Shadowman is a registered trademark of Red Hat, Inc.. We don't know what stormy
thoughts he truly has of the other - "red sky at morning, sailors take
warning; red sky at night, sailors delight", perhaps.
centOS plugs itself as the
Community Enterprise OS, and intends to be and remain
binary compatible with... ummm, a Certain North American Software Company.
No relation at all to the Taiwanese company CENTOS which makes internal and Cardbus
peripherals. I didn't know you could get SATA support in pc-card form...
yummy!
The "penny" I've offered for these thoughts is drawn by Heather Stern
using the Gimp, based on a photo of a real penny, a monitor from the
NeXTstep family of icons, the text circle script-fu, beveling tricks, some
gradient and alphamasking tweaks, and the word portion of the centOS logo
stretched and mangled by Curve Bend (found under Filter/Distorts) with
some correction of its outer edges by the perspective tool.
My friend silentk online (Pete Savage) provided the photographs of
UK cloudscapes. If you have need for photos for open source projects, he has
been taking photos for a couple of years now and has a considerable collection.
He can be mailed at pete at progbox dot
co.uk to make requests.
Tux is drawn by Larry Ewing using the GIMP.
Heather is Linux Gazette's Technical Editor and The Answer Gang's Editor
Gal.
Heather got started in computing before she quite got started learning
English. By 8 she was a happy programmer, by 15 the system administrator
for the home... Dad had finally broken down and gotten one of those personal
computers, only to find it needed regular care and feeding like any other
pet. Except it wasn't a Pet: it was one of those brands we find most
everywhere today...
Heather is a hardware agnostic, but has spent more hours as a tech in
Windows related tech support than most people have spent with their computers.
(Got the pin, got the Jacket, got about a zillion T-shirts.) When she
discovered Linux in 1993, it wasn't long before the home systems ran Linux
regardless of what was in use at work.
By 1995 she was training others in using Linux - and in charge of all the
"strange systems" at a (then) 90 million dollar company. Moving onwards, it's
safe to say, Linux has been an excellent companion and breadwinner... She
took over the HTML editing for "The Answer Guy" in issue 28, and has been
slowly improving the preprocessing scripts she uses ever since.
This was going to be a piece about a nice juicy stumper in
netfilters. However, that's been solved after all (see the answer gang
column about that) and the Gang working overtime has nailed all the medium
sized questions. So, I'm forced to reveal a stumper of my own - any
readers are welcome to chime in. If you've got something for me, send mail to
TAG... if it
inspires an article of some sort, check out our author submission
guidelines and contact the articles@ staff. --Heather
My hardware is a nice Pentium Coppermine 550 I've had for ages.
The Tyan motherboard I have is not quite up to match it - the CPU used to
be part of a dual pair, but that motherboard failed and the twins were split
up. By not up to snuff, I mean that I have 1 Gig of memory in it, which
would suit the CPU fine, and it works, but the mb is only rated for 768Mb.
Still, on 2.4.x kernels it has always been sturdy; on 2.6.x it
was crashy and I'd always end up going back. I tried 2.6.5, 2.6.8, 2.6.10.
I also wanted to get a few simple objects all happy at the same
time:
sound - Ensoniq 5880, uses the es1371 drivers.
USB - I use a small wacom graphire II. Also occasionally a thumb drive.
I used to scare bugs out of USB all the time by having my stuff plugged into
the hub on my monitor, and turning my monitor off at night. I don't do that
anymore. It eats USB numbered nodes up quick, and my monitor (a big LCD)
seems to prefer to stay on anyway.
CD burning - I burn a lot of discs. If this can't behave without
making coasters, I don't care how new the kernel is.
Ext3 - obviously this should work. Before people go leaping that
$other_journal is better, my criteria is being able to see the fs no matter
which rescue disc or test disc I'm booting - and when it's shutdown clean
this acts like ext2. However I do lots of dev work and my disks haven't
been small for awhile now; tack on a journal and I'm not forced into a
lunch break as punishment for my coding follies.
my low end Radeon video card. RV100 QY according to lspci. A measly 64
Mb. It's not up to the snuff of my monitor either, but the price was right
when I was eager to enjoy my new glory. It prefers framebuffer, but in the
2.4 series it drives the radeonfb bonkers, so I get stuck in vesa. I'd
prefer some higher octane in my video gastank
I get heisencrashes. At times it will lock up for apparently no
reason. After my speedy reboot, no disk farts in the logs, no shuddering of
a webserver process, no clue to the heart attack. Shall I blame the CPU?
Ah, but if it were failing hardware wouldn't it get worse? Or the kernel's
changes have no particular effect? But the newer kernels send the
boogeyman away - almost for long enough for me to believe that whatever it
is, they nailed it. Hooray! I go on my merry way hacking insanely huge
graphics in Gimp, compiling kernels or X or whatever in background while
juggling chat windows and pondering an article for some future
issue.
Of course it'd be handy if I could provoke it. For awhile I
almost thought I could... it was more fragile after I discovereed ayttm,
webcam support for Yahoo. But there'd be days on end it did nothing even
when I played with that heavily, others, pow kerblooey out
it goes. One night it was late and I left it at its console prompt (I don't
startx unless I feel like it) and it hung overnight. I haven't crashed in
awhile (a few weeks I think) on my 2.6.11 kernel - but I *have* crashed this
way on it, and I wasn't doing coding at the time, so I can't take the blame
myself for that one. The truth is out there -- somewhere.
Now the techsupport type in me says "and, Heather, what sort of
Doesn't Work does it do? Glad you asked. I'm buzzing along
typing, clicking, getting some work done - or goofing off - and the
keyboard's not working. I think my mouse slipped... but moving the mouse
doesn't get any effect. clicking it does toggle my wacom pad's signalling
light - but the computer isn't dealing with the interrupt. hyperspace. Ah!
I hope, I'll just ssh in from the sparc or my laptop. Nope, net's dead too.
Magic SysRq? (not that this is any help if I was in X.) nope. dead as a
doornail.
I'm going to be adding a serial console soon - as soon as I find
table space for some more cords. More ideas, and especially any idea of
what to look for are quite welcome. -- Heather
Suggestions on building an initrd in Debian without devfs?
Suggestions that he doesn't need an initrd if he's building
his own kernel won't be looked at, we already have those for him
I suspect he means building it the debian way.
I am a very novice administrator, who has had a Spam Assassin box thrust
at me recently. My knowledge is very limited with regards to linux and
though I have been searching the web I have not been able to find out
how to release quarantined email.
Hmm. Not much information to go on. A SpamAssassin box, eh? I assume we're
talking about a standard box with a mail server and SpamAssassin?
Which mailserver are you using? The interface to SpamAssassin can vary
according to the mailserver in use. What does the part of the configuration
for SpamAssassin look like?
Even the Linux distribution could be a clue? Do you know if it was installed
from your distribution's package manager or built from source to a local
configuration? If it was installed from the package manager, what packages
were used? If it was built locally how was it configured?
I have found the mail, have even been able to discern the contents but
not been able to figure out how to get it from there back into the queue
for it to be delivered again.
Again not much information. The location you found it in could be a clue.
I assume that you're working from a command line and not a web interface?
If you have a moment and could point me in the right direction I would
very much appreciate it.
It's hard to give you much of a pointer without more clues.
Have you tried googling for SpamAssassin and quarantine? It throws up a lot,
but I don't have enough information to tell what might be relevant to you. It
could be narrowed down a lot by adding the name of your mailserver (e.g.
sendmail or exim).
I've CCed this to the Answer gang. Answer gang discussions can be published
in the Gazette, so let us know if that's a problem.
Read your issue for March 2005. Came across the lines
Nero, the popular Windows CD/DVD burning software, has recently been released
for Linux. NeroLinux is a closed source application, and is available free of
charge following registration on the Nero website.
A Full Version of Nero Software Version 6.3 or higher
Retail Version or Downloaded Version
Please note: This offer is not for OEM or demo version users.
As an OEM user you can upgrade for a special discount offer if you register
your product.
...............
May I suggest you come out with a clarification for your readers.
Thanks! We'll be happy to publish this in our Mailbag; hopefully, nobody will
get tripped up by the original statement.
According to the review I read in Linux Format yesterday, it isn't any
good either
Their verdict was to use K3B. (Or perhaps Gnome Toaster
-- NeroLinux's GUI is based on that, so there's little difference).
-- Jimmy
I would also like to take this opportunity to say thanks for a great magazine
which I have been enjoying for the last two years or so.
Thank you very much for reading Linux Gazette! It's always great to hear
from our loyal readers and have them tell us how LG helps them.
I came across your magazine with the debian CD's ( Woody and subsequently Sarge
). Your magazine collections have kept me company ( Along with all the
Documentation available in Sarge ) while my ship has traversed most of the
oceans ( Atlantic North and South, Indian Ocean, the south China Seas,
Mediteranean and European waters).
Thank you all very much.
You're welcome - and we're glad to hear from you (the nature of
publishing is such that you hear more complaints than praise;
unsurprisingly, people are more ready to react to bad than good.) Thanks
for letting us know!
