Linux-Hardware Digest #682, Volume #14 Tue, 24 Apr 01 23:13:08 EDT
Contents:
Re: AMD and CPU idle state ("Peter T. Breuer")
Re: oldest terminal you have used (Larry Ebbitt)
Re: Networking -- switches vs hubs ?? (Joe Pfeiffer)
Re: kppp problems... (Bill Unruh)
Re: AMD and CPU idle state (lino)
Re: kppp problems... (Dances With Crows)
Re: Flat panel (TFT/LCD) with Linux anyone? (Greg Hughes)
Re: Could Linux be used in this factory environment ? (Paul Repacholi)
Re: Networking -- switches vs hubs ?? (Martha H Adams)
Re: Pentium I 133 +32 MB enough ? (David E. Lamy)
Re: Could Linux be used in this factory environment ? (Charles Lyttle)
Re: Switchboxes for keyboard, mice, video? (Keith R. Williams)
How to ifconfig ADM8511 usb adapter under Linux 7.1 ? ("stephen")
Re: Switchboxes for keyboard, mice, video? (Keith R. Williams)
Re: Switchboxes for keyboard, mice, video? (Robert Redelmeier)
Re: sb1000 cablemodem & 2.4.2 ("Dave Thompson")
----------------------------------------------------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: AMD and CPU idle state
Date: Wed, 25 Apr 2001 01:03:24 +0200
lino <[EMAIL PROTECTED]> wrote:
>> Well, frankly, without calibration, how can you tell? Put a
>> thermomenter inside and see what it says.
> As I wrote before, I've compared the temps from lm-sensors, BIOS and Win
> until, and they agree within the differences due to the reboot times.
> I've even measured them with an expernal thermometer, and all works
> fine.
This is very improbable. How did you choose the correct lm-sensors
algorithm?
(I speak as a person who had to reverse engineer the internal wiring
of my mobo to deduce the correct algorithm, and you'll see me quoted
in the lm-sensors documentation on the qdi brilliant mobo for that).
Peter
------------------------------
Date: Tue, 24 Apr 2001 19:20:15 -0400
From: Larry Ebbitt <[EMAIL PROTECTED]>
Subject: Re: oldest terminal you have used
Paul Ryan wrote:
>
> > I was gonna throw in an ASR33, but I'm whupped at the gate <g>.
>
> I would love to get hold of one of them. Anybody in UK got one they don't
> want?
>
Check with any Ham friends you have. They used to be popular with them.
--
Larry Ebbitt - Linux + OS/2 - Atlanta
------------------------------
From: Joe Pfeiffer <[EMAIL PROTECTED]>
Subject: Re: Networking -- switches vs hubs ??
Date: 24 Apr 2001 17:06:53 -0600
[EMAIL PROTECTED] (Martha H Adams) writes:
> I've just picked up some hardware for making a network, and reading in the
> accompanying manuals, I see "switches" and "hubs" in a baseT network are
> either same thing for practical purposes -- or different; and I can't guess
> which out of the manuals. The "switch" I have here is a Linksys EZSX55W
> 5-port switch Version 2.0. Apparently it's appropriate and I can use it to
> network a few desktops and laptops that I have in my home; but I'd like to
> be sure.
A couple of other people have given correct answers about the
difference; the only thing I'd add is that what it adds up to is that
they work interchangeably in a network, but switches have higher
performance.
--
Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605
Department of Computer Science FAX -- (505) 646-1002
New Mexico State University http://www.cs.nmsu.edu/~pfeiffer
SWNMRSEF: http://www.nmsu.edu/~scifair
------------------------------
From: [EMAIL PROTECTED] (Bill Unruh)
Crossposted-To:
comp.os.linux.help,comp.os.linux.misc,comp.os.linux.questions,comp.os.linux.setup
Subject: Re: kppp problems...
Date: 24 Apr 2001 23:35:46 GMT
In <3ae5ec8c.0@news> "Karim R. Sobhi" <[EMAIL PROTECTED]> writes:
]Apr 24 21:52:46 Edgar pppd[927]: Peer is not authorized to use remote
]address 217.52.7.110
] i'm sure the username and password are correct so i don't know where the
]problem is. any ideas?
