Linux-Development-Sys Digest #200, Volume #6 Sat, 2 Jan 99 03:16:22 EST
Contents:
Re: Kernel 2.1.131 + ppp + APM (Hans-Detlev Fink)
Re: Interesting (and surprising) way to crash X (=?ISO-8859-1?Q?Bj=F6rn?= Elwhagen)
Re: System temperature Linux vs. MS Windows (Gordon Scott)
Linux Journaling Filesystem (Michael Pronath)
Re: Access Control Lists within Linux (Christopher Browne)
Re: Much easier way to crash X (=?ISO-8859-1?Q?Bj=F6rn?= Elwhagen)
Re: BogoMips (Mathieu ARNOLD)
Re: Virtual PC (Richard RUDEK)
Re: lp0 on fire in 2.1.131 (Dr Dale Mellor)
Re: du and df give different results (Laurent Martelli)
Re: CDROM Warning: (don't try this) (Jeff Johnson)
Re: engineering practices in Linux/OSS (Philip Brown)
Re: RAD Tools (Glen Turner)
Re: Interesting (and surprising) way to crash X (Bill Davidsen)
XFree86 mode's for laptop LCD displays? (Richard Tilmann)
What about "Linux.. the home game"?? (a consumer version) (Jeff Johnson)
Re: Programming CDROM (Gyorgy Krajcsovits)
Re: Virtual PC (Tony Hoyle)
Re: Virtual PC (Mikko Rauhala)
Re: HELP w/ Yamaha OPL3-Sa Sound (Daniel N. Sands)
Re: Interesting (and surprising) way to crash X (Stefan Monnier)
Re: Interesting (and surprising) way to crash X (Mike McDonald)
----------------------------------------------------------------------------
From: Hans-Detlev Fink <[EMAIL PROTECTED]>
Subject: Re: Kernel 2.1.131 + ppp + APM
Date: Sat, 02 Jan 1999 07:38:21 GMT
"Henning Plöger" wrote:
>
> Hi,
>
> 1.I have compiled and installed the Kernel succesfully,
> but when I try to connect to my ISP with pppd
> it reports inn /var/log/messages "network down(100)".
> With the 2.0.36 Kernel there are no proplems.
> I am using pppd 2.3.3 on Redhat 5.1.
You will need ppp-2.3.5 for kernels > 2.1.99.
>
> 2.I have compiled the kernel with the APM-Option
> "Power-Off on Shutdown", when I shutdown
> the System it stops with "System halted".
> The APM-Functions are enabled in BIOS and
> with Windows there are no problems.
> With kernel 2.0.36 it ends up in printing endless
> lines of Numbers and Characters on my Screen.
> I am using a P2-450 with an BX-Mainboard.
Richard Jones answered to that already.
-Hans-
>
> Any Solutions?
------------------------------
From: =?ISO-8859-1?Q?Bj=F6rn?= Elwhagen <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x
Subject: Re: Interesting (and surprising) way to crash X
Date: Sat, 02 Jan 1999 07:38:30 GMT
In comp.os.linux.x Bill Davidsen <[EMAIL PROTECTED]> wrote:
> In terms of being able to access /dev/mem I guess, but in kernel you
> don't need to deliberately open a file and do i/o, a broken pointer can
> do the job when you're running in an absolute address mode. And if
> something hangs in user mode, any user, you at least have a reasonable
> chance of getting some clock ticks elsewhere.
I was merely saying that root has every permission there is to have on a
unix-system and that root had every permission that the kernel has. I am
too familiar with the hassle of a single pointer can have when directed
wrong in the address-space. I don't disagree at all for what it's worth.
>> As i said earlier. Go to the homepage of GGI and check what it's all
>> about. I haven't read much about it so i'm not he guy to tell you what
>> it does exactly.
> Why do GGI people think that if you don't agree with them you must not
> understand? I don't agree with the microkernel people, either, but I'm
> closer to them than the folks who think that everything belongs in the
> kernel. I firmly believe that things which can be done outside the
> kernel should be.
I don't consider myself GGI-people. As i said i have not read enough to
even begin to tell anyone what it's really about. That's why i
redirected him to the GGI homepage.