Capt Jasbir Singh Dhillon
[Heather]
You and our Editor in Chief would surely get along grandly -
he's sailed a bit widely, but I'm not sure he's gotten to every
ocean yet...
I've sailed some of those waters myself - the North Atlantic, as well as
a bit of the Pacific - and look forward to sailing the rest. Maybe
we'll even run across each other someday. Meanwhile, enjoy LG - and feel
free to let us know if you see any opportunities for improvement (this,
of course, applies to all of our readers.)
Fair winds and following seas - Ben Okopnik
Re: The article "Staying Connected" in issue 115 of the LG
When I read about your shell script called "google" I thought
that you might be interested in a program called surfraw
(http://surfraw.sourceforge.net). Surfraw does what your google
script does and it also does it for many different search
engines and types of search engines too.
Thanks, Grurp; I'm familiar with "surfraw", mainly through installing it
and then finding out that 'google' suddenly worked differently.
I
prefer my own version, since I've got my fingers 'trained' for it - I
don't have to think about how I use it, it "just happens".
However, our readers might appreciate hearing about it, so I'll CC this
to TAG.
I enjoyed reading your shell script tutorial - it's always good to remind
oneself of the basics.
I did however find one slight error - at least I think so. I've attached two
possible patches for it, I can't tell which of them (if either) is
correct...
[laugh] I guess I should just be glad that I didn't say "grab a seat,
this may take a while..."
Thanks, Ville; glad you're enjoying the articles.
-- Ben
Well, I guess that would require a 'while' loop, and perhaps an 'if' clause
since it says "it may take a while". But then I have no idea what the
'if' condition should be, and when (if ever) to terminate the 'while' loop...
I thought it'd just let you all know
that I am going to be without Internet access for at least a month.
Yow - poor Thomas. I don't know how you're going to survive, poor lad...
we'll have to snail-mail you some SSH modulus strings, or maybe the
contents of /dev/random, to keep you from falling apart completely. -- Ben
After that, I might have some, but it will be very intermittent, to not
at all, at best. Perhaps in October I will have something more
stable... I can't say.
We'll miss you! -- Ben
So it means my time working on LG is going to have to take a break for a
while. Obviously I'll do what I can as and when I get Internet access.
Thank you, all.
-- Thomas Adam
We'll try to keep the good ship LG afloat.
Thanks for all your hard work - much of it may be behind the scenes, but
I'm certainly cognizant of it. Have a great time playing in The Big
Room, and we'll see you when you get back. -- Ben
[Mike Orr] Farewell Thomas, we'll see you in the fall. Are you taking a walking tour about England?
Nothing quite as luxurious, alas. It's the end of the academic year
-- so the summer, although I am also having to move house for my final
academic year, which starts in October. Due to the logistics of things,
that's the earliest time that I can see myself being able to get
internet access.
Whilst I am at my parents though, they're by the coast --
Ah. The perils of student life
Have a good summer. -- Jimmy
Have a safe, wonderful summer, Thomas.
I look forward to seeing you in the fall.
-- Heather
This page edited and maintained by the Editors of Linux Gazette HTML script maintained by Heather Stern of Starshine Technical Services, http://www.starshine.org/
The Answer Gang
Linux Gazette 116: The Answer Gang (TWDT)The Answer Gang 116:
...making Linux just a little more fun!
The Answer Gang By Jim Dennis, Jason Creighton, Chris G, Karl-Heinz, and...
(meet the Gang) ...
the Editors of Linux Gazette...
and
You!
We have guidelines for asking and answering questions. Linux questions only, please.
We make no guarantees about answers, but you can be anonymous on request. See also: The Answer Gang's
Knowledge Base
and the LGSearch Engine
Greetings, everyone, and welcome back to the world of The Answer Gang.
When last you left the party, I was worried that we weren't seeing enough
good questions - and their answers. I'm pleased to say that things are looking up... even if it is my arms that are loaded up with this big armful. Thank you everyone, for writing - keep 'em pouring in... the fact is, while Real Life was busy waylaying me, the Gang's been going like, ummmmm, gangbusters.
So feel free to Ask The Gang anything linux-y you'd like.
#include really_cool_notes_about_CSS.h
Hmm. Those are about half done, since I didn't know if I wanted to make
them a blurb for you people or continue fleshing it out into a more
serious article. I didn't realize we had a history lesson on
soundcards going on over in the TAG lounge... all my coffee's gone and this thing still isn't pubbed - time to give it to you gentle readers already! Enjoy!
Open Solaris
From Jimmy O'Regan
Answered By: Ben Okopnik, Rick Moen, ILUG member Niall Walsh, Thomas Adam, Raj Shekhar, Heather Stern
Not Linux, but important because we'll probably be using some of this
software in a few weeks/months time: the source code of Open Solaris has
been released (http://www.opensolaris.org/os).
[Thomas]
Woooo.
Sun aren't finished releasing code,
That was June 15, keep an eye on their site as things fill in though.
-- Heather
but this release has most of the
kernel and userland tools. (No Thomas, nothing CDE related,
[Ben]
I'm teaching a Sun class right now, and the response from my students
was uniformly sceptical: "They're just trying to stay in the game." I'm
just going to note that Apple has been playing exactly the same game -
kinda-sorta release some-but-not-all of the source, maybe - and it
hasn't done very much for them.
Heh. Their FAQ has a nod to Apple:
...............
Do I need to register the OpenSolaris source code I have downloaded from
the site?
No. There is no registration. There is no click-to-accept license. Enjoy!
...............
Sun look to be more sincere about what they're doing than Apple--they're
using a licence that's already accepted (the CDDL is just the MPL
rebranded),
[Rick]
And improved, in my view. You may already have seen my related post to
the Irish Linux User Group mailing list,
Alas, no. Too many bounces in my last great disconnect, and my current
connection isn't reliable enough to justify resubscribing.
[Rick]
but here it is:
Quoting Niall Walsh (linux (at) esatclear.ie):
From there how long from there until we have a Debian GNU/OpenSolaris?
Or is CDDL even DFSG Free (the patenting clauses anyway makes me
wonder)?
[]
CDDL does strike me as being every bit as DFSG as its MPL predecessor.[1]
From what I remember of the debian-legal discussion, the gist of it was
"the licence may or may not be free, depending on how the licensor chooses
to act".
[Rick]
Ah, debian-legal. {sigh} A fellow on ILUG raised that, too, and my
response on that is further below...
[Rick]
However, of course, the OpenSolaris codebase apparently (as everyone
suspected) includes quite a lot of binary-only, proprietary "secret sauce"
components.[2] How useful the CDDL (and other open source / free-software)
portions would be without them is an open question. My own expectation
would be "Not very; probably even substantially less than the
corresponding case with Apple Darwin."
The only thing I remember seeing mentioned specifically is drivers, which
is understandable--too many NDAs. IIRC, Sun has a compiler that recompiles
Linux drivers for Solaris--device support isn't one of great plusses of
Solaris that Sun have been cheerleading in any of the blog entries I've
seen (DTrace this, DTrace that :)
[Rick]
Of course whether it's worth it or not is another matter. I
guess both will be projects like rebuilding a Freely redistributable
Suse/Novell Professional image, a test of just how many people care.
[]
I think the latter would be a great deal easier -- and more useful.
One of the big things in Open Solaris is that services can start in
parallel at boot time. Finito (http://web.isteve.bofh.cz/finito) does
something similar for Linux.
Heh. I think there will be a lot of public contributions to the code, if
only to clean it up. I never really understood why so much research goes
into code comprehension (well, aside from the fact that C lecturers write
some of the worst C around): open source projects need to have code that's
as clear as possible, otherwise noone can contribute. Heck, even Wine,
which (of necessity) has some of the most cryptic code I've ever seen
(self-modifying blobs of x86 binary code, C emulations of C++ exceptions
and classes, etc), is a model of clarity, as far as it can be.
Not so Open Solaris. Their tar implementation is one 182k .c file; their
nroff implementation uses the original Unix naming convention for files
(n1.c etc). I'm sure they'll get plenty of clean-up patches from people
cringing on Sun's behalf :)
[Rick]
[1] Sun Microsystems solicited my feedback on the licence before its
publc release. Unfortunately, I did not have sufficient time in to help
them during the very short comment period available -- but my impression
of the licence is overwhelmingly favourable.
instead of writing their own just barely open source by just
enough people's definition, but oh-so obnoxious licence for the purpose
(I suppose they've already done that enough times), and have put all the
released code in CVS from the start.
[Ben]
I didn't have much time to really look into this yesterday (but I did
get some really good coding done, and had the very satisfying experience
of solving a thorny problem that's been bugging me for a couple of
years!); today is a new day, and I'm wasting^Wspending my morning
checking out the buzz on this. Fascinating - it looks and smells real.
In fact, reading Tim Bray's blog and the link tree that branches out
from it - thanks, Raj! - I'm slowly growing convinced that Sun has
finally clued in, bought the stock, drank the Koolaid. All I can say, in
stunned admiration, is "Bravo" - I had not expected it, and was actually
rather cynical about it, having built my expectations based on what I
saw from my little corner of the Sun culture. It seems I was wrong.
