Linux-Development-Sys Digest #631, Volume #6 Fri, 16 Apr 99 21:14:13 EDT
Contents:
Re: Clueless Users Are Bad for Debian - was Re: After Week 1 With Linux -- licking
wounds. (Christopher B. Browne)
Re: Templates (Andrew Eiler)
Re: Yet Another Audio Chip (Dave Platt)
Hp Kayak (Anis Hanna)
Re: Help with Linux Schedulers Pease ("Sascha Bohnenkamp")
Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0 (Alexander Viro)
Linux system ID, is there such a thing? (Neil Radisch)
bug in rpc.rstatd-3.01 ? rpc XDR return pack is 132 bytes long instead of 128.
("Alain Coetmeur")
Re: SMP Linux, Any Catches? ("Clint Byrum")
Re: Help with Linux Schedulers Pease (Christopher Mahmood)
Re: SMP Linux, Any Catches? (Andreas Dilger)
savecore? ("Jeff Chiu")
Re: SMP Linux, Any Catches? (M Sweger)
RE: Download file like NetVampire or Getright ??? (Dr H. T. Leung)
Re: Linux system ID, is there such a thing? (Ken)
Re: Linux system ID, is there such a thing? ("Clint Byrum")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Christopher B. Browne)
Subject: Re: Clueless Users Are Bad for Debian - was Re: After Week 1 With Linux --
licking wounds.
Reply-To: [EMAIL PROTECTED]
Date: Sun, 04 Apr 1999 22:59:24 GMT
On Sun, 04 Apr 1999 21:13:38 GMT, Jim Roberts
<[EMAIL PROTECTED]> posted:
>Got to agree with this. Those that are not willing to do the research
>sould stay with a MS product. A product that has little choice or
>selection availible within it.
Tom Christiansen says it aptly, if in a somewhat insulting fashion:
"Huh? Windows was designed to keep the idiots away from Unix so we
could hack in peace. Let's not break that." -- Tom Christiansen
It is remarkably "costly," whether in terms of dollars, lost hair, or
wasted time, when you try to support people that are ignorant and that
don't really *care* that they're ignorant.
>Dejenews and other resources would answer 99% of the questions new
>users have if they would just use their heads.
This is about the same as the claim that man pages would answer
everyones' questions. It *is* fair to say that the coverage is a bit
wider with "web resources," in that the art of creating good man pages
seems to be on the wane.
But there is still the problem that people have to ask the right
questions in order to get useful answers. This is true whether
they're:
a) Figuring out how to read a man page with the view of finding
answers,
b) Formulating a web search, or
c) Asking a question on Usenet.
An intelligent question can result in good results by any of these
approaches; a sufficiently lame question will find failure in all
these methods.
>You should see some of the comments posted in the SUN/Solaris/SCO
>groups since the MFG's have started to release free single user
>license for their X86 Unix products.
Doubtless.
>One more rant, how many times do we have to hear about WINMODEMS.
>
>I'm sure I'll take some flames here but.....
You should be burned at the stake, right...
The essay was quite nicely put.
Last night I figured out how to get Debian Linux onto my Alpha/UDB
box; the documentation out there was sufficiently fragmentary that if
I didn't already have a fair idea of what I was doing, I would not
have gotten the task done. As it was, I got stalled for a couple
hours trying to get the initial boot to happen. (That *should* be
better documented...)
Now that I've had success with it, I probably never need to worry
about it ever again, as Debian is almost auto-upgrading.
The process will "put off" those that are not reasonably competent.
Which has the good and ill that those that use Debian are pretty
likely to be "more competent than the national average."
Which is *actually* an argument for someone to make better
documentation and/or to create install processes to make Debian more
usable to the "less sophisticated," and *not* an argument for much of
anything else.
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
[EMAIL PROTECTED] - "What have you contributed to free software today?..."