To further clear things. I would hate if the kernel got to bloated with
alot of unnessecery stuff. But i do think that, if it will work
properly, lowlevel-drivers for graphics in the kernel would be a good
thing.
I'm not in any way flaming anyone just for thinking different. I never
have and i never will. But i WILL tell my opinion about things as I see
it. When people stop beeing critical things are getting dangerous. I
actually think it's nice if someone corrects me if i'm wrong. But that
won't make me lay down and not check for myself first.
Best regards
// Marwin
--
Björn Elwhagen
Student at Wexio University
Sweden
------------------------------
From: [EMAIL PROTECTED] (Gordon Scott)
Subject: Re: System temperature Linux vs. MS Windows
Reply-To: Gordon Scott <[EMAIL PROTECTED]>
Date: Sat, 02 Jan 1999 07:39:03 GMT
Another interesting feature of an OS that halts when it has nothing to
do, is that it's RFI quiter than one that spins -- and the difference
can be very noticable. An old 386 I used to have would normally be fine
under Linux, but when running Windoze would completely blitz FM radio
reception within a range of several metres.
G.
--
Gordon Scott Opinions expressed are my own.
[EMAIL PROTECTED] (official) [EMAIL PROTECTED] (backdoor)
[EMAIL PROTECTED] (home) http://www.apis.demon.co.uk
Linux ............... Because I like to _get_ there today.
------------------------------
From: Michael Pronath <[EMAIL PROTECTED]>
Subject: Linux Journaling Filesystem
Date: Sat, 02 Jan 1999 07:38:33 GMT
Hi,
Windows NT has a journaling filesystem, as well as the new
OS/2 Warp Server. Is there anything like that for Linux yet
or a project to create some?
Is there a webpage with an introduction into the concepts of
journaling filesystems?
Michael
--
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Subject: Re: Access Control Lists within Linux
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:40:20 GMT
On 11 Dec 1998 08:21:55 -0600, Peter Samuelson
<[EMAIL PROTECTED]> wrote:
>[Anthony Ramm <[EMAIL PROTECTED]>]
>> In the source file linux/ext2_fs.h a struct ext2_inode defines
>> I_file_acl and i_dir_acl which are supposed to be addresses of acces
>> control lists. This is, however, not implemented in Linux 2.0. Does
>> anyone know if there is a version with it implemented
>
>Rumor has it that Stephen Tweedie and Remy Card have already done
>something with ACL's in ext2 but IIRC it was still alpha code. Don't
>expect to see it mainstream for awhile. (My guess: mid- to late-2.3.)
That sounds rather similar to the rumors I've heard.
Theodore T'so did a talk on this a year and a half ago at the Linux
Expo; indications *then* were that there was enough "hooks" already in
the ext2fs kernel code to make the kernel side easy. The problems then
are to design:
a) Library calls that "grok" ACLs,
b) Defaults that may be used with programs that are not "ACL-aware,"
c) As you mention later, it is necessary to have utilities that *are*
ACL-aware, including such luminaries as tar, cpio, cp, ln, mv, ...
d) It's not real useful to have ACLs unless they are well-managed,
which, in the presence of systems being "managed" using rpm/dpkg, means
that rpm and/or dpkg need to become ACL-ized, and packages themselves
need to be ACL-ized.
The early layers are pretty easy to handle, but every successive layer
of functionality requires more effort, and introduces substantial
additional complexity.
I'd say this is why we don't yet have ACLs in common use.
>> and also where I might find some documentation on the implementation
>> of access control lists within Linux.
>
>Can't help you there. But ACL's don't sound fundamentally hard to do
>-- I like Solaris's implementation, which involves I think two system
>calls and two userspace utilities. Problem: apps like `tar' need to
>know about ACLs since backup software often relies on it, yet you don't
>want to break the standard tarfile format.... (Or are there hooks in
>the tarfile format for ACLs?) One possibility is storing a pseudo-file
>in the archive that has ACLs in it (analogous to UMSDOS --LINUX-.---).