Yes. You have an ethernet card with a default route defined on it.
Place
route del default
at the end of the rc.local file
(in /etc/rc.d on redhat/mandrake)
If this is a remote machine logging in to your machine, you fogot to
put the fourth entry into your {pap,chap}-secrets file.
username * password *
------------------------------
From: lino <[EMAIL PROTECTED]>
Subject: Re: AMD and CPU idle state
Date: Tue, 24 Apr 2001 23:33:46 GMT
"Peter T. Breuer" wrote:
>
> lino <[EMAIL PROTECTED]> wrote:
>
> >> Well, frankly, without calibration, how can you tell? Put a
> >> thermomenter inside and see what it says.
>
> > As I wrote before, I've compared the temps from lm-sensors, BIOS and Win
> > until, and they agree within the differences due to the reboot times.
> > I've even measured them with an expernal thermometer, and all works
> > fine.
>
> This is very improbable. How did you choose the correct lm-sensors
> algorithm?
Yes, ok... but trust me, the temps are correct at least within 2 C or
even better. I've made a few hours of test, swapping system, kernel,
inserting an external thermometer, opening and closing the computer
case... the right modules were (as in the startup file rc.local):
---[...]---
## i2c and lm-sensors
modprobe i2c-viapro
modprobe i2c-isa
modprobe eeprom
modprobe via686a
/usr/local/bin/sensors -s
---[...]---
> (I speak as a person who had to reverse engineer the internal wiring
> of my mobo to deduce the correct algorithm, and you'll see me quoted
> in the lm-sensors documentation on the qdi brilliant mobo for that).
>
> Peter
In any case, the problem of the AMD idle state persists. I hope
somebody has a solution for this... In the Changelog files of the newest
kernels like 2.4.3 or 2.4.4-pre6 it doesn't seem to me to see anything
about this...
Thanks for the suggestions.
Lino
------------------------------
From: [EMAIL PROTECTED] (Dances With Crows)
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: kppp problems...
Reply-To: [EMAIL PROTECTED]
Date: 25 Apr 2001 00:14:58 GMT
[NGs trimmed, don't crosspost so much!]
On Tue, 24 Apr 2001 16:18:41 -0400, Stuart Williams staggered into the
Black Sun and said:
>"Karim R. Sobhi" wrote:
>> i'm trying to connect my Linux Mandrake 7.1 to the internet through
>> my modem. the modem is compatible and when i query it it works fine.
>> when i run kppp, on the other hand, i get this on the log file:
>>
>> Apr 24 21:52:38 Edgar pppd[927]: pppd 2.3.11 started by root, uid 0
>> Apr 24 21:52:38 Edgar pppd[927]: Using interface ppp0
>> Apr 24 21:52:38 Edgar pppd[927]: Connect: ppp0 <--> /dev/ttyS1
>> Apr 24 21:52:46 Edgar pppd[927]: Peer is not authorized to use remote
>> address 217.52.7.110
Make sure that "noauth" is in /etc/ppp/options . If the line that says
"noauth" is commented out, then pppd will require the remote machine to
authenticate itself, which is not what you want if you're dialing up to
your ISP. (Hey, you know which phone# you called, right?) This is
strange, "noauth" should be on by default for the mythical Normal User
that seems to be Mandrake's target market. Again, find the line that
says "noauth" and remove the # from the front of that line.
>ok i dont know if this will work but you can try it. use modemtool to
>make your modem /dev/modem then connect using /dev/modem I did
>this last time I used a modem(been a long time). hit alt-F2 in kde
>and type modemtool
Irrelevant. /dev/modem should probably be a symbolic link to /dev/ttyS1
in this case, but if /dev/modem was pointing to the wrong place, the OP
would get different error messages.
--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin / Workin' in a code mine, hittin' Ctrl-Alt
http://www.brainbench.com / Workin' in a code mine, whoops!
=============================/ I hit a seg fault....
------------------------------
From: Greg Hughes <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Flat panel (TFT/LCD) with Linux anyone?
Date: Wed, 25 Apr 2001 00:53:11 GMT
Hello
I've got an SGI 1600sw flatpanel hooked up to a G400 using the MultiLink
adapter.
