Linux-Development-Sys Digest #160, Volume #7 Mon, 6 Sep 99 13:14:13 EDT
Contents:
Re: [Linux] Calling C from Fortran: Function always returns zero? (Nix)
Re: select() and FD_SETSIZE (Nix)
Re: Linux standards compliance (Nix)
Re: gdb Reference (Nix)
Re: LINUX AND COREL (Nix)
Re: LispOS? (Nix)
Re: The conceptual sandbox? (Nix)
Init pb : respawning too fast ("Gregory Lepere")
encripted disk (Camil Coaja)
Re: [Linux] Calling C from Fortran: Function always returns zero? (Taketoshi Sano)
Richards Stevens died (Scott Lanning)
Re: The conceptual sandbox? (Jonathan Guthrie)
Re: Can I compile the kernel using a cc other than gcc? (David T. Blake)
Re: what's a "jiffy" in /proc? (Mladen Gogala)
Re: Flamage - Why? (Errin Watusikac)
Bottom Half(NET_BH) (Fernando Ortega)
want to move my russian keyboard to linux ("Bert Douglas")
Re: The conceptual sandbox? (James Andrews)
----------------------------------------------------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: [Linux] Calling C from Fortran: Function always returns zero?
Date: 06 Sep 1999 07:54:43 +0100
[EMAIL PROTECTED] writes:
> Given the compilers you're using, this is unlikely, but could C be using
> a different register than Fortran for returning values?
This is a property of the md file in gcc, not of the front end, and has
remained unchanged for linux platforms for donkey's years.
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: select() and FD_SETSIZE
Date: 06 Sep 1999 01:08:17 +0100
[EMAIL PROTECTED] (Alan Curry) writes:
> I'd rather edit
> /etc/termcap than go through the tic/untic/infocmp compilation cycle.
Users cannot edit /etc/termcap.
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: Linux standards compliance
Date: 06 Sep 1999 01:14:08 +0100
[EMAIL PROTECTED] (Christopher Browne) writes:
> It's not evident that much actual work has been put into rewriting
> Linux device drivers to conform with the UDI API
AFAIK, none at all has.
I doubt it will happen at all (unless the kernel forks) while the Powers
that Be in the Linux kernel world are so against it. (Oh dear, *what* a
pity. ;} )
> (This may be not unlike the situation with the adoption of a common
> ABI for IA-32 systems
Dead, shut down a few months ago. Linux/glibc2 is the common ABI,
according to the project's web page ;}
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: gdb Reference
Date: 06 Sep 1999 07:52:11 +0100
[EMAIL PROTECTED] (Peter Samuelson) writes:
> AGREED. I don't know why the FSF considers man pages obsolete. Now I
Once it was because they were poorly structured. Now, I don't know what
RMS thinks (not being RMS) but *I* for one think that writing stuff in
*roff yourself is obsolete because you can translate texinfo into it.
This is what texinfo is *for*, being translated into other formats.
> don't particularly care for roff format either (which is why God
> invented POD and other stuff which can be converted to man pages), but
Like texinfo. texinfo is a legible and widely-used-outside-perl form of
POD ;P
> the *idea* of man pages -- synopsis of usage, complete reference of
> functionality, few if any examples -- makes info files look very clunky
> for small programs. Sure, large programs probably need more of a
> manual, but `cut' doesn't.
Agreed. Probably invocation instructions should be automatically
translated into *roff at the same time as the complete info file is
makeinfoed.
Ah, a job for {awk|perl|elisp|guile} ;}
> As someone who already uses emacs/xemacs for most of my editing work, I
> find its info browser natural and convenient.
Except that XEmacs 21.0.? < x < 21.1.5 had a number of horrible horrible
bugs in the info browser. (Some are still there).
> For a vi user, I imagine
> it would not be.
It's damned good for a searchable hyperlinked format. IMHO it still
beats HTML (searchability, TeXability) and man pages (better
hyperlinking and structuring), *except for quick reference*, where man
has the edge.
It amazes me that RMS has apparently been unable to see this; after all,
in his own work (Emacs) the Info documentation was for major
reference/tutorial purposes and he had a separate system for quick
reference (*and* multiple search tools over that system).
The Emacs documentation system still has no better in any system I have
ever seen or used; perhaps the Lisp machine's system was better, but
that is regrettably dead :(
I'd almost go so far as to say that without that documentation system I
would never have started using Emacs. That system, and its seamless
integration with the customization/enhancement facilities in Emacs, is
definitely Emacs's major advantage over vi (even more than the
multi-buffer handling &c).
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Crossposted-To: comp.os.linux.misc
Subject: Re: LINUX AND COREL
Date: 06 Sep 1999 07:52:59 +0100
[EMAIL PROTECTED] (Peter Samuelson) writes:
> remember when real version numbers
> all but disappeared from 'doze software in favor of two-digit year
> numbers?)
I remember when things were going the other way (Clipper Summer '87,
followed by Clipper 5)
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: LispOS?