The interesting bit about this is that the Solaris kernel has been a
strictly "hands-off" affair since time immemorial; you could tweak some
user-space settings to influence its operation in some mild ways, but
that was about all (and learning to do even that required a three-day
class and using ADB [shudder].) As well, some of the operational
algorithms - e.g., the exact process of deciding how to accept/reject
TCP connections - were generally hinted at but the details were
intentionally shrouded in mystery (security through obscurity, even
though Sun showed itself to be totally aware, in other places, that this
is not a useful approach.)
There's also the fact - and I'm not snarking at Sun in the least, but
applauding their moving with the times - that Solaris installations are
growing far fewer percentage-wise; this latest change may just be a
simple recognition of fact and adaptation to it. I've spoken to many
people in the large corporate culture, in many places and many different
companies, and have heard the story/complaint/simple fact repeated many
times: "so this new guy comes into IT, and before anybody knows it, he's
got 30 Linux boxes up in one day..." Cost of software: zero. Cost of
machines: hauled out of a closet where it was stored due to being old
and unusable. That's a hard combination for a commercial machine+OS to
beat, stony hard - and companies that _aren't_ having to spend a million
bucks a year on buying new are starting to notice. Many of them -
including banks, the traditional stronghold of conservatism - are
quietly dropping their "no non-commercial OSes" policy.
[Heather] I've done my share of dragging old Sparc hardware out of the closet too.
It runs a mite warm - and seemed happier on Debian/sparc than I'd hope
to put a modern Solaris on it - on the other hand, that SparcStation 10
was actually a hot commodity in a netlounge serving a few dozen people
at a time.
[Ben]
Sun is an excellent hardware company; the cost of their gadgets is
quite high as compared to the commodity PC, but the quality and the
reliability that you get from the stuff can be absolutely stunning to
someone who's never experienced them. I've never had Sun stuff "fight"
me the way, e.g., a PC SCSI card has; the documentations was always
available, and the relevant setup was as obvious as it could be made and
robust. Solaris, seen in that light, has always been a house divided
against itself - and Sun appears to have finally resolved the ambiguity.
Bravo; bravo indeed. The Open Source world has grown - and it's quite
the growth spurt. Not that we needed validation, but this is an
unlooked-for bit of big-leagues legitimacy that does not hurt at all.
Checking the debian legal lists it seems it may be DFSG compliant...
[]
In a better world, it would be possible to consult debian-legal (or
the various related "summary" Web pages) to reliably determine whether a
licence is DFSG-compliant. That fictional alternate debian-legal
wouldn't be populated by context-challenged monomaniacs woefully
ignorant of applicable law.
Ah well.
Suffice it to say that I draw a distinction between the concepts of
"DFSG compliant" and "approved by certain net.random wankers posting to
a rather painful-to-read public mailing list".
[Ben]
[laugh] A critical distinction in the world of Open Source, and one that
needs to be drawn far, far more often than it actually is.
[]
The "patents" comments are tautologically true -- but would be so
regardless of licence. That is, any codebase adversely encumbered
by patents is non-free/proprietary, irrespective of what licence
provisions would otherwise apply.
That other bit about fixed attributions making a work
non-free/proprietary is one reason why, although I'm a long-time
subscriber to debian-legal, I only rarely read it, in order to safeguard
my blood pressure: Author attributions may not be stripped in
derivative works by default action of copyright law , so it is utter
lunacy to assert, as poster Garrett and numerous others do, that clauses
to that same effect make the work non-free through it "failing the
Chinese Dissident Test".
That is a perfect example of the aforementioned problem of certain
posters being context-challenged and ignorant of the law.
You must include a notice in each of Your Modifications that
identifies You as the Contributor of the Modification."
So that the theoretical dissident programmer could distribute a
modification without having to admit to ownership.
[]
That licence requirement could be met by "Module foo.c was contributed
2005-06-14 by Anon Y. Mouse of the Fugitive Coders Group". The cited
clause's intent is not to require the equivalent of biometrics and a
mugshot from any contributor, but rather to distinguish cleanly between
original code and subsequent contributions.
And that is an example of what I mean by "context-challenged". ;->
Licences are not code for a Turing machine: They're designed to be
interpreted by judges, who (being modestly optimistic, for a moment)
have brains and are supposed to apply them to such things.
Yeah. I remember seeing similar assertions and having a definite pain in
the forehead after reading them.
[Rick]
[Proprietary carve-outs in OpenSolaris:]
The only thing I remember seeing mentioned specifically is drivers, which
is understandable--too many NDAs.
[]
Quite. Third-party rights. I remember seeing some other things, too,
but cannot remember specifics.
I don't mean to derogate the benefits to some of Sun's move -- just to
remind people that OpenSolaris's carve-outs mean it's doomed to be
crippled as an open-source operating system, e.g., for purposes of porting
performed by anyone but Sun themselves.
[Ben]
I tend to take an optimistic view of these things that look like naivete
from a certain perspective - any move in the direction of Open Source by
the historically-proprietary software companies is good for us. One of
the many underlying reasons for that is what I like to call "time in
grade": the longer these people spend using Open Source methods - to
whatever degree - the more these get embedded in their culture and the
environment around them. After a while, they're impossible to eradicate.
As for "the abyss looking back at you", I rely on the GPL and its ilk to
keep the resulting influence on the world of Open Source to a tolerable
minimum.
(Yes, the sum total of all the influences and issues here is far more
complex than this. However, abstracting this bit and staring at it for a
while contains its own lessons, and they're interesting ones.)
Well, though Sun want to hype up the release by saying "Solaris is open
source now", their own roadmap says otherwise.
Another tried his hand at predicting the future of system speeds. "As of
this writing (1996) a clock rate of more than about 10 kHz seems utterly
ridiculous, although this observation will no doubt seem quaintly
amusing one day," he wrote.
Religion was a common theme in the code. "Oops, did not find this
signature, so we must advance on the next signature in the SUA and hope
to God that it is in the susp format, or we get hosed," said one developer.
"God help us all if someone changes how lex works," wrote another. "Oh
God, what an ugly pile of architecture," moaned a third.
...............
Booting a "Live CD" image without a CD
From Ben Okopnik
Answered By: Kapil Hari Paranjape, Robos, John Karns, and a very useful webpage by
Matthias Müller
Hi, all -
Got a curious Linux problem here that I'm trying to puzzle out, and
after struggling with it for a bit, I remembered that I'd heard of this
thing called The Answer Gang... :)
I'm trying to boot Linux on my fiancee's laptop, a Sony VAIO F590K -
something that she'd be quite happy to see, since her opinion is that
Micr0s0ft should have stopped when the going was good - i.e., DOS5.0
Simple, right? Uh, well... the only problem is that it's got a dead
CD-ROM drive. She's going to order a new one soon, but until then...
The VAIO doesn't support booting from USB. However, I've managed to load
Puppy Linux onto a 1GB USB FlashDrive and burn the appropriate disk image
(provided by Puppy) to a floppy - it's an ingenious system (the floppy
boots FreeDOS, which searches for and boots the FlashDrive) that could
probably be easily adapted to boot other distros... if I only understood
exactly what to tweak and how.
I'm afraid that I've met my match (at
least for the moment) in trying to understand the whole shebang.
I've looked at many LiveCD distros in the past few days. A number of
them can be run from USB - but require that the machine boot from the
USB, not an option here. I've even carefully studied the "Booting
Knoppix from USB" HOWTO, which assumes the same thing, to see if I could
somehow mingle Puppy's floppy boot and Knoppix on a USB stick... no
luck.
[Kapil]
Use Knoppix boot.img on a floppy and copy the KNOPPIX directory to the
USB stick. This should work provided the kernel+initrd on the boot.img
supports USB---I think it does but there may be a kernel boot option.
The Knoppix boot mechanism is:
Recognise possible hardware where the KNOPPIX hardware may reside.
Look on all block devices for the KNOPPIX directory and under it the
Transfer control to the cloop stuff. The remaining hardware detection
I think puppy uses Xorg and also possibly the vesa driver only. You may
have better luck with Knoppix.
Oh, and PXE booting is out as well: the F590K does support network
booting... however, PXE does not (yet) speak PCMCIA.
So, given all of the above - what do you folks think? Have any of you
had experience in booting something like this, or do you have any ideas
that I've perhaps missed?
Argh. So much for writing email while talking on the phone and being
derailed in the midst of it all by questions about tea selection (from a
large number of options, I might add - Kat and I are both heavily into
tea. I think I'll try her kelp tea this time... or maybe the
hibiscus...)
Puppy failed to recognize the video hardware in the VAIO. Key factor
I neglected to mention.
Ben tries Kapil's knoppix solution, but...
-- Heather
Ooops. Seems like Knoppix stopped using "boot.img" at v3.4 (I've got
3.7) - they use "isolinux" these days. There seems to be a bit of
discussion on the Net that mentions using the "boot.img" from the v3.3
CDs - I've seen reports that it Just Works - but I don't have one
available, and won't be able to download 3.3 until I get to a high-speed
connection. Would anyone here happen to have such a thing handy?
