it is the job of the programming language to express
to the computer in a non-ambiguous fashion what actions to perform.
As an analogy of the differences between your position and mine,
let us pretend that instead of a computer we have someone who
speaks only Russian, and we speak only English. Your approach is
to teach him English. My approach is to learn Russian.
Now, let's say that this person wasn't playing with a full deck,
if one forgives the euphemism. Computers, as a rule, are not
particularly bright. It's a lot easier for us to learn their
language than for them to learn ours.
> Computer engineers tend, all to frequently, to assume that a person
> can easily learn to think in computer terms.
This is because it's empirically true. The same way people can learn
to think in different languages.
> The assumption that 'A' is different to 'a' because the ASCII
> codes are different is just another example.
To a computer, these symbols ARE different. We're talking to a
computer. Ergo, it makes sense (IMHO) to express ourselves in
terms that the computer will understand.
In fact, though, your entire argument is a little shaky. 'A' IS
different to 'a'; it's just that the context in which this is the
case is way too difficult for us to express to the computer.
No matter which way you look at it, we HAVE to "dumb it down" to
some degree.
> > This says more about your imagination than anything else. ;-)
>
> I am quite prepared to be enlightened as to why somebody would want
> precisely what is described above.
What about the several examples I and others have given you about
computer generated filenames?
>
> > Then open up a GUI FileManager, and use it to do your file creation,
> > editing, and removal. I'd say that a significant percentage of the
> > files I create (as opposed to those that I have automatically
> > created) are done via some form of GUI interface, whether it's on
> > Unix OR Windows.
>
> What if I am doing a File..Save As into another directory, say?
What's the problem? You're talking about possible spelling errors
to retrieve files, not create them.
> Just saying "use a GUI" hides the problem. Suppose you are opening a
> file in a directory with hundreds of different files? You remember
> that it has a name like myFilesomething - in a case sensitive
> system that is no help.
And what if all I remembered was that it was called "Filesomething"?
I forgot about the "my" part?
> I don't believe that shells or GUI's should lie to me.
??? What are you talking about? They DON'T lie to you. Nobody was
suggesting that they did.
> I would much rather have a case-sensitive system that is honest
> about it than one that says "you're a foolish GUI person, and
>we will conceal the gory innards of the file system from you".
You've completely misunderstood me here.
I understood you to be complaining that you often mistyped
"myFilename" when you meant "MyFilename". My reply was that, if
you use a GUI (and therefore aren't typing EITHER) this mistake
doesn't happen. You seem to have construed this as me implying
that the GUI would misrepresent the case of the files; that wasn't
my point at all.
> > I mean, why stop there? Even with case insensitivity, you might
> > accidentally create MyFileName.txt and MyFileName.text, or
> > MiFilename.txt. Getting rid of case sensitivity isn't going to
> > stop many typo problems.
>
> The .txt/.text problem is easily dealt with by explicit file typing.
> When you save your file, it knows that it is a text file, or a C++
> program, or whatever, and the .txt suffix is unnecessary. (Yet
> another break with Unix tradition, of course). Mifilename is an
> obvious misspelling of MyFileName - MyFilename is obviously the same
> name.
So you're saying that if I want to create files with arbitrary
extensions, I have to create them with some editor that will name
them ".txt" and then later manually remain the files? Or am I not
supposed to be able to think of any extensions that weren't "in
the box"? Even DOS based systems don't "know" the extension of your
file; at best, they have a default.
And how is "Mifilename" an "obvious" misspelling of "MyFileName"?
What if MI were my initials, for example?
There will always be cases where users misspell file names. If they
do it frequently - as I do! - then they should avoid having to type
them at all, rather than trying to "second guess". I use File
Manager tools for most of my opening and editing files. I doubt that
I'm unique or even unusual in this respect - among Unixphiles also.
> > I challenge you to find a Java programmer who fits that profile.
> > Or most Perl programmers. Or a lot of Smalltalk programmers.
>
> I just looked at my big fat Perl book, and I didn't see a single
> example of mixed case, except, funnily enough, in the Win32 modules.
Then you aren't looking very hard. What do you have, Programming Perl?
I just flicked mine open at random; try page 514. Better yet,
perldoc perlstyle.