I have it working under a GNU/Linux install with kernel 2.4.3 and XFree86
4.0.3 w/ Matrox HAL/Driver 1.2.0 (from their web site). The set up is
running in analog mode.
This set up works pretty well. The display quality is impeccable -- just
as good as w/ the DVI adapter once you tweak the phase adjustments for the
analog signal. My only beef w/ this set up is that XFree86 4.0.3's DRI
modules for the G400 crash a lot, locking up the whole machine when I'm
playing Quake 3 (after about 20-60 minutes). And that has nothing to do
with the flatpanel (except that I need XFree86 4.x -- see below).
The one thing about the set up is that you have to use XFree86 4.x to get
the 1600x1024 resolution to work properly (3.3.x doesn't do the right
things with the modeline). I also need the 1.2.0 HAL for the Linux
drivers to get a proper signal output (otherwise there is a black bar at
the left of the screen).
However, the MultiLink box correctly detects and supports the VESA
standard resolutions including 1280x1024 and below, as I suspect most
flatpanels would do, so tweaking the modelines and whatnot to get weird
timings will probably not apply if you get a flatpanel supporting these
standard modes.
I would recommend against getting the DVI add-on for the G400. I have it,
and it sucks. The analog output of the G400 by itself is so good that the
DVI adapter only provides marginal quality gains. However, the DVI
adapter also brings a whole new set of problems.
The G400 routinely loses sync with the MultiLink adapter through the DVI
output, causing the flatpanel to turn off and then on again after a couple
of seconds. This happens every 5 minutes or so at the lower resolutions,
so it's frequent enough to be extremely annoying, to the point that the
DVI adapter is unusable at 640x480 or at the console. The blanking occurs
in all resolutions, all colour depths and happens in all operating systems
and at the BIOS POST screen. There are also quirky problems getting X to
work properly with the DVI adapter (I had to 'roll my own' version of the
Matrox-supplied X drivers to get it to work properly), and when you exit X
it doesn't properly reset the console video mode (no signal sent to the
flatpanel) so I have to reboot the machine.
In short, the card works very well in analog mode with a flatpanel, but do
not purchase the DVI add-on for the G400 until Matrox gets their act
together and fixes all the problems with it (which I doubt they will).
- Greg
On Tue, 24 Apr 2001, it was written:
>
> Hi, Does anyone have any success with using a flat panel monitor with
> Linux? If so, I would appreciate if you could post the details.
>
> Iam currently running Linux Mandrake 7.2, but would be moving to 8.0
> soon. I have a Matrox Millennium G400 (Dual head) AGP card. Iam
> thinking of buying a flat panel monitor, but want to make sure I can
> use it under Linux.
>
> Thanks, nram
>
> ==================================
> Poster's IP address: 38.161.118.161
> Posted via http://nodevice.com
> Linux Programmer's Site
>
------------------------------
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.misc
Subject: Re: Could Linux be used in this factory environment ?
From: Paul Repacholi <[EMAIL PROTECTED]>
Date: 25 Apr 2001 08:28:05 +0800
[EMAIL PROTECTED] (SammyTheSnake) writes:
> In article <[EMAIL PROTECTED]>, Jonadab the Unsightly
> One wrote:
> >Above all, you don't want the computer stopping for no reason, so
> >run it on Unix or VMS, and you won't go down unless there's a
> >hardware problem or some idiot runs a forkbomb.
VMS is not affected by this sort of thing, the 'bomber' runs out of
quota and gets an error.
> not even then if you use a recent version of linux[1]. If each user
> has their own login ID then other users won't even notice :)
Problem is, it is normal that peole DON'T log in in a factory. The
terminals are auto-logged in by station, and run what ever is needed.
VMS has several ways of doing this, up to AMDS.
The oply use for logins is in the lunch room. A few extra 'unofficail
unsuported' terminals in the lunch room and some games does wonders
for the fault call rate, and operator ability to operate a terminal.
--
Paul Repacholi 1 Crescent Rd.,
+61 (08) 9257-1001 Kalamunda.
West Australia 6076
Raw, Cooked or Well-done, it's all half baked.
Spam-To: [EMAIL PROTECTED],[EMAIL PROTECTED],[EMAIL PROTECTED],[EMAIL PROTECTED],
[EMAIL PROTECTED],[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Martha H Adams)
Subject: Re: Networking -- switches vs hubs ??