If someone happens to have a Knoppix 3.3 image, let me save you a bit of
time (obviously, you'll have to change the ISO image name to whatever it
actually is, and "/mountpoint" to some existing directory that you don't
mind using as a mountpoint for a few seconds.)
# Mount the image
mount KNOPPIX_V3.4.iso /mountpoint -o loop
# Flip the image to me
mutt -s "boot image" -a /mountpoint/KNOPPIX/boot.img ben@linuxgazette.net
# You're done!
umount /mountpoint
I knew there had to be a reason I kept that old thing around. Actually,
at first it was paranoia, but then I realized that Klaus Knopper's push
to support the bleeding edge hardware from his pocket disc was giving
plainer PCs headaches. So now I try to keep the spare older versions
too. For the record the image linked above is from Knoppix 3.3,
2004-02-16, english edition, with an md5sum of
a761779c73e01185585e879c800ddede.
-- Heather
(re hardware detection) Man, this sounds very cool. I'll be very happy
if it works...
I'll definitely report the results as soon as I've had a chance to try
it out. For now, it's after midnight, and I'm off to bed.
Didier Heyden actually got Ben his diskette... and sent its md5sum with
it, just to be sure.
-- Heather
[laugh] I should have asked that people respond on the list first; I've
had at least three people send me the images, and at ~2MB apiece, my
poor little cellphone connection is groaning. Thanks, all who responded.
[Didier]
Let us know how it goes!
Not too wonderful.
The floppy boots the Knoppix miniroot, scans a
bunch of devices in /dev, and says
Can't find KNOPPIX filesystem, sorry.
Dropping you to a (very limited) shell.
Press reset button to quit.
However, examining /modules on that miniroot image gives me a clue: it's
full of SCSI modules but no USB modules. I'll "unwrap" that image
file, yank the miniroot out of it, mount that, and see about replacing
the SCSI modules with the USB ones; hopefully, Knoppix will be able to
recognize the USB FlashDrive as a result. The only thing I wonder about
is if the scanned devices in /dev need to be specified somewhere... I'll
look in the various .cfg files to make sure.
This does sound quite promising. Thanks for the hints and the help,
everyone - I'll report more as I know more!
I have now "rebuilt" boot.img - a very interesting and educational
process, actually (in fact, I've hacked a script that I found on the Net
to do it all for me, since I had to do it repeatedly.) However, I have
left the USB pendrive with my fiancee for the week - she's in New York
while I'm teaching a class in San Francisco; I have no floppy drive on
my machine, and thus no way to experiment with it until I get back.
Thank you, everyone, and Kapil in particular for your help with this -
I'm beginning to feel like I actually understand that booting process
(as contrasted against the "standard" one.) Again, I'll report the
results as soon as I have some - that will be sometime after this
weekend.
[Robos]
Well, if you only want to show her the livecd, how about you use and emulator?
Lately I've been using qemu a lot and it works like hell! Fabrice Bellard is
really a god. Ah, and if she has winblows on that machine, how about this
link:
Ah. BitTorrent, I'd guess - which I know almost nothing about.
[Robos]
Short intro into bittorrent: data is packed into small chunks. You download
a chunk on your machine and the next user that wants also this chunk
downloads it from your machine, thus relieving the main server of the high
load since the load is spread over all downloading (and uploading) clients.
The download doesn't have to be sequential, packets come in from different
downloading and uploading parties. Even after you got everything your
machine keeps uploading for others - this is one thing to watch out for.
Presumably, this would download a Knoppix ISO onto her machine? That's
not a problem; our laptops are Ethernetted together. Although it's good
to see that 3.8 is out (<clickety-click><download>...)
As we go to press I have a Knoppix 3.9 DVD here. Mind you, a lot of
machines for "just trying out" might not have a DVD drive to boot from.
Time marches on :)
-- Heather
Hey, it's got exactly the same message as I was getting!
[Robos]
and the folks that made the .bat file made some errors IMHO: -m should be
more in the range of 128 (smaller than what shm has, whatever that means on
winblows) and I think the error message in the pic also came from an error
of theirs but hey, it's close.
Well, I guess (haven't looked at this again) that the japanese knoppix
developers that made this image made some config error somewhere. I mean,
you essentially have the knoppix image, it's only still on the cd... And you
have qemu. On linux this works then like this:
qemu -cdrom=/dev/cdrom -boot d
and there you go. Where is the device under win? That's the problem there.
You did notice that I ran the knoppix in an emulated win session? :)
BTW: qemu with kqemu (the kernel module he has now on linux) gives really a
speedup to ~70% native cpu speed. Awesome (that's why win4lin pro uses qemu
underneath)
Helps?
Cheers
Robos
I'm not sure - but I'll be back in NYC this Saturday, and will have time
to play with computer stuff a few days later, after we get back to St.
Augustine. But I'll give it a shot if the current pre-programmed
solution (the modified Knoppix boot floppy, the preparation of which was
so superbly aided by Kapil's suggestions) doesn't work. If it does
work, I've built a great script that automates the somewhat painful
process of creating it, and I'll definitely publish it as well.
Bellard was born in 1972 in Grenoble, France. He went to school in Joffre
(Montpellier), where he already created a widely known program, the
executable compressor LZEXE.
...............
Ahh, that's it. He must have been all of 17 or 18 years old when he wrote
that, as I remember it from 1989 or '90. I think he sold the idea to Phil
Katz who incorporated it into his product. It was an algorithm that would
uncompress a DOS executable, transparent to DOS, after being loaded by the
program loader, before being executed. Handy for saving space in the days
of 40 MB disk drives.
and look at his projects - you can't be anything but
impressed to the max. Winner of obfuscated c contest and such. Also take a
look at his "old projects page". Such people are inpiring to me.
Answered By: Thomas Adam,
Heather Stern,
Jay R. Ashworth,
Kapil Hari Paranjape
I compiled playmidi-2.5 from source code, using a version of linux
(Basiclinux 3.3) based on Slackware 4.0 (libc5) with the Slackware 7.1
kernel and modules (kernel 2.2.16) because the source code would not work
with glibc2. A version I downloaded from somewhere for Slackware 7.1
(2.4.9) told me 'No playback device found'. I have no midi device other
than sound card and speakers so I typed playmidi -f for FM synthesis
because -e for external midi device is default. 'No playback device
found' either way.
I am using OSS sound, sb.o, Vibra16 AWE32 card which works perfectly in
DOS and has two 4MB 30-pin SIMMs. My DOS players appear to load sound
libraries automatically.
To compile without X or GTK or svgalib I commented out the lines in the
Makefile referring to splaymidi, xplaymidi, etc. because there was no
--without-X type option in the configuration program. I configured to use
SB AWE32 as the default device. I installed the developers' ncurses
package when it complained about not finding the right ncurses files. It
appeared to compile.
I am using the bare.i (basic) Slackware kernel, or a kernel which I
compiled for sound myself, and have inserted:
I can play wav files with sox and mp3 files with mplayer.
I also typed mknod /dev/sequencer b 14 1 and mknod /dev/midi b 14 2
(my distribution came without sound devices and I previously made dsp
audio mixer and sndstat to play wav and mp3 files with).
I tried to use playmidi in Basiclinux 3.3 where I compiled it dynamically
(it needs libc5 and libncurses, both there) and if I typed playmidi -f
(for FM synthesis) test.mid 'No playback device found'. Without the '-f',
using the default AWE32 which I compiled for, I presume, it gives me a
message about AWE32 and the author and looks like it should be playing but
there is no sound.
I put the FM synthesis all the way up with aumix.
What am I missing? The Makefile is attached.
My package (including the libncurses dependency) is at:
Pnpdump(8) comments the file rather well, so you can definitely try it
and see.
I was told I also needed to edit the resulting file.
[Thomas]
Maybe. Maybe not. pnpdump is intelligent in its detection, and will
often leave reasonable defaults uncommented. But yes, do obviously look
at it before hand.
Do I really need
to do this if I can figure out the settings by trial and error, or from
what worked for other people? I can tell which IRQ and DMA the sound card
is using when I boot (or with a diagnostic program).
[Thomas]
Which diagnostic program is this? They're not that reliable. Have a
look at using 'modinfo' on the module,
hwinfo seems to be reliable (DOS shareware, Slovak) for irq and dma
[Thomas]
As is the Linux equivalent.
Since I boot into linux with loadlin from DOS, the DOS version works best
for me.
[Thomas]
That doesn't account for accuracy, of course.
SB16 has so far identified settings perfectly - if they work in DOS, they
work in Linux. Except for one time when DOS needed IRQ2 and linux needed
IRQ9.
[Heather] Otherwise known as the "cascade interrupt" - irq 2 is irq 9, the
manufacturers used that trick to extend the original bank of interrupts.
Someone must have thought we'd never need 8 devices at once the way we'd
never need 640k for one program to abuse...
So is the DOS SB16 diagnose program - if I use the same irq and dma and
mpu_io for linux it always works (with insmod sb).
[Thomas]
You ought to be using 'modprobe' (insmod only loads the device you tell
it to -- modprobe loads dependant devices on it). As for DOS diagnoses
programs, they're iffy at best -- and not always is it the case that
IRQs are transferrable.