In general:
- File handles are all uppercase by convention
- Package names have their first letter uppercase by convention
- Local variables are all lower case by convention
Frankly, if you're actually being honest here, you have a pretty
bad Perl book.
> Java does seem to use it rather more. Smalltalk I don't know.
>
> Funnily enough, the SQL book I am looking at switches from all
> upper to all lower case halfway through the book.
Which is confusing, and shouldn't be permitted. IMHO.
> "Software Tools In Pascal" uses all lower case - who wrote that -
> my goodness, it's Kernighan! "Oh! Pascal" uses all lower case, but
> uses bold and italic text to make the programs exceptionally
> readable.
So what you're saying is that even case-insensitive language tend
to use a single kind of case? I don't see how this supports your
argument.
> > Free us from the chains of insensitive case, and we will use our
> > freedom to create more readable programs. ;-)
>
> I've read a few C/C++ programs. Also some Pascal.
And?
I hope you weren't intending that statement to be self evident in
meaning. I've personally seen readable programs in all three
languages, but I've never seen mixing case irresponsibly to be a
particularly valuable feature in this respect. If I name an
identifier "bar", then I expect a reference to "Bar" to be rejected.
If I'm reading a program and I see:
var bar:integer;
Bar := 54;
then my immediate impulse is to start looking for where Bar was
defined. Which is why that sort of thing is discouraged even where
it is syntactically legal.
> > About the only language that fits your viewpoint is C - and even
> > then, all uppercase is standard for constants. The point is that
> > WITH case sensitivity, we can be guaranteed of these "special
> > cases" and they leap out at us while we're parsing a file. Pascal
> > gives me no guarantees that A_SPECIALLY_CASED_IDENTIFIER wasn't
> > declared as a_specially_cased_identifier somewhere earlier, so I
> > don't have these sorts of visual clues UNLESS the programmer is
> > careful to use different cased identifiers and avoid "clashes".
> > If they're going to that much trouble, why not take the next
> > logical step and make it case sensitive?
>
> A_SPECIALLY_CASED_IDENTIFIER makes it a lot more readable if you
> are using a 1980's style editor. Modern programming editors should
> make keywords, constants, strings etc. visible without the
> programmer having to use his own typographic tricks.
Err... what about languages where the difference between a constant
and a variable is one that is not intuitively obvious to the editor?
Not all languages have a "const" keyword; and besides, you can never
have too many visual clues.
And what if we have to print our sourcecode on a black & white
printer?
> > Why not just use a File Manager and double click on the appropriate
> > file?
>
> A file manager is just an interface to the file system. If I do a
> search with filemgr, and I get the case wrong, I won't find the
> file, just the same as if I used the shell. GUI's aren't magic -
> they are just another way to interface with the same system.
The same thing would happen if you got the name wrong. I'm sorry,
my experiences just don't bear out your conclusions.
> > If you're REALLY pedantic, just create a hard link to
> > MyFirstProgram.java called myfirstprogram.java and be done with it.
> > They're now the same file. ;-)
>
> The horrifying thing is that Unix encourages this kind of
> "solution" to the problem.
You've missed the point. The point is - I don't consider that it IS
a problem.
"That's the way (uh huh uh huh) I like it (uh huh uh huh)".
And since you can have it the way YOU want it as well in Unix, why
would you try to stop me having it the way I want it?
> My Unix shell does exactly that. It's horrible. My point is that
> words in English have a correct spelling.
Debatable. How does one spell "gaol"? "Colour"? "Theatre"?
To any American reader, the above line contains 3 spelling errors.
> They do not have an associated correct case, any more than they have
> a correct font. It is perfectly legitimate to reject a misspelled
> word - it is wrong to reject a word for incorrect case.
Excepting of course that filenames need not be and frequently are
not English words - so "correct" spelling is somewhat moot.
> > You could even write a shell that did this, should you so desire.
> > The shell could try first for an exact match, and then a case
> > insensitive match, and translate myfi[TAB] to MyFilename.txt;
> > there's absolutely no reason this couldn't be done. Go for it, if
> > that's what you really want!
>
> What I want (and I have said this many times) is, I believe,
> impossible with any form of Unix.
And I believe you're incorrect. It's certainly possible to have a
shell that ignores case sensitivity.