Date: 06 Sep 1999 00:51:25 +0100
[EMAIL PROTECTED] (Peter Samuelson) writes:
> [Nix <$}xinix{[email protected]>]
> > Shouldn't be too hard to add it though. I'll work up some patches for
> > xemacs, I think, just as a gimmick ;)
>
> Don't forget to disable "\C-x\C-c"....
ITYM `kill-emacs'.
btw, wonderful comment/code juxtaposition (in XEmacs-21.1.6 at least) in
`shut-down-emacs'):
: #if 0
: /* This is absolutely the most important thing to do, so make sure
: we do it now, before anything else. We might have crashed and
: be in a weird inconsistent state, and potentially anything could
: set off another protection fault and cause us to bail out
: immediately. */
`Yes, this is very important, so let's not do it'. ;}
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: Nix <$}xinix{[email protected]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: The conceptual sandbox?
Date: 06 Sep 1999 00:53:11 +0100
[EMAIL PROTECTED] (Christopher B. Browne) writes:
> You are thinking of the "ACM Classic" entitled "Reflections on
> Trusting Trust."
[snip]
> Note that the idea is apparently not due to Ritchie; he quotes an
> unavailable Air Force paper on Multics...
Ritchie != Thompson ;}
--
`Have a read, but keep a bucket handy' --- Alan Cox on UDI
------------------------------
From: "Gregory Lepere" <[EMAIL PROTECTED]>
Subject: Init pb : respawning too fast
Date: Mon, 6 Sep 1999 13:39:52 +0200
Hello,
I'm trying to embed Linux. So, I start with a minimal system.
But when I boot Linux. The kernel loading is ok. The init loading is ok.
It start with the inittab. And, it didn't start getty. It says :
"Id 1 respawning too fast: disabled for 5 minutes" when trying to start
getty.
I think getty is dying as soon as it starts up.
So, I check the librairies it depends on.
I update them to correct versions and do a ldconfig.
But it's always the same result.
Could someone help ? Thank you.
Greg.
------------------------------
From: Camil Coaja <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: encripted disk
Date: Mon, 06 Sep 1999 12:31:10 GMT
I would like to protect the data on a removable hard disk against
unauthorised acces from persons who might get phisical access to it.
Is there a way to transparently encript the data writen and decript the
data read from the drive in question?
Thanks,
Camil
================== Posted via CNET Linux Help ==================
http://www.searchlinux.com
------------------------------
From: Taketoshi Sano <[EMAIL PROTECTED]>
Subject: Re: [Linux] Calling C from Fortran: Function always returns zero?
Date: 06 Sep 1999 08:51:51 +0900
Hi.
In article <7qr4qh$[EMAIL PROTECTED]>
[EMAIL PROTECTED] (John H. Chauvin) writes:
> -------------------
> program Main
> integer result,number
I think, in FORTRAN, you must declare the function testit()
as type of integer.
I get the correct result with adding "testit" to the line above,
such that "integer result, number, testit" and all of the rest
is the same as you wrote.
And you can get the correct result with the change of the name
of that function from "testit" to "itestit". In this case,
you use the implicit type declaration of FORTRAN.
--
Taketoshi Sano: <[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED] (Scott Lanning)
Crossposted-To: comp.unix.programmer,comp.os.linux.development.apps
Subject: Richards Stevens died
Date: 5 Sep 1999 18:46:57 GMT
http://www.bigdealclassifieds.com/classified/plsql/
classlevel3_step?wClass=0002&wPubdate=Friday&wRowstart=1&wLessOrMore
--
Scott Lanning: [EMAIL PROTECTED], http://physics.bu.edu/~slanning
"How can we make our teaching so potent in the motional life of man,
that its influence should withstand the pressure of the elemental
psychic forces in the individual?" --Albert Einstein
------------------------------
From: Jonathan Guthrie <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: The conceptual sandbox?
Date: 6 Sep 1999 14:16:02 GMT
You seem to be writing to me, so I'll respond.
In comp.os.misc James Andrews <[EMAIL PROTECTED]> wrote:
> If your OS controls who gets to write to the initialisation scripts,
> then how exactly would a program foirce itself to be restarted with the
> machine/OS?
I think you misunderstand. I'm not talking about ANY program, I am talking
about THE program. What IBM calls the "control program" or what a Unix
architect refers to as a "kernel". To break your security, you build a
compiler for whatever language your kernel is written in that has some
advantages (such as producing faster or smaller code---there's always room
for improvement in the output of a compiler) and has one big disadvantage:
It removes the kernel's anti-corruption code.
When you reboot to take advantage of the faster kernel, your system becomes
trashable. Since the kernel MUST be run at system startup time, there
isn't a question of controlling initialisation scripts and no implementation
defects are required, either. There are, no doubt in my mind, other ways to
accomplish the same purpose, but they require knowledge of just how your
system is intended to work, which even you don't know.