HP and AIX both have slighly different semantics that also have merit.
It could be argued that if Linux picks up some implementation, builds a
good set of base tools, as well as some interoperability tools (to make
it easier to migrate thinking from one of the other ACL
implementations), that this could result in widespread adoption of "our"
ACL implementation since it would be fairly widely used and understood.
--
Q: How many Newtons does it take to change a light bulb?
A: Faux! There to eat lemons, axe gravy soup!
[EMAIL PROTECTED] <http://www.ntlug.org/~cbbrowne/lsf.html>
------------------------------
From: =?ISO-8859-1?Q?Bj=F6rn?= Elwhagen <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x
Subject: Re: Much easier way to crash X
Date: Sat, 02 Jan 1999 07:38:36 GMT
In comp.os.linux.x Jeffrey Tsao <[EMAIL PROTECTED]> wrote:
> p.s. Please don't post a "solution" saying "just log off
> everyday". I need to stay logged on to check the stdout
> of some programs running in the background.
I'm sorry i can't solve your problem for you, but i thought i should
inform you of a very nice program called 'screen' that makes virtyal
sceens on tty's that you can put in the background and easily get back
in front again. It's pretty powerful although i don't think it supports
colors too good, not ANSI as i use in BitchX or programs like it at
least. I always have my rc5des-client running on a screen and ive never
experienced any problems with it. You can ofcourse have alot of
different screens at the same time.
Regards
// Marwin
--
Björn Elwhagen
Student at Wexio University
Sweden
Note: Remove 'nospam' from email-address
------------------------------
From: Mathieu ARNOLD <[EMAIL PROTECTED]>
Subject: Re: BogoMips
Date: Sat, 02 Jan 1999 07:38:45 GMT
Vincent Lai wrote:
>
> Does anybody know the meaning and measurement reference of the bogomips in
> linux? I have 2 PCs (200MHz MMX Pentium and 415MHz Pentium II) and they got
> about the same bogomips (around 400). Why?
you should go and take a look to the Bogomips mini howto.
--
Cordialement
Mathieu ARNOLD PGP key id : 0x2D92519F
IRC : Dalnet #mygale _mat ICQ uin:1827742
mailto:[EMAIL PROTECTED] http://www.multimania.com/arn
------------------------------
From: [EMAIL PROTECTED] (Richard RUDEK)
Subject: Re: Virtual PC
Date: Sat, 02 Jan 1999 07:41:34 GMT
Richard Jones <[EMAIL PROTECTED]> wrote:
>Philippe Rochat <[EMAIL PROTECTED]> wrote:
>: Does it exist on Linux something similar to Virtual PC (On Mac OS) ?
>
>: With such a sys, it would be possible to install Linux everywhere, and then
>: install win or NT on top of it if really needed by users ... it would also
>: provide a smooth way to migrate from monopolistic world to freedom !
>[...]
>
>It's a good idea, but not simple to implement because of the
>severe brokenness of the x86 architecture. There are broadly
>two approaches: (1) try to virtualize the CPU and I/O
>peripherals as far as possible; (2) emulate CPU instruction-
>by-instruction and also virtualize the I/O peripherals.
[snip]
>
>I am inclined to think that (1) is the most feasible
>approach. It may be possible to actually patch the NT kernel
>directly to remove any problematic code. It's highly
>unlikely that there are many places in NT's code which
>actually fug around with the %eflags register as I described
>above (or do any of the other odd non-virtualizable stuff).
>
I've never got into it, NuMega has a range of SoftICE products that
effectively do this - it's a software based InCircuit Emulator (ICE). It
allows you to break in, examine, modify, trace etc practically anything -
even OS's. I seems to work because many device driver developers and people
reverse engineering the internals of Win95 and NT swear by it, not at it
:).
It would be great to be able to have an OS that could change personalities
on the fly, each personality continuing to run in the background if
required (a support person's dream) - A robust, highly optimised Virtual PC
could do this.
__ __ _______________________________
//)) //)) | Richard RUDEK. MicroDek. | Hey, it's Friday night...