------------------------------
From: Andrew Eiler <[EMAIL PROTECTED]>
Crossposted-To:
alt.linux,alt.os.linux,comp.lang.c++,comp.os.linux.development.apps,comp.unix.programmer,comp.unix.questions,gnu.g++.help,gnu.gcc.help
Subject: Re: Templates
Date: Fri, 16 Apr 1999 09:35:10 -0400
Or include the cc file at the bottom header file. The only issue you run into
would be a circular refrence, so you need to use some #defines to avoid that.
eg...
--- template.h ---
#idndef TEMPLATE.H
#define TEMPLATE.H
//class def
#include "template.cc"
#endif
---eof---
---template.cc---
#idndef TEMPLATE.CC
#define TEMPLATE.CC
//#include <...>
//class def
#endif
--eof---
-andrew
Pascal Proulx wrote:
> Martijn,
> read the info files provided with gcc. You are right: template code must
> reside in a header file.
>
> >All functions in template classes must be inline.
> I do not agree!
>
> Pascal
>
> Martijn Lievaart wrote:
>
> > Alexander Dymerets wrote in message <[EMAIL PROTECTED]>...
> > >Hi!
> > >> I'm trying to design template classes in C++ (under Linux and Solaris)
> > using
> > >> the gcc compiler, but I'm running into trouble. It seems that the
> > template
> > >> functions must be written
> > >> within the header (.h) file. Whenever they are written in the .cpp file
> > I
> > >> get errors about them not being there.
> > >
> > >All functions in template classes must be inline.
> >
> > Apart from some bugs in some compilers, this is simply not true. It is the
> > most portable though.
> >
> > Martijn
> > --
> > My reply-to address is intentionally set to /dev/null
> > reply to mlievaart at orion in nl
------------------------------
From: [EMAIL PROTECTED] (Dave Platt)
Subject: Re: Yet Another Audio Chip
Date: Fri, 16 Apr 1999 18:38:32 GMT
In article <[EMAIL PROTECTED]>,
Nix <$}xinix{[email protected]> wrote:
>> productized
>
>*shudder*
92nd Commandment: Thou Shalt Not Verbify.
--
Dave Platt [EMAIL PROTECTED]
Visit the Jade Warrior home page: http://www.radagast.org/jade-warrior/
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
------------------------------
From: Anis Hanna <[EMAIL PROTECTED]>
Subject: Hp Kayak
Date: Fri, 16 Apr 1999 10:44:01 -0300
Has anyone installed Redhat 5.2 on HP Kayak XA machines?
I did, works fine, except my xfree86 version 3.3.3 hangs the machine
within a few minutes of starting Xwindows. my configurator detects the
video card correctly (cirrus clgd5465). I tried everything in the
troubleshooting section of the manual, disabling blibt, and
acceleration, nothing works. Anyone had this problem?
Anis Hanna
------------------------------
From: "Sascha Bohnenkamp" <[EMAIL PROTECTED]>
Subject: Re: Help with Linux Schedulers Pease
Date: Fri, 16 Apr 1999 16:59:12 +0200
imho there exist only two seperate schedulers with linux
the normal one, priority (ageing)
and the rtlinux one, which supports realtime-processes in a seperate
scheduler ...
well inside the source of the scheduler there are some line which would make
the scheduler
round-robin etc.
------------------------------
From: [EMAIL PROTECTED] (Alexander Viro)
Crossposted-To: comp.os.linux.misc
Subject: Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0
Date: 16 Apr 1999 10:21:08 -0400
In article <W9AR2.1416$[EMAIL PROTECTED]>,
Christopher Browne <[EMAIL PROTECTED]> wrote:
>On 8 Apr 1999 17:07:45 -0400, Alexander Viro <[EMAIL PROTECTED]> wrote:
>>In article <[EMAIL PROTECTED]>,
>>Johan Kullstam <[EMAIL PROTECTED]> wrote:
>>>not if you're from a commonwealth country - which includes new zealand
>>>(where is the old zealand btw?). in *english* (as opposed to american
>> ^^^^^^^ - Zeeland.