The first question I ever had for TAG concerned where to find modprobe so
I could modprobe mdacon. Our linux has only insmod. I was told there is
a list of module dependencies somewhere, and I figure it out either by
doing web research, or looking at the error messages and guessing which
modules will fix them (similarly named to parts of the messages).
[Thomas]
I don't recall this question. Mind you, if modprobe is being
distributed separately from insmod, and modinfo, then that's in grave
error, IMO.
It was two years ago. I wanted to use a 2-floppy disk linux with TTL and
VGA monitors. The discussion veered off into suid and other things that I
had never heard of in my two weeks of linux experience.
[Thomas]
Welcome to TAG. "The answers themselves may range from friendly to
gruff and often contain sharp humor, horseplay, fluff, pedantry, and
pontification; the discussion spawned by your question may well wander
off-topic and possibly back on again - with all of this reflected in
your mailbox. Thin-skinned folks, those who expect "their answer" and
nothing else, and narrow-minded people are urged to take the appropriate
precautions. (Yellow helmets and fire- and bullet-proof underwear are
available in the shop just off the main lobby.)"
Syschk is hopeless - it told me 6 IRQs were free when only one really
was.
[Thomas]
as some allow for options to be passed to them:
io=..., irq=...
The sound cards (at least for mad16 and sb) let you specify settings, as
do ISA ethernet cards.
[Thomas]
OK.
For one sound card
(an opti) I needed two settings pertaining to midi (mpu_io and mpu_irq).
Is there some way to figure them out without isapnptools?
Is opl needed for awe32 or just for FM synthesis? He also said my FM
synthesis is broken - will awe32 work without it? I have four awe32
cards, which I think are all plugnplay.
[Thomas]
Yes, it will. FM Synthesis is optional.
My Makefile was created by removing the parts of the standard Makefile
that I thought were irrelevant. I probably removed too many pieces
[Thomas]
More than likely -- when you embark on this, start from the centre (the
'install' directive, and work outwards.)
Thanks. The precompiled package I found for playmidi does not seem to
support AWE so I attempted my own. I was unable to compile the source
code with glibc2 (I got an error message about GLIBC2 and atexit) but it
worked with libc5.
[Thomas]
atexit and glibc2 (that is libc6) is a classic case of "oops, this
version of glibc is not going to work. libc5 is fine and will coexist
with libc6 just happily. I have a number of motif apps that are
dependent in this way.
I put libc5 and libm and ld-linux on my glibc linux, and glibc2 on my
libc5 linux, to provide a wider choice of programs.
[Thomas]
You can just say "libc5" here, as I know what it contains.
Some things (like
Opera) don't come in source code so you are stuck with their choice of
library, meaning I also had to upgrade glibc2. The problems occur when
they have dependencies other than libc (libpng, for instance) and you have
to specify to use the libc5 not the libc6 version when you have both of
them. I have a little script for that.
[Thomas]
This is not good, as you will get stuck at some point. Libc5 is not
going to handle newer versions very well, and there will be a point when
it will fail.
I have been using libc5 mainly just to run programs that I don't have
working with libc6, such as this playmidi, and also a few other things
that our basiclinux list members have compiled for libc5, or specifically
for the setup we have which uses it (files are in sort of odd places) such
as Abiword. I can chroot between the two linuxes to wordprocess but
playmidi ought to work in both since I added libc5.
It is old source code, but I thought that problems
arose only when you tried to compile newer source code with an older libc,
not vice versa.
[Thomas]
Not always.
I will check with modinfo as to whether I can specify settings for opl
which are the same as for sb. I will have to install modinfo first - our
linux has only the essentials (the kernel did not even support sound).
[Thomas]
Interesting, as modinfo is part and parcel of insmod and modprobe and
the like. An omission of it is surprising...
Our 'basiclinux' started as a 2MB download, with just insmod. We have
busybox versions of many things which don't always work as required, and
have to install the full versions. I just pick out a file or two from a
package, to keep things small. We also have Xvesa (about 1MB) instead of
a larger X. Sometimes these busybox versions are what keeps things from
compiling. Playmidi may not have liked my minimal X, which is why I
removed it from Makefile.
[Thomas]
I'd say. This "picking and choosing" process is... an interesting idea,
but perhaps a bit too thorough for some things.
Yes, the author of our linux thinks I am crazy too, but I prefer to just
install the things I am actually using. If they have dependencies, I have
to install those too, and make symlinks, but I learn a lot that way.
Sometimes I install the whole package and then delete most of it, for
instance SANE supported many scanners and I deleted all but HP, along with
networking stuff.
The neighbor suggested it but I don't know enough about compiling.
Playmidi was one of my bolder adventures (modifying the Makefile
manually).
[Thomas]
There's an "instruction" manual weighing in at 300tonnes.
Can I delete most of it after compiling? It would take me a few years to
download.
I could add some package to get modprobe and modinfo but it is more fun
figuring these things out on my own.
[Thomas]
I'd go with the quickest option -- it's something you need, and not
something you can fool around on a whim as and when you come across it.
I would rather understand things than be quick.
[Thomas]
I know nothing about this distro -- but a distro as minimalistic as that
which gives one the option of not having modprobe as "standard" is
fundamentally flawed.
He was trying to fit into 2 floppy disks. We can add any Slackware 7.1
packages we like after installing to hard drive, but he suggested the most
basic ones. I also deleted 1MB of locale files in the first 11MB
installed to HD.
[Thomas]
Why the brutual stripping? Surely your hardware can't be that
antiquated? Or is this just for fun?
Why would I want 1MB of locale files? It just takes longer to find things
I do want on there and I don't speak Korean. Or 50 copies of LICENSE or
COPYING or TODO? I like to only keep things on the computer that I know
I am using.
[Heather] You'll be pleased to know that Debian policy points at only one copy of
the standard licenses, with a notation in /usr/share/docs that's much
shorter, or a whole license if an app is special. Also it's easy to set
up your incoming sources so you don't use the contribs and non-free,
thus avoiding the more curious and/or onerous license types.
Several of us put together a little script to convert man pages to html
and display with a browser, but lynx does not display some characters as
intended.
[Heather] Lynx also makes a fine pager over plaintext files, if you don't mind
also giving up boldness. If you set your TERM=dumb and then use man to
dump your pages (redirected out to a file of course) then the formatting
is stripped very quickly without resort to scripting tricks and missed
characters. We did this for LNX-BBC... and then found that many man
pages were superceded by the --help the programs had buried in them.
Scalpel time
<IMG SRC="../gx/dennis/smily.gif" ALT=":)"
height="24" width="20" align="middle">
I do find myself wondering if you'd enjoy trying an LNX BBC for
your baseline though. Everything was from scratch at the time and are
libc6 binaries.
It saves 8MB of groff and man programs. We have Xvesa with
only the fonts and files needed to run it in an rxvt (all this in a linux
that fits on 2 floppy disks), but Xvesa has a few problems. If we run
into problems we install larger versions or additional files. There was
no sound or scsi support so I learned to compile my own kernel, and
removed support for some things I don't use, to keep it small (around
450K), such as firewall and samba. I included awe support in the kernel.
What would I have learned if I just used a standard distribution and
kernel?
Our star compiler put together a static (libc5 or libc6) mplayer (7MB
executable) but it won't play midis. We can run linux on a 486 with 16MB
RAM and 100MB hard drive if we don't need to compile or use Mozilla.
Yes, SuSE is one of the more reasonable distributions, in fact the only
other one besides slackware that worked on our 486s. But it still
installed itself to run processes at regular intervals, and took much
longer to boot because it assumed you had new hardware and went looking
for it. Our little linux boots in 14 seconds on a 486.
[Heather] Now I know you didn't try debian; I had 386 and 486 hardware running on
it fine. Though admittedly I haven't tried to install Sarge. However,
every installer based system has a kernel that goes looking for the
kitchen sink - it can't know what your hardware is until it looks. So
the first "high tech" thing people do at installfests is build their own
kernel to get that zip. Though it matters a lot less if you're not
going to reboot your machine constantly, about shaving a minute off of
your boot time. Linux is more suited for being always-on.
I went looking for awesfx, which is a set of utilities also containing
sfxload. I found the author's site, with his comment that he had
switched from bz2 to gz because of download problems, but the source
code was bz2 and the download was corrupted. I found the binary package
at linuxpackages.net - for SW10. I found nothing of the sort for SW7.1 or
SW8.1. Any ideas where to look? I don't want to upgrade to the SW10.0
glibc.
[Thomas]
No idea, sorry. I'd only be doing the same thing you are, via google.
I learned that SB Live also accepts sound fonts (soundbanks) but they seem
to be different ones from SB16 AWE32. And that SB AWE64 still does not
work in DOS (or probably therefore in linux) because awe will still not
initialize even after the card starts working with standard settings.
[Thomas]
It is a very old card...
The awe64 is the newest card that I am using. I have mostly 1992-1995
SB16s with jumpers. They sound nice. Most of them have four jacks (line
and speaker are separate) and a couple have volume dials.