Unless, of course, you are DEFINING what you want as being "that
which is impossible with any form of Unix", which is a little
unsporting. ;-)
--
Cheers,
GAZZA
------------------------------
From: "Fr�d�ric Hoerni" <[EMAIL PROTECTED]>
Subject: I/O error on SCSI disk
Date: Thu, 15 Apr 1999 18:16:09 +0200
Hello,
I got the following message a few times when booting linux :
/dev/sda1 contains a file system with errors, check forced
scsi disk i/o error : dev 08:01 sector .....etc...
Additional sense indicates Unrecovered read error.
Sometimes it goes on ok, but other times I have to run fsck and answer
'ignore <y>' to the problems.
Does anyone have an idea of the problem and is it easy to correct ?
Thanks
Fred ([EMAIL PROTECTED])
------------------------------
From: [EMAIL PROTECTED] (Markus Wandel)
Subject: Re: Mac SCSI Scanner on a Linux _PC_ SCSI system?
Date: 15 Apr 1999 12:42:36 GMT
In article <7f2lqv$pjt$[EMAIL PROTECTED]>,
westonpa <[EMAIL PROTECTED]> wrote:
>
>>In Linux, you're on your own for drivers anyway, and it's assumed you're smart
>>enough to install a generic SCSI card and find the right cable, so it just
>>comes down to can you find a driver program for the scanner. I wasn't too
>>impressed with SANE 1.0 when I tried it, but I found a very nice standalone
>>program called "epscan" for my current Epson scanner (had to patch it, mail
>>me if you need that particular info.)
>
>What Epson scanner is it?
Epson ActionScanner II. The ID string it returns isn't quite what
epscan expects so you have to hack it ("EPSON SCANNER..." instead of
"EPSON SCANNER..."). Other than that 100% compatible.
Markus
------------------------------
From: "Alan W. Jurgensen" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Creative Ensoniq AudioPci64
Date: Thu, 15 Apr 1999 15:31:34 +0000
Hi Yall,
REgarding: Creative Ensoniq AudioPCI Sound Card...
Looks like it works in 2.2.5 Kernel using es1371....
As Taken from linux/Documentation/sound/es1371:
...
This driver does not support the ALaw/uLaw sample formats.
ALaw is the default mode when opening a sound device
using OSS/Free. The reason for the lack of support is
that the hardware does not support these formats, and adding
conversion routines to the kernel would lead to very ugly
code in the presence of the mmap interface to the driver.
And since xquake uses mmap, mmap is considered important :-)
and no sane application uses ALaw/uLaw these days anyway.
In short, playing a Sun .au file as follows:
cat my_file.au > /dev/dsp
does not work. Instead, you may use the play script from
Chris Bagwell's sox-12.14 package (available from the URL
below) to play many different audio file formats.
The script automatically determines the audio format
and does do audio conversions if necessary.
http://home.sprynet.com/sprynet/cbagwell/projects.html
...
ONE Strange Thing is: cat /dev/sndstat Displays nothing!
Does anyone have any sure-fire procedure for properly loading
Sound Modules required by this 1371 Card? IE, is an OPL2/3
driver needed... uart401 needed? what else!?
Any Help/Comments are appreciated!!
later,
al-man
James Kosin wrote:
>
> Dear Rob v.d. Meer,
>
> Its supported in RedHat 5.2 and kernel 2.0.36 and later.
>
> I haven't figured out how to get Xwindows setup to do sounds like Windows
> 95/98 yet. I am able to play the .au files and stuff like that.
>
> Thanks,
> James Kosin
>
> Rob v.d. Meer <[EMAIL PROTECTED]> wrote in message
> news:7f2qdo$k5m$[EMAIL PROTECTED]...
> > Hya,
> >
> > I've just bought an Creative Pci64 soundcard (The Ensoniq clone) but Linux
> > doesn't seem to recognize him.. At bootup it tells: Unknown device at PCI
> > adres 1237:1381 and the it does not install the sounddriver configured
> > in sndconfig.. The HOWTO of the AWE64 doesn't get me any further either
> caus
> > the AWE is an ISA card..
> >
> > I'm running Redhat 5.0 with kernel 2.0.33... Anyone can help ??