>>
>> Gaak... Across the North Sea (looking from England, that is).
>>Netherlands. Heck, they *really* don't teach history and geography
>>in schools, or what?
>
>Not even in commonwealth countries... :-(
>
>I was not aware of that; the surprising factor is that it was named
>after a "foreign" nation, or am I showing *complete* ignorance in being
>unaware of a transition of New Zealand from Dutch control to English
>control?
Not exactly. It was rather Dutch "attention" & Maori control ->
British "attention" & Maori control -> sequence of really bloody wars ->
British control. IIRC New Zealand was discovered by Abel Tasman in 164?.
He gave the name. Don't forget, there was a Dutch control over Indonesia,
so...
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
------------------------------
From: Neil Radisch <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Linux system ID, is there such a thing?
Date: Fri, 16 Apr 1999 15:45:58 -0400
I'm looking for a way of identifying
a linux system. The number does not have to be 100% unique,
but it should be close. Under Irix and Solaris I can get
a motherboard serial number from the OS. Under Linux with
ethernet I use the ethernet ID.
What can I use if the Linux system has no network card?
Neil
------------------------------
From: "Alain Coetmeur" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.admin,comp.os.linux.networking
Subject: bug in rpc.rstatd-3.01 ? rpc XDR return pack is 132 bytes long instead of 128.
Date: Fri, 16 Apr 1999 15:37:07 +0200
for an application, I'm using the rpc.rstad service based on sun rpc.
I've programmed a C++ RPC system respecting sun rpc protocol,
and I contact the "rstat" server on linux and solaris.
on a RedHat 5.2 I'm using the latest version
rstatd-3.01
on some suse 6.0 nodes I'm using the rpc.rstatd
included in the standard nkitb packaged with suse 6.0,
but it behave the same as the 3.01 rstatd.
the problem is that the linux rpc.rstatd daemon returns
an incorrect XDR packet on UDP when responding.
it is longer by 1 long (4 bytes) that the theorical length.
the added data have the value "1L" which may be a flag.
clearly this is irrespective to the /usr/include/rpcsvc/rstatd.x interface definition.
this does not seems to annoy classic RPC clients like rup
or rsysinfo, because they mau be sloppy with UDP packet length,
but my RPC layer detect when there are remaining data and choke.
does anybody know if the rpc.rstatd
does this for "bug compatibility" ? if it is an errror ?
who should I report this bug to, if it is a bug ?
===========================================================================
receiving from a Linux rpc.rstatd server :
Socket#3.recv(8800,
00000000 00000001 00000000 00000000 00000000 00000000 000023dc 000001c4
000028f5 0205acd1 0001e014 00000000 00000000 00000000 00016eb4 000224ea
00000000 00000000 0214bad2 00000000 000ba295 00000000 00014ec0 0025e4fd
00000000 00000000 00000000 371203e9 00000000 371731f1 00012cdb 00000000
00000001
^^^^^^^^^^ unexpected data !!!!