[Thomas]
I'm well known for having and running antiquated hardware. I also know
when to give up and concede that there are some things just not worth
the headache. That said, you still have some outstanding issues to try
-- and I'd probably push with the pnpdump(8) idea.
I first need to get hold of uncorrupted source code for the sound bank
loader and see if the DOS initialization was adequate. I wrote the
author. He may have posted the .tar.gz package at Creative's opensource
site, which always times out.
I might give up on four Opti 929A/Crystal cards, but SB is under control
and our 'distribution' even has a package for it.
I am
considering PCI mainly because we are running out of ISA slots - MGA card,
modem, sound card, ethernet card. Only the sound cards seem to be pnp.
I also learned that I can set my ISA SB16s to use standard settings by
disabling USB in CMOS, running SB16 diagnose in DOS (which no longer
complains that IRQ5 and DMA0 and DMA1 are taken), then reenabling USB (on
one computer - on the other I have to leave it disabled or the conflict
comes back). Would a PCI sound card avoid this problem?
[Thomas]
Anything this side of the late renaissance
(for which, alas, your poor SB16 card does not) would, yes.
What sort of sound cards did they use in harpsichords?
Someone just gave me a 'Live'.
I may write the author of awesfx to ask where to get a .gz version of
source code.
[Thomas]
Remind me again why bzip2 is not preferrable? If it helps, I can always
convert it to .gz for you and shove it on my webserver.
The download gets corrupted. I even tried downloading to a local bbs with
a fast connection and bunzipping it there - premature end of file. We had
the same problem when someone posted a .bz2 package for our group. I used
lynx to download and other people downloaded with other programs.
awesfx*.tar.bz2 (version 4-4 or 4.4) if you want to attempt the download
and bunzip.
I will search for an rpm binary of this package.
[Heather] Debian contains version 0.5.0b-2. The perl script alien (which I highly
recommand for anyone's sysadmin kit) could turn it into an rpm.... or
even into a slackware package, which it looks like your box would prefer.
But the important thing is that debian has a well scattered mirrors
system, and its source package contains an orig.tar.gz as well as their
diffs, and you can read the maintainer's changelog to see what the
diffs were done for. I've used the trick of fetching debian parts for
other distros before - though I mainly deal with the mainstream ones,
your little distro should be served easily as well. There is even an
archive server if you insist on hunting down older packages.
There are versions of source code and binaries for ASP 7.3, Alt 2.2, Alt
Sisyphus, PLD 2.0, TurboLinux .0, SuSe 8.1, Redhat.2, and later (and other
linuxes). If I am just after source code, does it matter which version of
linux I get the source rpm for?
[Thomas]
No, since you'll be linking to whatever versions of libs you have on
your machine. Yes, the program might complain that you need a newer
library version or whatever, but that's not too much trouble.
I don't know much about compiling so it would be easier if I did not have
to specify or modify details about libraries.
I would rather use a precompiled binary but I don't know which glibc all
the rpm linuxes use, except that Redhat 7.2 is glibc2.2.x and I have
upgraded from SW7.1's 2.1.3 to SW8.1's glibc2.2.5. (Never mind that
playmidi is compiled for libc5 - I will use it in a linux which contains
both lib5 and libc6). Is there some site that lists the libc's used by
various versions of linux?
[Thomas]
Nope, not that I know of.
I don't care which directories the packages
put things in as I am just planning to take the package apart and extract
the few pieces I want.
I can use rpmtotgz to convert rpm binary packages to slackware binary
packages - will this also work for the source code? I know how to extract
Debian .tar.gz with ar +x (and keep the larger of three files produced).
One of our members explained how to extract tar.gz's from rpms using a
complicated little script based on a small utility (that I picked out of a
larger package), which I tried on a playmidi rpm but it failed (corrupted
download?).
[Thomas]
Really? rpm2cpio is just fine for all cases. And of course, there's
"alien".
That may have been the utility (rpm2cpio).
You don't even have to wander from my original subject, I have done it for
you. Minimal distributions are so much more educational.
I will give this a try with a soundbank some time soon and then if needed
look into modinfo and isapnp.
and
therefore do not have FM synthesis. I compiled my own when a precompiled
version did not work for me as FM synthesis (due to opl being missing).
It would not compile before I edited Makefile because I did not have GTK.
I only wanted the CLI version of playmidi, not X, GTK, or SVGA.
Apparently if you boot with loadlin from DOS, you can initialize a
plugnplay sound card in DOS and don't need to bother with isapnp. On three
computers with awe cards, the one with awe64, which does NOT play midi
files with an awe midi player in DOS (but does in Win98 in a DOS window),
informs me 'awe initialization failed', which implies it worked on the
other two.
What I am missing appears to be:
I might need to add to the insmod sb line mpu_io=0x300
From awe32 tools package I definitely need to add the actual sounds:
This is probably why it appeared to be playing but no sound came out -
the sound bank was empty so that is what it played.
For FM synthesis I need to use a version of playmidi that was properly
compiled, and insmod opl, possibly initializing it with isapnp.
Can someone explain how isapnp and plugnplay work and what initialization
does to a pnp card? The person trying to help me says it is very
complicated. He got it working with FM synthesis and one Vibra16 card.
Time to wander again. The computer with the AWE64 card, which was working
fine in Windows and almost fine in DOS and linux (no midi, would not
initialize awe), suddenly stopped being able to boot Win98FE after I
changed the ethernet card (to another PCI model) and put in an internal
PlugNPlay modem (which is the only other plugnplay card in there) because
I needed the external modem for a computer with fewer ISA slots. Removing
the modem did not help, nor did disabling onboard USB (which uses an IRQ,
and which helped previously to get sound to work, and I could then
reenable it and sound kept working).
Removing the ISA sound card fixed the problem. With PCI sound card (SB
Live, of which I now have two) Win98 boots. I don't have drivers for it
for DOS, linux, or Win98 and will have to work on that problem for a
while (it is not SB16), or figure out how to get a SB16 ISA to work in
there again, maybe a different one, before continuing with playmidi and
awesfx. My other computer is in pieces.
I have no free IRQs. The DVD-ROM drive can't even find one and disables
DMA (I was told there is a way to do this with hdparm but maybe I need DMA
to play DVDs?). I am using a SCSI CD-ROM drive to read CDs. I put a scsi
card in here, and ethernet, and ISA modem, MGA (disabled the parport to
free up an IRQ) and sound. The PCI video card also wants an IRQ. So to
free up IRQ5 for the ISA sound card (instead of IRQ7, which it took away
from the printer port) I disabled USB onboard, then was able to run DOS
SB16 diagnose, which reset the ISA sound card to standard settings, then
when I reenabled USB the sound card still worked with standard settings.
Is there a better way to do this? I have hwinfo, which tells me which IRQ
and DMA settings are being used by what. I will keep hunting around the
CMOS setup for ways to set IRQ and DMA to be used by legacy devices but
this is not the BIOS I am used to (it has little icons - AMI instead of
Award or vice versa).
Do some devices also work even without IRQ and DMA and can I disable their
use of them manually in BIOS?
[Heather] Parallel ports can use a polling mode that is slower. If you have both
IDE and SCSI devices, I'd see if you can disable the other entirely and
stick with only one drive interface. Beware that some 486's can only
boot off scsi hard disks not scsi CD since they expect a bootable CD to
be atapi - that's assuming they can boot from an off-board controller at
all. I suppose you need the floppy in your conditions - it uses an IRQ,
too, but could be your boot device - even just a master boot record can
fit on it nicely.
I am thinking of trying different SB16 cards in there before tackling the
'Live' drivers. I still have two jumpered SB Pro's.
No wonder our little linux came without sound support.
It was not necessary to initialize any of the ISA SB sound cards (or two
clones) in linux (they were being initialized in DOS), or the ISA Opti 931
(which pnpdump said is not pnp). I hope to avoid isapnp for a while
longer. The modem (pnp) was working in another computer without
initialization, as com2, just like a jumpered modem would have worked.
The author of awesfx steered me to the .tar.gz at
www.alsa-project.org/~iwai.
Version 5 includes ALSA and version 4 does not and it also does something
else differently which might be better with older cards.
I just acquired a second Soundblaster Live, which also accepts sound fonts
using this same program. Win98FE, which suddenly stopped booting with the
AWE64 ISA card (after I changed PCI ethernet cards and added MGA), boots
with this one but the driver needs SE not FE (which would imply they don't
support DOS). Linux OSS worked perfectly (soundcore.o and emu1k10.o - ?)
and this PCI card also cured my problem of the DVD-ROM drive not finding a
free DMA and IRQ.
So I am set to compile awesfx and test it on a card which I am certain
does not need isapnp because it is not isa, and then on an AWE32 with and
without initialization in DOS, to see if it really needs to be DOS or
isapnp initialized. Supposedly the sb driver initializes the card itself.
The mad16 definitely did so for the opti 931 and a cs4232 appeared to do
the same (but no sound came out).
Ubuntu linux did not find our ISA ethernet or sound cards.