//\\ //\\ | Chatswood, Sydney. Australia. | C:\WORK> CD \
`-------------------------------'
------------------------------
From: Dr Dale Mellor <[EMAIL PROTECTED]>
Subject: Re: lp0 on fire in 2.1.131
Date: Sat, 02 Jan 1999 07:41:42 GMT
Talking about bad error messages:
Keyboard not found, press F1 to continue
Dale
------------------------------
From: Laurent Martelli <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: du and df give different results
Date: Sat, 02 Jan 1999 07:38:39 GMT
Veksler Michael <[EMAIL PROTECTED]> writes:
> In comp.os.linux.development.system
> Laurent Martelli <[EMAIL PROTECTED]> wrote:
>
> > My partition mounted on /usr is reported as full by `df', but if I do
> > a `du' on /usr I get a very much different result. I also check it
> > with e2fsck to see if it wasn't short of inodes, but it seems to be
> > ok.
>
> > I am using kernel 2.1.131
>
> > Any idea my my disk space may have gone ? Am I missing something ?
> > Have I found a bug ?
>
> [lot of info snipped]
>
> Consider the scenario:
> $ (dd if=/dev/zero bs=1024k count=128; sleep 1h) > log &
>
> Great, now your disk gets this 128MB log file.
> $ rm log
>
> You remove the file, right? Well, not exactly. The file can no longer be
> accessed from your file system (not counted by du), but the offending
> process can still write/read this file (which df is aware of). The file
> will get removed only when this process dies.
> $ killall sleep
>
> will probably kill this process, and free your disk space (not recommended
> since it may kill some useful sleep processes :-).
>
> I guess that this is your problem (or else you have some major disk
> corruption - not likely)
I did an e2fsck on the partition, and it turned out that a great many
inodes had a wring dtime, and that inode counting was wrong almost
everywhere. The number of free inodes was smaller than what it should
have been. Which explains the difference between du and df, since I
guess that df uses the sub-totals for performance reasons.
However, I doubt that your description corresponds to my problem. I
doubt that I had some process like that one. In fact, df seemes to
give constant results as long as nothing is written on the disk. I
reinstalled my debian 2.0, and still had the same effect. So I tried
to install and remove a package several times. Free space was
narrowing. This was with kernel 2.1.131. I'll try with a 2.0.xx and
see how it behaves.
--
Laurent Martelli
------------------------------
From: Jeff Johnson <[EMAIL PROTECTED]>
Subject: Re: CDROM Warning: (don't try this)
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:39:15 GMT
Walter Lundby wrote:
> So, I figured:
> "dd bs=32768 if=/dev/cdrom of=/dev/null"
>
> Worked great to prove that the cd wouldn't spin down...
>
> BUT SMASHED THE READ HEAD after it finished reading the CD!
> The drive was a total loss.
> [Awful firmware: No range testing]
Bravo!
Its not everyday that someone can cause physical damage to a device
by simply doing a data transfer test. I manage a peripheral
testing/compatibility lab and cannot remember the last time I heard of
something like this. I'd love to be able to hand a device back to a
sales person and say "you're device really sucked, so much so that it
literally blew itself up during testing..."
Just my $.02
-Jeff
------------------------------
From: [EMAIL PROTECTED] (Philip Brown)
Crossposted-To: comp.os.linux.misc
Subject: Re: engineering practices in Linux/OSS
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:39:34 GMT
On 12 Dec 1998 06:36:27 GMT, [EMAIL PROTECTED] wrote:
>When those I receive files from upgrade, I can't tell them to send me
>it in an older format!
why the heck not?
All you do is send it back to them, which will open up their program.
Then ask then to "save as", and send it back.
you can put in a jab about your department "not feeling the need to waste
money on pointless upgrades" if you like.
--
[trim the no-bots from my address to reply to me by email!]
--------------------------------------------------
"initiating.. 'getting the hell out of here' maneouver" - Lennier, babylon5
------------------------------
From: Glen Turner <[EMAIL PROTECTED]>
Subject: Re: RAD Tools
Date: Sat, 02 Jan 1999 07:37:36 GMT
Meng Howe wrote:
>
> Anyone knows of any RAD tools in Linux?