Date: Wed, 25 Apr 2001 00:58:03 GMT
Hi Joseph and others. Thanks for responses. Now all I have to do is get
my network set up and get it to *work.*
My myself-only application for a network in my home is, to have a machine
in a dedicated spot for my fiction writing; another machine there which
runs without keyboard and monitor for some remote-computing work, probably
fractals; two or three machines in my livingroom for general hacking; and
a dedicated machine running a firewall. ...It wasn't all that long ago,
some Big Name pronounced there *might* be work in this world for three or
four computers. Well! Cheers -- Martha Adams
------------------------------
From: [EMAIL PROTECTED] (David E. Lamy)
Subject: Re: Pentium I 133 +32 MB enough ?
Date: 25 Apr 2001 01:20:54 GMT
In article <[EMAIL PROTECTED]>, Jaap L.A. wrote:
>Hi,
>
>I'm planning to install Linux Suse 6.0 on a P133 with 32 mb. I'd like to use
>the KDE desktop. Will it work or will I get just as frustrated with it as I
>am with win9x etc.
>Has anybody experience with this ?
>
If you use a lighter weight window manager such as Sawfish, Blackbox, or FVWM
(others probably exist in addition to these), then X will run just fine. Just
avoid Netscape and Mozilla as well.
I use an old IBM Thinkpad with the same hardware specifications as Jaap's box
and FVWM provides a positive X experience for me.
--
__o As of 04/24/01, free software is still the state of an
_`\<,_ open mind. Develop it, share it and use it!
(*)/ (*) David Emile Lamy [EMAIL PROTECTED]
pgp key available at http://pgp5.ai.mit.edu/~bal
------------------------------
From: Charles Lyttle <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.misc
Subject: Re: Could Linux be used in this factory environment ?
Date: Wed, 25 Apr 2001 02:10:26 GMT
Erik Funkenbusch wrote:
>
> "Charles Lyttle" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Erik Funkenbusch wrote:
> > >
> > > "Charles Lyttle" <[EMAIL PROTECTED]> wrote in message
> > > news:[EMAIL PROTECTED]...
> > > > Mind you, I can't vouch for the accuracy of that report. But is was
> > > > offered as proof that the crash wasn't the fault of the OS. "It was a
> > > > misbehaving application that caused the OS to crash." The Navy had to
> > > > clear the OS or justify to congress why it insisted on MS when most
> > > > contractors were saying it couldn't (or shouldn't) be done. The
> > > > contractor had to clear the OS because he promised that it could be
> > > > done. But he was late, and the Navy had to either cancel a test at a
> > > > loss of millions, or go to test with a beta version. So the vendor
> says
> > > > "we were only a little late, if they had just waited a few more days".
> > >
> > > This is all completely untrue.
> > >
> > > Read all the info collected by Jerry Pournelle on the issue
> > > http://www.jerrypournelle.com/reports/jerryp/Yorktown.html
> > >
> > > The contractor in question also stated specifically that the navy had
> gone
> > > against their recomendation of installing newer software that didn't
> have
> > > the problem PRIOR to the event.
> > > http://www.sciam.com/1998/1198issue/1198techbus2.html
> > >
> > > "... the fault was with certain applications that were developed by CAE
> > > Electronics in Leesburg, Va. As Harvey McKelvey, former director of navy
> > > programs for CAE, admits, "If you want to put a stick in anybody's eye,
> it
> > > should be in ours." But McKelvey adds that the crash would not have
> happened
> > > if the navy had been using a production version of the CAE software,
> which
> > > he asserts has safeguards to prevent the type of failure that occurred.
> "
> > >
> > > You should also read the original article which is the source of all
> this:
> > > http://www.usni.org/Proceedings/digiorgio.htm
> > >
> > > Notice that in an article of great length, only 2 paragraphs are devoted
> to
> > > NT, and none of the say the OS crashed.
> > >
> > > Further, the same author that wrote the gcn article (which quotes from
> the
> > > usni article) also clarifies his statements in a followup article:
> > > http://www.gcn.com/archives/gcn/1998/november9/6.htm
>
> > That last one is even worse than my story. A divide by zero in the
> > controller for a fuel valve caused the entire LAN to go down crashing 27
> > remotes?