[Kapil]
As far as I know "automatic" detection of ISA cards is fraught
with problems. You need to "know" which driver to load and
occasionally you also need to know other parameters (irq, ioports
etc.).
I replaced the ethernet card with PCI. The ISA sound card is a Crystal
which I could not get to make sound when I manually insmodded with what I
hoped were the correct parameters, but someone said to try aumix in case
it starts with 0 volume. Ubuntu is not my main problem, just tried to
test a sound card with it. It came from a friend and probably works.
[Heather] Crystal sound isn't an emu10k1, that'd be a module in the sounds area
named cs(some numbers). Probably cs4232 since I find it shown in more
of my kernels, but cs423x or cs4281 are possible. It won't be a cs46xx -
that's a PCI card.
[Kapil]
In your case, since you "know" that you have an SB Live card,
you can do "modprobe snd-emu10k1" and that should be enough.
The SB Live card is since yesterday working with OSS in a different
computer in a small Slackware-7.1 based distribution. insmod soundcore
insmod emu10k1 did it. Replacing the SB awe64 ISA card with the PCI card
also let Win98FE start booting again, and fixed a problem with the DVD-ROM
drive, which no longer has to disable DMA access or wait for IRQ to time
out. I had three ISA cards in there, now only two. And PCI video,
ethernet and scsi. I think the awe64 was grabbing DMA 0 and/or 1, which
the DVD-ROM drive needed. I am losing track - at one point I also had an
awe32 in there.
[Kapil]
After that "sfxload" should be used to load the wavetable with
the appropriate soundfont to play your midi file.
The author of awesfx wrote to tell me where to find source code in .tar.gz
format after the .bz2 package failed to download completely (premature EOF
- some code put in by bzip2) and I need to compile it in order to have
sfxload.
[Kapil]
I think playmidi has a switch to indicate the AWE card output, but
you need to set up the oss emulation mode. However, you can directly
use the ALSA sequencer via "pmidi".
I will let you know when I succeed in getting sound from a midi file.
Thanks for the help.
I am not using ALSA, but OSS. I compiled playmidi myself to use AWE32 as
default. The later version (5) of awesfx works with ALSA, version 4 does
not but will work for OSS. awesfx is said to work with awe32/64 and Live
cards, but the sound banks are different for the Live cards (of which I
just got another at a yard sale in case a future computer balks at ISA).
On another computer I had to disable USB support in BIOS to get the ISA
sound card to work. On this one I could disable USB, get the ISA card
working, and reenable USB, and sound still worked in linux but Windows
would not boot at that point (I had also added two more ISA cards but
those were not the issue). I don't really know what happened but the PCI
sound card fixed it. Now I need to upgrade Win98FE to SE in order to use
the Live card.
Someone said DOS assigns DMA to all PCI cards, maybe Windows did the same
and ran out?
I will report back when I get sfxload and midi working.
I have done some more experimenting on the ISA AWE cards.
I have a Vibra16 CT3930 which has jumpers (not pnp) and it is the only one
that initializes AWE32 in DOS and plays AWE in DOS. I have two CT3600s
one of which is AWE32 and has 512K free RAM and the other SB32 without the
free RAM, but all three cards have two SIMM slots (with 2x4MB) - one had a
jumper to enable this RAM, the others did not.
The CT3600s and two AWE642 (4520) play awe in a DOS box under Win98, which
means Windows is initializing the awe part. They play all but AWE in
plain DOS. I have not yet tried the Vibra16 in linux with playmidi.
None of these cards needed sound banks loaded in DOS or Windows (unless
Windows loads its own automatically) so I don't think that is my problem,
I think it is initialization.
I got Creative's CTCU.ZIP (at the pipeline driver site) and played with it
for a long time. I was able to get both CT3600's working in a computer
where they had conflicts by resetting the MPU IRQ and DMA with CTCU and
then running CTCM. The AWE32 retains these settings permanently, the SB32
has to have them reset every time. I also had to disable the gameport by
running ctcu and ctcm (and possibly the IDE secondary controller) to get
them to work - otherwise DIAGNOSE (SB16 - similar to CTCM but cannot
disable things) cannot find a free address on one card or a free IRQ on
the other card. They both came set to MPU 300 and IRQ 10. One is now
IRQ5. I can now use these cards in linux with the same settings that work
in DOS, except for AWE, which works in neither despite being set
(apparently) by CTCM (from ctpnp.cfg file, which I edited manually).
The two AWE64s refuse to permanently change their IRQ settings and
Diagnose tells me IRQ5 is taken - by the card, somehow, perhaps for the
second 32 voices? I have not managed to disable whatever takes up IRQ5
and they won't work at all on this computer but work on other computers.
(One is lacking FM synthesis, so who cares). But they ought to work in
linux where the PCI cards are more polite about grabbing IRQs, if I can
get AWE initialized, probably with isapnp.
With the remodeled CT3600 AWE32 in this computer, Win98FE boots correctly
and makes noises, and the DVD-ROM drive is no longer complaining about
IRQs and DMA. IRQ5 DMA 1 and 5.
I found sbldos.zip - DOS drivers for the PCI SB Live - not yet tried.
Win98FE has no drivers, SE does.
and a cs4232 appeared to do the same (but no sound came out).
[]
I still have to try the Vibra16 CT3930, which is not pnp, in linux, with
playmidi. (FIrst I had to run some experiments with video cards and
mplayer and DVDs, which led me to conclude that you don't need more than
4MB RAM, and that an ATI AGP card with 4MB and no heat sink, and to a
lesser extent a PCI Voodoo with 16MB RAM and heat sink, works much better
than one with 64MB RAM and a fan).
I was able to disable IDE controller and gameport with jumpers on this
card.
Ubuntu linux did not find our ISA ethernet or sound cards.
[]
It did find the PCI card that played silently in my linux, and adjusted
the volume so it works now (Trident).
None of these cards needed sound banks loaded in DOS or Windows (unless
Windows loads its own automatically) so I don't think that is my problem, I
think it is initialization.
[]
The non-pnp Vibra16 CT3930 sound card, which should not need pnp
initialization, plays AWE in DOS but not in linux. It has 1MB GM (general
midi) sound onboard so should not need sound loaded. It also plays
silently with playmidi in linux.
I made character device sequencer.
I suspect my problem is that I compiled playmidi wrong. Someone with a
plain SB16 card said that FM synthesis is broken. I remarked out a bunch
of lines and parts of lines referring to things I did not want, so I
would not have to install GTK and other libraries related to producing
a fancy graphical output rather than just sound, and probably also broke
the AWE part. There is a precompiled playmidi that worked for other
people but not for me which I can also try to get working.
At the same site as the awesfx utilities for loading sound banks
(awesfx-0.5.0d.tar.gz source - 94K)), I found an alternative midi player
in the package awemidi-0.4.3c.tgz (366K) which I will attempt to compile
(hopefully it can be configured NOT to need the GTK-based interface or
someone here can suggest how to modify Makefile properly to avoid it).
The player is 'drvmidi'.
The author suggests that if you have a pnp card you can initialize it in
DOS and then use loadlin. I have four pnp AWE cards. I downloaded some
more AWE32 drivers, which are available at a Finnish site - 130-AWE1.ZIP
through 4.ZIP, and also awe_REV4.zip (all from 1995) and some more faq and
info files, and will try to get these cards playing AWE in DOS. I also
found sbldos.zip - DOS sblive drivers, for my two PCI (non-pnp) cards with
a similar EMU80?? awe chip.
There is also an sb64basic.exe (the site was not working when I tried)
and, likewise inaccessible yesterday, 700031A.exe through E.exe for AWE32
(hopefully for DOS not just Windows), A being drivers and C utilities.
You cannot initialize these cards in Win98 and then reboot to DOS and keep
them initialized - I tried that too.
playmidi plays silently
From sindi keesan
Answered By: sindi keesan, copied to BasicLinux: baslinux@lists.ibiblio.org
But trimmed since we see the whole thread above :)
-- Heather
PLAYMIDI PLAYS MIDIS NOW!!!!! (So does drvmidi).
Short summary to date: Playmidi was playing silently on my AWE32 card.
I suspected that I had compiled playmidi wrong (FM synthesis was said to
be broken, and I compiled by remarking out anything in Makefile referring
to X, gtk, or ncurses). Someone at your list said I needed to load a
sound bank with sfxload. I thought it was not really needed for the card
that worked in DOS without me loading anything into it. I was also told
to initialize my pnp cards, so I tried first with the only jumpered AWE3(2
I have (CT3930 Vibra16), which does not need initialization and works in
DOS (the others do not yet, won't do AWE initialization in DOS). aweutil
/s initializes the card - won't play AWE without that.
I was indeed getting silence because I had not loaded any sound banks
(fonts) into my AWE card. See below for the details. Drvmidi is much
easier to compile without GTK or ncurses than is playmidi, and does not
require libc5 like playmidi did but both work now.
with playmidi in linux. I made character device sequencer.
[]
(This was not needed - perhaps only for external midi devices)
I found an alternative midi player in
the package awemidi ... The player is 'drvmidi'.