Depends on your application: a lot of Windows RAD
tools are DB-oriented, whereas most Linux RAD tools
are web-oriented (but still allow DB access). This
is no bad thing -- most RAD vendors are rapidly trying to
"web-enable" their products.
Furthermore, Linux RAD tools presume an existing programming
environment, and thus do not provide one. Xemacs and the
GNU programming utilities actually build a very effective
programming environment, but one which requires more
configuration out-of-the-box than Windows products. The
main advantage of this approach is when you use your
*second* langauge -- the Linux development environment
doesn't change.
There is also a good Borland-like programming environment,
but I haven't used it.
An off the top of my head list is:
Perl
Tcl
Python
Php
For a traditional RAD developement, Tcl or Python
would be the go. Perl has the advantage of having
a huge amount of reusable code already available.
The Apache web server offers significant support
for Perl and PHP.
All the languages except PHP allow `tricky' code
to be written in C and used by the language with
fewer linguistic hoops than is required in Windows.
Tcl's C integration is particularly attractive if
you use it to build a vocabulary of application-
specific words -- see the Scotty SNMP library for an
example.
Sun's Java is also worth the look -- especially as it
very neatly allows the construction of backend
application servers and frontend application clients.
It is a good choice if you need to simplify complex
DB transactions into application transactions.
Finally, all significant DB vendors have recently announced
support for Linux. Thus a lot of the DB-vendor-specific
languages will shortly be on Linux.
Cheers,
Glen.
--
Glen Turner Network Specialist
Tel: (08) 8303 3936 Information Technology Services
Fax: (08) 8303 4400 The University of Adelaide 5005
Email: [EMAIL PROTECTED] South Australia
------------------------------
From: Bill Davidsen <[EMAIL PROTECTED]>
Subject: Re: Interesting (and surprising) way to crash X
Date: Sat, 02 Jan 1999 07:37:31 GMT
Christopher Browne wrote:
>
> On Wed, 09 Dec 1998 14:27:06 GMT, [EMAIL PROTECTED]
> <[EMAIL PROTECTED]> wrote:
> >Why is this, anyway? Is it primarily a safety feature, or does it just have
> >to do with the architecture of the kernel? And what happens to the signal?
> >Does the kernel process it, realize it's for a process in kernel space, and
> >throw it away?
>
> It happens because kernel code can't reasonably be pre-empted. (No
> doubt as SMP gets more robust, this applies to smaller and smaller bits
> of kernel code.)
>
> Essentially we can "blame" this on the "architecture of the kernel."
To a large extent I agree, but the reasoning is that some parts of the
kernel CAN be preempted, and some i/o devices will do i/o with DMA (or
bus master) and therefore it is much easier to avoid trying to kill the
process and release than it is to abort the system calls and i/o. You
can spell "easier" as "safer," too ;-)
Some early version of UNIX had problems in this area, and many operating
systems had trouble when they first went SMP. If you lose track of
processors in the same process, it's like the song "don't let your right
hand know what your left hand do." Only there's a good chance that the
right hand will be making a rude gesture while the left hand waves
bye-bye.
This can be filesystem corruption, writing over data in memory for a
running program, etc. You don't want kill -9 to be *too* effective, I
believe.
--
bill davidsen <[EMAIL PROTECTED]>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
------------------------------
From: Richard Tilmann <[EMAIL PROTECTED]>
Subject: XFree86 mode's for laptop LCD displays?
Date: Sat, 02 Jan 1999 07:44:05 GMT
Anyone know where the docs are for writing the XFree86 mode lines for
laptop LCD displays? How does the H & V synch timings and clock
timings translate to the LCD?
------------------------------
From: Jeff Johnson <[EMAIL PROTECTED]>
Subject: What about "Linux.. the home game"?? (a consumer version)
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:39:14 GMT
Okay...