)=132
err=CantDecodeArgs
receiving from a solaris rstad server:
Socket#3.recv(8800,
00000000 00000001 00000000 00000000 00000000 00000000 0212881b 002dd896
06bd35a7 4fafdebc 00000000 00529e1a 00000000 00000000 000acdb6 0000cee4
00000000 00000000 07c597e0 02cb1cf1 00000000 00000001 002da841 168b9731
00000000 00000002 00000003 36343081 00000000 3717343c 0002c071 0265d9d8
)=128
err=Ok
here is the XDR structure declared
const RSTAT_CPUSTATES = 4;
const RSTAT_DK_NDRIVE = 4;
struct statstime { /* RSTATVERS_TIME */
int cp_time[RSTAT_CPUSTATES];
int dk_xfer[RSTAT_DK_NDRIVE];
unsigned int v_pgpgin; /* these are cumulative sum */
unsigned int v_pgpgout;
unsigned int v_pswpin;
unsigned int v_pswpout;
unsigned int v_intr;
int if_ipackets;
int if_ierrors;
int if_oerrors;
int if_collisions;
unsigned int v_swtch;
long avenrun[3];
rstat_timeval boottime;
rstat_timeval curtime;
int if_opackets;
};
here is a manual interpretation (I used to read XDR fluently)
reply header (packet type, credentials) : 00000000 00000001 00000000 00000000 00000000
00000000
cp_time[0-3]: 000023dc 000001c4 000028f5 0205acd1
dk_xfer[0-3]: 0001e014 00000000 00000000 00000000
pgpg in/out: 00016eb4 000224ea
pswp in/out 00000000 00000000
intr: 0214bad2
ipackets,ierrors, oerrors, collisions: 00000000 000ba295 00000000 00014ec0
swtch: 0025e4fd
avenrun [0-2]:00000000 00000000 00000000
boot time: 371203e9 00000000
cur time: 371731f1 00012cdb
opacket: 00000000
BUG!!! what is this : 00000001
------------------------------
From: "Clint Byrum" <[EMAIL PROTECTED]>
Subject: Re: SMP Linux, Any Catches?
Date: Fri, 16 Apr 1999 13:07:21 -0700
Hold on, you guys are confusing me. My understanding of beowulf was that it
allowed clustered processing using Linux. I also thought that it would
essentially treat the "clustered" systems as additional processors, sending
jobs to them through a dedicated "cluster network". What would be the
benefit of running Beowulf on a standalone SMP system? I just figured that
with Beowulf installed, the kernel would spread its processes out among all
the beowulf machines.
Please, enlighten me, I'd like to use clustered processing as a
"scalability" option for our customers.
Andreas Dilger wrote in message <7f83ba$[EMAIL PROTECTED]>...
>I dislike the term "Beowulf system", since it is very misleading. Nothing
>runs on a "beowulf" system, but rather uses parallel processing libraries
>like PVM or MPI. PVM I know for a fact can run multiple jobs on a single
>system (hopefully SMP, but not required), and it can also run on multiple
>networked systems (the normal method). This allows you to compile a
>single executable for an architecture like Linux-x86, and run on SMP or
>networked single-CPU systems, and they all behave the same way.
------------------------------
From: Christopher Mahmood <[EMAIL PROTECTED]>
Subject: Re: Help with Linux Schedulers Pease
Date: 16 Apr 1999 09:52:11 -0700
there's one at sunsite, it supposedly uses GA.
-ckm
------------------------------
From: [EMAIL PROTECTED] (Andreas Dilger)
Subject: Re: SMP Linux, Any Catches?
Date: 16 Apr 1999 19:34:02 GMT
In article <7f7rmo$sp5$[EMAIL PROTECTED]>,
M Sweger <[EMAIL PROTECTED]> wrote:
>Instead of running Beowulf across multiple platofrms, have Beowulf run on
>one platform that has multiple processors, whereby each processor and its
>databus is treated as a virtual network within the platform. In essence,
>pretend that the multiprocessor machine is really a clustered set of
>machines.
I dislike the term "Beowulf system", since it is very misleading. Nothing
runs on a "beowulf" system, but rather uses parallel processing libraries
like PVM or MPI. PVM I know for a fact can run multiple jobs on a single
system (hopefully SMP, but not required), and it can also run on multiple
networked systems (the normal method). This allows you to compile a
single executable for an architecture like Linux-x86, and run on SMP or
networked single-CPU systems, and they all behave the same way.
Cheers, Andreas
--
Andreas Dilger University of Calgary \"If a man ate a pound of pasta and
Micronet Research Group \ a pound of antipasto, would they
Dept of Electrical & Computer Engineering \ cancel out, leaving him still
http://www-mddsp.enel.ucalgary.ca/People/adilger/ hungry?" -- Dogbert
------------------------------
From: "Jeff Chiu" <[EMAIL PROTECTED]>
Subject: savecore?