[]
I compiled awemidi-0.4.3c.tgz (which is only 70K - very simple
Makefile.sample could be edited to remark out all references to both
ncurses and gtk/X11, unlike playmidi which insisted on them) and got
drvmidi which also played silently. I downloaded the source code for
awesfx-0.50.d and it would not configure because it wanted alsa, so I got
version 0.4.4 and it compiled perfectly (just type make, no configure
needed for this version). All the files compiled statically by default
except for needing glibc/libcm/ldlinux.
I found in the DOS package (s64basic.exe - just run it to install SB16
with AWE support then delete the windows directories) along with
aweutil.com one sound bank synthgm.sbk.
sfxload synthgm.sbk (other sound banks are at the driver site)
Now both drvmidi and the playmidi that I compiled play midis now with
awe, that sound just as good as the DOS players. SB16 must have been
loading this sound bank without me knowing it.
The precompiled playmidi apparently predated the AWE addition and works by
default on an external midi device. Only other options were gravis
ultrasound or FM synthesis. THis is why it played silently, and the
playmidi I compiled DOES work for AWE but needed the sound bank installed.
Thanks for this suggestion. I have learned a lot along the way. I still
need to try the other AWE cards (that won't work yet in DOS but do in
Windows) with linux.
I am successfully playing midi files in linux without any need for isapnp.
The solution was to get newer versions of ctcu.exe and ctcm.exe
(Creative's ISA configuration utility and manager) as found in
ctcmbbs.exe, which comes inside s64basic.exe. I had replaced those files
with files from ctcu.zip because of a corrupt download of s64basic.exe in
which those two files would not work. The good ones are from 1997.
To get this going in DOS, run s64basic.exe and follow instructions to make
a ctcm directory. (If you don't actually plan to use the card in DOS, it
can be separate, otherwise within SB16). Run ctc -- modify any settings,
test them - A220 I5 D1 H5 E620 worked for me and I was able to pick an
audio configuration with no midi port at all (300 or 330) and to disable
game port and controller) followed by ctcm. Copy ctpnp.cfg to the sb16
directory. Run diagnose and let it edit autoexec.bat and config.sys. I
remarked out what it puts into config.sys and anything about ctcm in
autoexec.bat and instead aded to autoexec.bat the line c:\ctcm\ctcm (with
or without /s). Leave the set sound and set blaster and other settings,
and diagnose, mixerset, and aweutil /s, which initializes awe32 (tho it
did not used to do so).
After exiting ctcu, run ctcm to load the new settings. If you have a
separate ctcm directory, copy ctpnp.cfg to the sb16 directory. Diagnose
sets up DOS to run from SB16 directory.
I was then able, on both AWE32 and SB32 cards, to play AWE files with the
'diagnose' utility and with three DOS midi players in AWE mode. (One of
them refused to play with io set to 240 but worked at 220). PMB MSP and
CDP.
I booted into linux with loadlin and insmodded the usual sb modules, and
then awe_wave, which loaded for the first time.
I used sfxload to load synthgm.sbk (it needed the path to find it) and
then drvmidi to play it. I used the vol control on the radio which I am
using as a speaker (aux input) to adjust volume.
In both DOS and linux, the radio makes regular clicking noises when I am
not playing a midi file but otherwise it all works perfectly.
I have posted awesfx and drvmidi packages for glibc2.2.5 (statically
compiled otherwise) at http://keesan.freeshell.org. s64basic.exe can be
found at several places on the web (search by name) including pipeline in
Australia.
This was not yet tested on AWE64 and I have not yet attacked the PCI card
SB Live, for which Creative also provides DOS drivers (sbldos.zip).
If you do not set up SB in DOS first, you will have to deal with isapnp.
I tried for a couple of hours to edit the isapnp.conf file produced by
pnpdump (isapnp /etc/isapnp.conf) but kept getting IRQ or DMA or IO
conflicts since I don't know what I am doing.
I think only the AWE part of SB (and maybe FM synthesis) needs to be
initialized (in DOS with ctcm or in linux with isapnp) because I was able
to play wav and mp3 files in linux after I booted DOS without system
files. I was also able to play wav and mp3 files on an Opti 931 (mad16)
and an ESS 1868 (sb) in linux without initializing the sound card first.
The ES1868 has an IDE controller which would have to be disabled somehow
if you want to use the onboard secondary controller in linux as hdc/hdd
instead of hdf/hdg. So do many other older isapnp cards. The non-pnp ISA
cards have jumpers to disable things with, much easier.
keesan (at) sdf.lonestar.org
SDF Public Access UNIX System - http://sdf.lonestar.org
Normally I trim people's sigs entirely, sometimes I leave a scrap in.
In this case - people ask us now and then about whether there are any
decent public-access or pay-by-call providers, maybe this will answer
some of them too.
And with that, sindi, welcome to the Answer Gang. Tell the bartender
your drink of choice - you've earned it!
-- Heather
x86_64 distributions
From Rick Moen
Answered By: (folks on the ILUG mailing list) Kevin Lyda, Rick Moen, Braun Brelin,
David Golden
[Mailing list participant Braun Brelin inadvertently triggered a noisy
100-post distribution-advocacy flamewar by asking what's the "best
64-bit Linux distro". These meta-comments followed:]
-- Heather
[Kevin]
So, after that long thread, I'm wondering what Braun thinks of his
choice.
No, no, not the distro choice; his choice of ILUG as a place to email
a question about Linux distros...
[Rick]
Are you suggesting that there's at least one on-line Linux forum on the
globe where soliciting "Suggestions for best 64-bit [sic] Linux distro"
would not constitute the hapless posting of flamebait?
Someone with a bit more common sense might have posted "Here's the
Linux distros I'm aware of, so far, with active x86_64 ports: Ubuntu,
Kubuntu, Gentoo, Debian, Fedora, Mandriva, SuSE, RHEL, CentOS, Tao
Linux, Scientific Linux, CRUX, Lineox, cAos-2, RockLinux, Knoppix64,
Slamd64 (Slackware), StartCom, SourceMage, Rocks, rpath Linux, Kanotix,
White Box. Any others worth mentioning?"
Someone might even have chased down a long batch of URLs for
publication.
-- Heather
[Kevin]
What can I say, I'm a Pisces; I dream a lot. Dreaming or not, I do
think a nearly 100-message thread on it borders on farce.
Any others worth mentioning?"
[]
It would be nice to get some ideas of what people are using. If you
want to use a distro that will have local support, I think the question
has some merit.
[Rick]
We Tauruses, by comparison, are too stubborn to believe in astrology.
Just as a reminder, though, the question posed was "What's the best...?"
You're a Taurus? I've always wondered, how do you acquire flatware?
Just as a reminder, though: The question posed was "What's the
best...?"
[]
Yes, OK, that wasn't the wisest choice of words. But still, no matter
how distro questions are asked, these threads always seem to careen out
of control.
[Braun]
I was rather amused, actually. I was thinking of asking another question
on the list about GNOME vs. KDE, to see if I could spark another long
flamewar.
[David Golden]
Bah. Neither approach the professionalism of classic X desktops.
Okay, in most respects they do, but one thing really annoys
me about both KDE and GNOME is the way they still don't handle
applications running from different machines and home directories
properly at all. They make the assumption, time and again, that
all the apps on the screen are running on the one display with prefs
in the one home directory. If I wanted that, I could get a fucking
Macincrap.
Older X applications used the xrdb, so that UI configuration
was a property of the display. All your resource db-supporting apps
could look just how you wanted them, no matter which host they
were running from.
Then 10000 Windows-developer-weenies apparently jumped ship to Linux,
bringing an incoherent mess of files stuck in the home directory and
slagged off xrdb while completely missing the point of it, so we have
things like the unadulterated horror of GConf. A few did eventually
grasp what xrdb brought, and I acknowledge that xrdb would need heavy
tweaking bordering on rearchitecting for "modern" preference datatypes,
but they really threw the baby out with the bathwater by forsaking its
core ideal of dynamic, server-brokered preferences.
P.S. See "man Xaw" for an interesting hack: you can now specify the
vector-drawing commands used to draw a widget in the "displayList" xrdb
resource of an Athena widget...
[Braun Brelin]
The reason I asked the question was not because I wanted to know the
best "distro", I wanted to know what was the best "64-bit" distro.
Given that 64-bit chips are sti
![[BIO]](misc/cover/shadowman-80.png)
![[BIO]](misc/cover/centos_icon_60.png){kind=icon}
centOS plugs itself as the
Community Enterprise OS, and intends to be and remain
binary compatible with... ummm, a Certain North American Software Company.
yummy!![[BIO]](../gx/2002/note.png)
Heather got started in computing before she quite got started learning
English. By 8 she was a happy programmer, by 15 the system administrator
for the home... Dad had finally broken down and gotten one of those personal
computers, only to find it needed regular care and feeding like any other
pet. Except it wasn't a Pet: it was one of those brands we find most
everywhere today...
Stumping The Answer Gang, Take II
[Heather] Apparently accepted, since his blog mentions having a mentor to work with.
Congrats Pradeep!
Open Solaris
Greetings from Heather Stern
The floppy boots the Knoppix miniroot, scans a
bunch of devices in /dev, and says