This is still an idea formulating in my head so I am sure that many
of you will be able shoot holes in this idea. (fire away)
I was thinking about the whole Linux/OSS versus Winbloze technology
and the evil empire in Redmond Washington. As it stands now Linux
threatens WindowsNT workstation and server applications. These are the
domain of the sys admin, computer technology professional and the real
saavy end user or hobbyist base. Surely no real estate agent or family
doctor is going to go to their favorite computer store and buy a new
system and a shrink wrapped Redhat box and run home to set it all up and
play on the internet or send emails to family and friends. That is the
domain of Winbloze 95/98.... or is it?
We already have a high-end, stable, performance operating
system/environment with Linux. Why not strip it down and come up with a
single-user, GUI only environment version of Linux strictly for the home
use, non-computer professional. I mean hell, behind the GUI desktop does
it matter whether its DOS32 or a Linux kernel running the show?
I said at the top you could shoot this full of holes and I am sure
there is enough here to start a steady barage of fire. I just think if a
verion of Linux could be created that boot/install started directly into
a GUI environment and a 'wizard-like' setup utility just like Winbloze
95/98 and was shipped with support for most of the current volume
shipping video cards. Include Netscape and some application like Star
Office or Applixware. Set it up in such a way where the real estate
agent or teacher or Dr Quinn medicine woman wouldn't have to know how to
recompile a kernel, linuxconfig or be super-user. Make it so they can
write a letter and print it, send an email, browse web pages, play Quake
and maybe remove the "ooh UNIX, Im scared" factor. Maybe add in some
sort of WABI like interface so people could run applications made for
the Winbloze environment.
Do that and Linux/OSS would challenge Winbloze on all three levels:
Server, Workstation and consumer desktop.
And no more Blue Screen of Death......
Like I said, its just a thought I had and am typing it out as I am
pondering it. I am just curious if it had ever been thought about or
considered...
Just my $.02, your mileage my vary. See dealer for details. Offer ends
12/31/99. Additional charges may apply in certain areas.
-Jeff
------------------------------
From: [EMAIL PROTECTED] (Gyorgy Krajcsovits)
Subject: Re: Programming CDROM
Date: Sat, 02 Jan 1999 07:42:09 GMT
[EMAIL PROTECTED] wrote:
: In article <751m8e$7mt$[EMAIL PROTECTED]>,
: "J.C Chuang" <[EMAIL PROTECTED]> wrote:
: > Hi,
: >
: > In Linux, when inserting a CD to CDROM, is there any mechanism
: > that system will notify me automatically that a CD is inserted or a CD is
: > changed ? I don't think polling is suitable for me ?
: Hmm, I don't know. Try looking at the kernel source but I remember somewhere
: that a lot of CDROM's don't support this notification so if you really want
: it to work, you might have to poll.
I was trying to do something like that, but never got it working realy.
What I did was to write a simple (about 45 lines) Tcl/Tk script
that has a button for (mount/umount and eject). It starts every
time I run X. Anyway, this will save you lots of "mount /cdrom",
although not automatic.
bye, krajo
: Martijn van Oosterhout
: Australia
--
George Krajcsovits | e-mail: [EMAIL PROTECTED]
None of the opinions in this post are those of
the Eotvos Lorand University of Science, Budapest,Hungary
-- Member of HuLUG (Hungarian Linux Users Group) --
------------------------------
From: [EMAIL PROTECTED] (Tony Hoyle)
Subject: Re: Virtual PC
Date: Sat, 02 Jan 1999 07:45:39 GMT
On 15 Dec 1998 11:30:46 GMT,
[EMAIL PROTECTED] (Andreas Mohr) wrote:
>
>> Hmmm ... this is not as strong as VirtualPC strategy (and especially doesn't let
>> you install NT which is possible on Virtual PC). Does it support games for
>> example ?
>Yes, it does:
>Unreal, Quake, Starcraft, ...
>(those are the ones mentioned most often)
Mostly apocyphal I think. Wine in its current incarnation can't even
run the NT version of File Manager. The only app I've ever managed to
run 'out of the box' is Notepad.
It'll be a huge leap when they get InstallShield working, as it'll let
mere mortals like me (who aren't into hand-installing Office 97) see
if some of these claims are partially true.
Tony
btw. Quake runs native on Linux anyway, why emulate it?