>
> The navy uses different terminology than the rest of the industry. To them,
> the "LAN" is their networked application, not the OS or hardware it's
> running upon. When the database contained invalid data, the applications
> running on the remotes all crashed as well, thus causing a Local Area
> Network failure.
>
> > Industry (mostly) fixed that problem 30 years ago. For what its
> > worth, I had an NT machine I was working with bring down an entire LAN
> > of over 1000 machines. It was called the "ping of death". Some
> > applications could cause the NT software to start issuing network pings
> > at high speed. These faults often also caused a BSOD, but not always.
>
> Ping of death was a specific type of exploit that needed special, invalidly
> formed packets to work. There was no way to cause normal applications to
> generate POD packets.
>
> Unix also suffered from this bug.
This was caused by my desktop issuing several thousand pings per second.
No one else could get on the LAN. Every now and then someone could slip
in between pings, but not often. The system was transmitting properly
formated packets, just lots of them. And ignoring everyone else.
--
Russ Lyttle
"World Domination through Penguin Power"
The Universal Automotive Testset Project at
<http://home.earthlink.net/~lyttlec>
------------------------------
From: [EMAIL PROTECTED] (Keith R. Williams)
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.sys.ibm.pc.hardware.chips
Subject: Re: Switchboxes for keyboard, mice, video?
Date: Wed, 25 Apr 2001 02:46:27 GMT
On Tue, 24 Apr 2001 05:49:18, [EMAIL PROTECTED] (Eric P.
McCoy) wrote:
> [EMAIL PROTECTED] (Keith R. Williams) writes:
>
> > > > Windows key. ...don't need no stinkin' Windows key! Never.
>
> > > Actually, I kind of like it. The one that does Ctrl+Esc (not the
> > > context menu key), I mean. I'm a die-hard command line junkie, so in
> > > Windows it's Ctrl+Esc, up, up, enter, command name - I use it a lot
> > > for calc and notepad.
>
> > When you'r not using WinBlows? I don't even want it when I
> > am! I don't like WinBlows specific hardware in any form.
> > The WB key is just another example.
>
> I'm sure X and your favorite wm can be configured to do something
> useful with it.
Last time I checked it blew WPS away, not that it was
unintended. I don't need the keys and rather like the
101-key layout. ...with a few gripes.
> > > > Home and End could be done better.
>
> > > That's why God gave us C-a and C-e. (By the way, do not try to use
> > > those in Windows - bad things, like erasing the entire message, will
> > > happen. I do it all the time, and it drives me crazy.)
>
> > Cntl-Home and Cntl-End? ...I'm sure Billy has figured out a
> > convoluted way of doing this too.
>
> Home, End, et al. are the fault of IBM, not of Microsoft. And to be
> honest, they're much more intuitive than whatever Unix correspondents
> there are.
Sure. I'd prefer HOME and END closer to the cursor keys,
but...
----
Keith
------------------------------
From: "stephen" <[EMAIL PROTECTED]>
Subject: How to ifconfig ADM8511 usb adapter under Linux 7.1 ?
Date: Wed, 25 Apr 2001 11:15:03 +0800
I have some question about the
ADM8511 usb adapter under Linux Redhat 7.1 with kernel 2.4.2:
After loading the "pegasus.o" , then "ifconfig" the adapter as below:
==> ifconfig eth0 192.168.1.12 netmask 255.255.0.0 or dhcpcd eth0
there are some message shows:
"pegasus.c : eth0 :link NOT established (ox7849)- check the cable
I can't connect to my Lan network, but executing "ifconfig" can
show eth0 and lo .
I also check the cable which is connected to Lan network
(10/100M switch), and the cable is pretty good.
PS: the leds on the usb adapter :
power : lit Link/Act : not lit 100M : not lit
Does anyone has similiar problem as mine? And
could anyone give me some resolution to the problem?
Thanks .
------------------------------
From: [EMAIL PROTECTED] (Keith R. Williams)
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.sys.ibm.pc.hardware.chips
Subject: Re: Switchboxes for keyboard, mice, video?