Date: Fri, 16 Apr 1999 15:00:20 -0700
Is there something like Unix's savecore for Linux?
If so, what to use to analyze the system dump? Something like crash?
Gdb?
Thanks for the info
.
------------------------------
From: [EMAIL PROTECTED] (M Sweger)
Subject: Re: SMP Linux, Any Catches?
Date: 16 Apr 1999 17:23:36 GMT
Reply-To: [EMAIL PROTECTED]
Clint Byrum ([EMAIL PROTECTED]) wrote:
: Where can i get the 2.2.x kernel? I haven't updated the kernel outright
: since I upgraded my 2.0.16 kernel to 2.0.22. As far as 14 processors go, I
: don't think I'd want all the eggs in one basket like that. I'd rather use
: beowulf and do clustered processing... More fault tolerant.
Hi,
Is it possible or easy to modify the Beowulf s/w to do this?
Instead of running Beowulf across multiple platofrms, have Beowulf run on
one platform that has multiple processors, whereby each processor and its
databus is treated as a virtual network within the platform. In essence,
pretend that the multiprocessor machine is really a clustered set of
machines.
This would be the micro view of the clustered configuration vs. the macro view
of the clustered configuration.
--
Mike,
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Dr H. T. Leung)
Crossposted-To:
alt.linux,alt.os.linux,comp.os.linux.x,comp.os.linux.development.apps,comp.os.linux.development,comp.os.linux.hardware
Subject: RE: Download file like NetVampire or Getright ???
Date: 16 Apr 1999 19:13:39 GMT
GNU wget. It is *much* better than anything in the windows world.
In article <[EMAIL PROTECTED]>, Vincent <[EMAIL PROTECTED]> writes:
|> I'm looking for a software which download file in multiple time like the
|> NetVampire or Getright under Windows ?
|> If this software exist, where can I found it ?
|>
|> Thanks
|>
|> [EMAIL PROTECTED]
--
--------------------------------------------------
"What you don't care cannot hurt you." Chap. 7a, AMS-NS
------------------------------
Date: Fri, 16 Apr 1999 17:05:25 -0700
From: Ken <[EMAIL PROTECTED]>
Subject: Re: Linux system ID, is there such a thing?
Crossposted-To: comp.os.linux.development.apps
I suspect you're addressing a problem at too low a level. Why do you
need a unique ID? Knowing that, there may be a better solution than a
hardware-based serial number.
It also then becomes possible to see what problems arise with your
application when using a hardware-based number, such as spoofing by
user-installed drivers. This, I think, is the fundamental problem with
Intel's PSN in their new P3. It's relatively easy to spoof. Another CPU
can pretend to be the user's CPU, effectively stealing his credentials
for services that use that number for authentication.
Neil Radisch wrote:
>
> I'm looking for a way of identifying
> a linux system. The number does not have to be 100% unique,
> but it should be close. Under Irix and Solaris I can get
> a motherboard serial number from the OS. Under Linux with
> ethernet I use the ethernet ID.
>
> What can I use if the Linux system has no network card?
>
> Neil
--
Ken
mailto:[EMAIL PROTECTED]
http://www.well.com/user/shiva/
------------------------------
From: "Clint Byrum" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Linux system ID, is there such a thing?
Date: Fri, 16 Apr 1999 13:10:34 -0700
I went through the same problem. On HP-UX one can get the system serial #
with the uname command. Linux(due to the use of Intel Hardware), has no such
ID. My advice would be to assign the numbers yourself, and stick them in
some configuration file(we make them have a certain user with the user's
"Comment" field in the passwd file as the Serial #).
Neil Radisch wrote in message <[EMAIL PROTECTED]>...
>I'm looking for a way of identifying
>a linux system. The number does not have to be 100% unique,
>but it should be close. Under Irix and Solaris I can get
>a motherboard serial number from the OS. Under Linux with
>ethernet I use the ethernet ID.
>
>What can I use if the Linux system has no network card?
------------------------------
** 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
******************************