====================================================================================
Hello, my name is Bill. I like blue. Do you like blue? Let me show you some blue.
====================================================================================
[EMAIL PROTECTED] http://sale.netfusion.co.uk
====================================================================================
------------------------------
From: [EMAIL PROTECTED] (Mikko Rauhala)
Subject: Re: Virtual PC
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:42:21 GMT
On Mon, 14 Dec 1998 16:53:43 +0100, Philippe Rochat <[EMAIL PROTECTED]>
wrote:
>Phil Hunt wrote:
>> WINE ("WINE Is Not an Emulator") is a Windows emulator for Linux.
>
>Hmmm ... this is not as strong as VirtualPC strategy (and especially doesn't
>let you install NT which is possible on Virtual PC).
The Wine approach (implement the Windows API) is certainly harder and
slower, but also on the long run a much better solution - not having to
buy a "real" copy of Windows being one, but not the only reason for it
(think better integration with the host OS, efficiency...).
>Does it support games for example ?
In principle, yes. Wine has a DirectX implementation, for instance, and
reports of more or less successfully running even modern games have
appeared (Starcraft, Unreal to name a few). However, it must be remembered
that Wine is still considered alpha software.
Actually, it would seem to me that games (especially action games) are the
easiest programs to get running, since they generally use a lesser subset
of the Windows API than applications such as office suites. That's my gut
feeling, anyhow.
--
Mikko Rauhala - [EMAIL PROTECTED] - http://www.iki.fi/mjr/
------------------------------
From: [EMAIL PROTECTED] (Daniel N. Sands)
Subject: Re: HELP w/ Yamaha OPL3-Sa Sound
Date: Sat, 02 Jan 1999 07:46:46 GMT
Ralph wrote:
#-> I have a Yamaha OPL3-Sa sound card that is supposed to be Sound Blaster
#-> compatible. When i configure the card under sndconf I've tried regular,
#-> 16bit, and Pro settings w/ correct IRQ, DMA, and I/O settings. My specific
#-> card is supposed to be Sound Blaster Pro Compatible, which I have never
#-> gotten to work at all. NO Sound. I was wondering if anyone could suggest
#-> anything to help my problem or tell me where i could download drivers.
#-> Thanx for your time.
OPL3-SAx is only SB Pro compatible. If you want 16-bit, use the WSS/MSS
drivers. For complete support, use the SAx-specific drivers.
---
Coming Soon - DOOM ]I[: What the Hell?
------------------------------
From: Stefan Monnier
<[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x
Subject: Re: Interesting (and surprising) way to crash X
Date: Sat, 02 Jan 1999 07:46:56 GMT
>>>>> "wily" == wily <[EMAIL PROTECTED]> writes:
> Well, you're right in a sense, but only if you use linux on a network.
> However, the keyboard and mouse DO include the entire system when you run
> linux as a standalone.
Don't know how it is where you live, but around here text terminals cost about
0$ (plus 10 bucks for the rs232 cable).
Stefan
------------------------------
From: [EMAIL PROTECTED] (Mike McDonald)
Crossposted-To: comp.os.linux.x
Subject: Re: Interesting (and surprising) way to crash X
Reply-To: [EMAIL PROTECTED]
Date: Sat, 02 Jan 1999 07:47:00 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (G T Francisco III) writes:
> On Tue, 15 Dec 1998 15:00:13 +0000, "John L. Masson" >Hear, hear.
> PMJI, ocassionally X will crash on me and I have to do Cti-Alt-Bkspace
> to shut it down. At that point that particular tty is just completely
> useless because the return key doesn't work, i.e. I just get ^M's when
> I hit the return key. Even if I go to another Virtual Console and kill
> the process, when I return I can't login because the enter key still
> won't work (still get ^M's instead).
>
> While not terribly inconvenient since this is a home machine and I
> have all the other VC's, do you have any idea how I can "reset" that
> particular tty/VC?
>
> Thanks,
>
> G.
Ah, and oldy but goody. Try "stty sane ^J" into the messed up shell.
Mike McDonald
[EMAIL PROTECTED]
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.development.system) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Development-System Digest
******************************