Date: Wed, 25 Apr 2001 03:00:54 GMT
On Tue, 24 Apr 2001 14:34:05, [EMAIL PROTECTED] (Jonadab
the Unsightly One) wrote:
> [EMAIL PROTECTED] (Keith R. Williams) wrote:
>
> > Well, since QWERTY was *designed* to limit typing speed...
>
> Sort of. IIUC it was designed to prevent typewriters
> from sticking, which involved preventing nearby keys
> (especially adjascent ones) from being hit in rapid
> succession, although a certain amount of limiting
> the typing speed was involved also. But that was
> the means, not the end.
Actually, no. QWERTY was specifically desigened to slow
down touch-typists. Note that the most commonly used keys
are on the least powerful fingers on the left hand. Try
typing anything without a left hand.
> But the Dvorak layout, supposedly designed for faster
> typing speed, is barely (if at all, which is debated)
> better. It makes most of the same mistakes as qwerty.
This was the theory, but in practice alphabetizing keys
isn't any better than QWERTY. There has been much study on
keyboards, but we're still stuck with QWERTY. So be it. I
won't live long enought to see it go away.
> > The "designed for Windows" sticker fell of my laptop
> > and stuck to the wastebasket. ...don't know how that
> > happened.
>
> Can't imagine. FWIW, this is a multiboot system,
> not a Windoze-only system.
My work laptop is the only single-boot system I have. My
work desktop and all systems here at home boot many OS's.
THe Laptop is for my principle work, which is necessarily
WinBlows. Other systems run whatever is needed.
> > > Actually, what I'd really like to try is to increase
> > > the number of buckies and reduce the number of regular
> > > keys by doubling up he letters and numbers.
>
> > Sounds too complicated.
>
> It is somewhat...
>
> > I liked the "chord" keyboards.
>
> And that's *less* complicated? It's the same idea,
> taken further to the extreme.
Sure, but with fewer keys it's easier to train oneself.
Court stenographers do this all the time. It can't be any
harder than learnign VE. ;-)
> > > The only thing I'll probably leave untouched are
> > > the cursor movement keys on the keypad. I like
> > > them just like they are.
> >
> > Home and End could be done better.
>
> I'm used to them, and I like them. They're within
> easy reach while you're using the arrows and page
> up and page down, which is the critical thing.
I'd rather they were down by the cursor keys and away from
the INSERT/DELETE keys. Actually I liked the PF keys on
3270s, but...
> Other keys I could do without are capslock and numlock.
> Capslock I *only* hit by mistake, and numlock I hit my
> mistake more often than intentionally. Actually, the
> numlock key is a good idea, it's just in the wrong place.
> It should be up in middle of the row above the function
> keys or someplace.
DON'T GET ME STARTED! ;-) Caps-Lock is an abortion for any
programmer/engineer.
> The guy who invented capslock should be shot. With
> sawed-off shotguns, loaded with rock salt. Tell the
> firing squad to aim for his feet.
Caps-lock worked a hundred years ago. No longer. I'm with
you on this one. I'm always re-typing lines because my
finger slipped onto the CL key. If I need all caps it's no
harder holding down a shift key than an C or A key. C-L is
different than Shift, obviously, bit no so much.
----
Keith
------------------------------
Date: Tue, 24 Apr 2001 10:02:04 -0500
From: Robert Redelmeier <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.ibm.pc.hardware.chips
Subject: Re: Switchboxes for keyboard, mice, video?
chrisv wrote:
>
> Robert Redelmeier <[EMAIL PROTECTED]> wrote:
>
> >> Maybe _you_ can't see it, but I sure can.
> >
> >And so can I! What I see is a contrast of two colors that
> >are not-quite-the-same. On 16bit color, a smooth transition
> >between two colors can produce a jaggy, digitized pixel line
> >(isochrom?) when one color flops over to another. This
> >can be highly disconcerting.
>
> When good dithering is used? And at high resolutions (which makes the
> dithering more effective)? If you say so, but 16-bit is essentially
> undistinguishable from 32-bit, to me under these conditions, and
> there's nothing wrong with my color perception that I'm aware of.
I don't know whether good dithering is used. The app is PowerDVD
playing DVDs [no hw accel]. I noticed that some very low constrast
and dark scenes would produce these distracting isochromes at 16bpp,
but they disappeared entirely at 24bpp.