> >
> > Thnx in advance, Rob
> >
> > ---
> > Email: [EMAIL PROTECTED]
> > The Young Technicians Delft
> > Faculty of Electrical Engineering
> > University of Technoligy Delft.
--
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
Alan W. Jurgensen - Full Compass Systems
phone: (608) 831-7330 - email: [EMAIL PROTECTED]
+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
------------------------------
From: "Billy B. Booker" <[EMAIL PROTECTED]>
Subject: Ditto 2GB help
Date: Thu, 15 Apr 1999 08:52:28 -0500
I have installed ftape and configured it to use the parallel port for my
Ditto 2GB drive. When I try to use it with tar, tar lists the first few
files, 1 buffer possibly, then hangs. I would appreciate any
suggestions.
--
Billy B. Booker
(281)488-3939
[EMAIL PROTECTED]
------------------------------
From: westprog <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.java.advocacy,comp.os.linux.advocacy,comp.os.ms-windows.advocacy
Subject: Re: All the current OSes are idiotic (was Re: Is Windows for idiots?)
Date: Thu, 15 Apr 1999 13:54:10 GMT
In article <7f30gi$pfk$[EMAIL PROTECTED]>,
"Jon A. Maxwell (JAM)" <[EMAIL PROTECTED]> wrote:
> westprog <[EMAIL PROTECTED]> wrote: (comp.lang.java.advocacy)
> |
> |> Object object
> |
> | [...] if you fail to press the shift key at the right time, you
> | get a valid symbol in the wrong place. This leads to confusion if
> |
> | There are plenty of other conventions which could be used - for
> | example, the Borland libraries use
> |
> | TObject Object;
>
> If you fail to press 'T' all the way down then the same objection
> applies.
>
> To sum up this thread, any number of trivial reasons have been given
> why Unix is 'faulty' somehow. Whenever one is shot down with solid
> reasons westprog comes up with another mind-numbingly inane rant.
Fine, I'll talk to GAZZA, who can disagree gracefully.
J.
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Mike Kuht <[EMAIL PROTECTED]>
Subject: Multia/UDB Linux Installation
Date: Thu, 15 Apr 1999 13:16:37 +0100
Anyone out there got experience of installing Linux on a cheapy Alpha
Box ? My main problem is that I've got a Multia/UDB with a single 340Mb
internal SCSI drive and only a single port scsi card, how can I get a
Linux distribution booted up in order to do an install ?
First thought was to use bootp/tftp to pull in a standalone/memory based
system which could subsequently use say NFS to install linux from a
friendly host machine (I have a 486 PC on the network capable of
supporting this). However, I know bootp is supported on Multia/UDB, BUT
is such a Linux image/distribution available which would allow this ?
Any suggestions would be gratefully received,
Many Thanks,
Mike Kuht ([EMAIL PROTECTED])
------------------------------
From: "Osvaldo Pinali Doederlein" <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.java.advocacy,comp.os.linux.advocacy,comp.os.ms-windows.advocacy
Subject: Re: All the current OSes are idiotic (was Re: Is Windows for idiots?)
Date: Thu, 15 Apr 1999 15:32:15 +0200
westprog <[EMAIL PROTECTED]> wrote in message
news:7f4kpb$aid$[EMAIL PROTECTED]...
> How many files
> are there on Unix systems named like that? Hardly any, because it is so
hard
> to retrieve the file if you misremember the case. I just did a random scan
of
> my Solaris system, and guess how many files I found that used upper case
in
> any way? Not one. That is what a case sensitive system turns out as in
> practice - a lower case only system.
I agree! On my "lousy NT" I have tons of directories and files with
mixed-case names, my directories are cool and beautiful to read. Meanwhile,
when using Linux I see that people more usually recurse to hyphens to
separate complex names (e.g. gnome-session instead of GnomeSession) because
even the religious Unix users know that it sucks to use the shift key all
the time, so an hyphen is easier (a single keystroke).
------------------------------
Date: Thu, 15 Apr 1999 12:08:29 -0500
From: morgan <[EMAIL PROTECTED]>
Subject: Re: XServer for NVidia RIVA 128ZX
xf86 3.3.3.1 fully supports that card.
www.xfree86.org
Morgan
Sven Bohlmann wrote:
>
> Hi!