It could be an artifact of PowerDVD or MPEG2 compression,
but then why does it disappear at 24bpp?
> If it were not for a few professional applications where color is
> vital, the world wouldn't miss "true color" at all, IMO. Of course,
> with modern hardware, it's like "why not" use true-color.
Well, 16 bit only gives 5 bits per color. At saturation, that
gives 3% between values which is almost certainly acceptable.
The problem is at zero [trace], each value increase produces
a much larger percentage change in output. Infinite in the
step from zero to one. 100% from 1 to 2. 50% from 2 to 3.
As I understand it, the eye has analog receptors (rods & cones
for color) on a pseudo-digital grid. The eye sees contrasts,
not absolute values. So it might have been better to have
colors encoded logarithmically, not linearly. The same
might apply for sound.
-- Robert
------------------------------
Reply-To: "Dave Thompson" <thompsw*nospam*@pobox.com>
From: "Dave Thompson" <thompsw*nospam*@adelphia.net>
Subject: Re: sb1000 cablemodem & 2.4.2
Date: Wed, 25 Apr 2001 03:02:29 GMT
I'm not sure why, but overriding the irq right in the source code got the
driver to work -- don't know whether there is a bug in my setup or in the
driver.
Dave
"Dave Thompson" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> I'm looking for help with a pnp device -- the general instruments
surfboard
> 1000 -- which is working ok under kernel 2.2.13 but I cannot get to work
> under 2.2.4. In fact, I'm back under 2.2.13 just to send this email
because
> this device is my connection to the internet.
>
> The sb1000 is a one-way cablemodem -- the upload is carried on through a
> serial modem and the download comes in through the cablemodem. I've
emailed
> the person who used to support the sb1000 but he does not have one of
these
> any more. He now has DSL which is not available at my house,
unfortunately.
>
> Here is the situation -- in 2.2.13 I setup isapnp.conf using the pnpdump
> routine, setting up irq 10 and I/o addresses 0x0100 and 0x0110. It works
> fine (has worked fine as my gateway/firewall for over a year now). In the
> kernel config PNP is turned on. The sb1000 is loaded as a module and
> produces three diagnostic messages into the /var/log/messages file
including
> one message that displays the i/o ports and the irq.
>
> With 2.2.4, I have pnp on and isapnp on in the kernel config (is this
right)
> ? I'm still using the isapnp.conf, except that I had to change the
> read_data_port to 0203 from 020b (not sure why, but pnpdump produces this
> under 2.2.14. In the prior version of the sb1000.c driver, you had to go
> into the driver to override the read_data_port to supply this value but
the
> new version for 2.4 does not have a place for this and seems to have code
to
> try and detect this itself. (this driver, by the way, is included with
the
> kernel as a config option if you are unfamiliar with it). When I run
isapnp
> /etc/isapnp.conf it seems to detect the card and does not complain about
> any conflicts. When I insmod sb1000 -- the log shows the first message
ok,
> the second message displays the correct i/o addresses but shows irq 0 --
by
> the way, I tried changing the i/o addresses to others in the allowable
> range, and this second message does NOT change so it is not finding the
i/o
> addresses from the card. The third message from the module never displays
> (this is simply a message that shows firmware level etc.). The module
stays
> running but does not seem to finish configuration.
>
> Also, I've tried NOT having the isapnp.conf, with effectively the same
> result -- the same messages, the same irq 0 ...
>
> So ... I've tried all sorts of permutations and combinations -- changing
the
> addresses or irq's within isapnp.conf is then reflected correctly when I
run
> isapnp -- but the driver cannot find the device, so it seems.
>
> Oh -- and I have the bios setup as 'not a pnp o/s' -- and -- it is NOT a
> dual-boot windows machine or anything like that.
>
> I would appreciate any helpful hints, options to try, windows to jump out
of
> ..
>
> This is not drop-dead critical -- I've been trying to upgrade this machine
> to 2.4.2 to take advantage of the additional facilities under iptables vs
> ipchains. 2.2.13 has been running for a long time without a problem ...
>
> Thanks.
>
>
------------------------------
** 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 by posting to comp.os.linux.hardware.
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
******************************