>
> I'm searching for an XServer for the NVidia Riva 128ZX because the normal
> SVGA Server supports just a resolution with 640x480 pix for this graphics
> card and the windows don't fit on the desktop. Please help me!
>
> tanks Sven
>
> ~~~~~~~~~~~~~~~~~~~~~~
> Sven Bohlmann
> [EMAIL PROTECTED]
> http://www.bohlmann.de
> ~~~~~~~~~~~~~~~~~~~~~~
------------------------------
From: "David Murray" <[EMAIL PROTECTED]>
Subject: Re: 2.2.5 Kernel is dead
Date: Thu, 15 Apr 1999 17:11:24 GMT
nevermind, I figured it out finally.. Seems I didn't compile in a console
display of any kind.. Wasn't used to "needing" this...
>
> Then the machine is dead as a can of spam. Never does another thing. If
I
> boot off of a floppy and change Lilo back to use my old 2.0.36 kernel it
> boots fine. I've tried recompiling the 2.2.5 kernel with different
options
> and still get the same thing. Any reason for this?
> --DavidM
>
------------------------------
Date: Thu, 15 Apr 1999 10:41:23 -0600
From: Bob van der Poel <[EMAIL PROTECTED]>
Subject: lost mixer with 2.2.5 upgrade
I just upgraded from a 2.0.36 to 2.2.5 kernel. All appears to be okay,
but my sound card isn't what it used to be. I can play midi files either
using the sound card, or an external synth. Audio CDs also play. I just
tried a .wav file, and no output at all.
The midi output does not sound very good.
I think the big problem is that the mixer is no longer working! I'm
using 'sounddesk', a tcl/tk program for mixing. Worked fine with
2.0.36...does nothing at all with 2.2.5. Well, it works, but the sound
volumes don't change (this applies to CDs and to midi).
I have checked my /dev/sndstat and it appears to be the same as the
2.0.36 output.
I have a gravis ultrasound.
Any suggestions as to where to start looking?
Thanks.
--
__
/ ) / Bob van der Poel
/--< ____/__ [EMAIL PROTECTED]
/___/_(_) /_) http://www.kootenay.com/~bvdpoel
------------------------------
From: "Ole Eirik Dalbye" <[EMAIL PROTECTED]>
Subject: ITK Colombus World ISDN adapter
Date: Thu, 15 Apr 1999 18:39:16 +0200
Is it possible to get the ITK Colombus World ISDN adapter to work in Linux?
Please help. Thanks alot!
------------------------------
From: "Kenneth H. Cox" <[EMAIL PROTECTED]>
Subject: Re: Travan TR4 / Seagate Hornet Tape Dirve
Date: Thu, 15 Apr 1999 10:30:44 -0400
David Graham wrote:
> You should be able to use make backups with something like:
>
> tar cvf /dev/ht0 {filenames}
>
> It was that simple for me, at least.
Do you know if it's possible to store multiple files on the
same tape? I want to use 'dump', and I expected that I could
do something like this:
dump 0uf /dev/nht0 /
dump 0uf /dev/nht0 /usr
And that appears to work, but then my attempts to use 'mt'
to move the tape to the start of the 2nd backup fail.
I tried:
mt -f /dev/nht0 rewind
mt -f /dev/nht0 fsf 1
Any suggestions would be appreciated.
Thanks,
Ken Cox
------------------------------
From: [EMAIL PROTECTED] (Piniek aka Piotr Ingling)
Subject: Re: Support for CD changers?
Date: Thu, 15 Apr 1999 17:54:36 GMT
Reply-To: [EMAIL PROTECTED]
Dnia Tue, 13 Apr 1999 17:15:55 GMT, [EMAIL PROTECTED] (Pete) napisa�(a):
>Hello all,
>
>I happen to have a Panasonic 5 disc CD changer. (IDE)
>
>I was wondering (other than treating it like a plain old CD drive)
>is there any other OS support for such a beast? Like the ability
>to switch cd's and automount?
>
Try cdchanger1.1 (it's somewhere on sunsite). It can change discs and takes
care of remounting them.
Piotr Ingling
e-mail: [EMAIL PROTECTED]
------------------------------
** 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
******************************