> What it does represent is
> a conglomorate of ideas that have not been combined quite in this way
> before, to use the whole, not the sum of the parts, to combat some age
> old problems. Whether it is the answer or not is irrelevant, its the
> trying that counts. The fear of failure and responses like those
> Vladimir has had here have resulted in less people trying to push the
> boundaries. If we all failed a little bit more, maybe we'd get
> somewhere,
If the potential for failure stifles attempts to "push the boundaries"
well, we certainly haven't seen that here. Vladimir hasn't DONE anything
so how can his efforts have failed? Being ridiculed is not failure. You
can, of course, claim that fear of ridicule is stifling creativity, but I
wouldn't believe you. You only get ridicule when you persist in telling
people of real experience that their experience is wrong without having
some real good evidence to back it up. If Vladimir and yourself are
ridiculed, perhaps you should spend more time trying to understand what
we're saying and less time telling us that WE'RE the fools.
An example from my life might prove illustrative:
I'm an idea person. I generate ideas for a living and I recognize that
90% of them are crap. Therefore, when I have an idea, I have to determine
whether or not it is worthy of development. Usually, it isn't.
Sometimes, I have an idea that I can't properly evaluate by myself, so I
seek out some experts to evaluate the idea. Usually, in those cases, the
idea is judged to not be likely to work or it is some standard part of
"the art" that I, as an outsider, wasn't aware of. Recently, I had an
idea that I thought would allow computers to work faster. However, I am
not at all skilled in computer design. When I went to someone who is
knowledgeable in supercomputers and described my scheme, he informed me
that what I had come up with (in this case, connecting multiple processors
with multiple memory pools through a crossbar switch) is not new. I,
therefore, dropped any ideas of developing the idea.
It is important to understand that I dropped the development not because
the idea was bad, but because the idea was well established among those
people whose job it is to develop high-speed computers. In order to
understand whether or not what I was working on was truly an advancement,
I would have to learn what the experts already know. That makes the odds
of success much lower than they might otherwise be. I've got other ideas
and a finite life in which to develop them. This is not high on my
priority list. (Although it did spark an effort to produce a digital
logic simulator so I could build in a program what I would have difficulty
building on protoboard.)
I can mention this in public and not be ridiculed because I don't
represent it as something that it is not. It IS a good idea, it's just
not all that original and anything that results from it is unlikely to
be a major advancement of "the art".
Should Vladimir pursue his goal? Why not, it's his life. Should you
help him? Why not, it's YOUR life. Should you both stop calling me
unimaginative because I think what you're doing is unlikely to advance
the state of the art? YES! We don't believe that either you or
Vladimir understand enough about what you're talking about to complete
a significant amount of your project. Although I can't speak for the
other fellows of real experience, I believe that, should you ever decide
to fix the structure of the Control Program and then implement it, you
would have a lot better appreciation for what we're saying.
Come back and show us after you're done. We'd like to see what you get.
--
Jonathan Guthrie ([EMAIL PROTECTED])
Brokersys +281-895-8101 http://www.brokersys.com/
12703 Veterans Memorial #106, Houston, TX 77014, USA
------------------------------
From: [EMAIL PROTECTED] (David T. Blake)
Crossposted-To: comp.os.linux.hardware,comp.os.linux.portable
Subject: Re: Can I compile the kernel using a cc other than gcc?
Date: 6 Sep 1999 13:58:27 GMT
Reply-To: [EMAIL PROTECTED]
Martin Kahlert <[EMAIL PROTECTED]> wrote:
> Yes they do: Compaq has a beta version of a fortran compiler
> for Linux AXP. It's a port of DEC's own compiler for their
> Unix for Alpha.
> I tried it on a 1 MB source analog circuit simulator
> and the binary was a factor of 2 faster than g77's
> (from gcc-2.95.1).
Ahh that would make sense too. The gcc implementation of the
ForTran compiler is pretty green, and Dec has been working
on their for a long time with optimization in mind.
--
Dave Blake
[EMAIL PROTECTED]
------------------------------
From: Mladen Gogala <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.development,comp.os.linux.help,comp.os.linux,comp/os.linux.misc
Subject: Re: what's a "jiffy" in /proc?
Date: Mon, 06 Sep 1999 12:02:32 -0400
Virginie Galtier wrote:
>
> Hi
>
> The proc man indicates that in /proc/<pid>/stat, utime is "The number of
> jiffies that this process has been scheduled in user mode."
>
> But I found 2 definitions for "jiffy" (in
> http://foldoc.doc.ic.ac.uk/foldoc):
>
> 1. The duration of one tick of the computer's system clock.
> 2. Confusingly, the term is sometimes also used for a 1-millisecond wall
> time interval.
>
> So in the case of /proc/<pid>/stat, should I consider the first
> definition or the second?
>
> Thank you for your help
>
> Virginie
Jiffy is one tick of the system clock (10 msec).
--
Mladen Gogala
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: Flamage - Why?
From: Errin Watusikac <[EMAIL PROTECTED]>
Date: 06 Sep 1999 09:09:03 -0700
James Andrews <[EMAIL PROTECTED]> writes:
[snipped]
> Yes, people in general like to see the colour of money. And it seems
> code is the currency of acceptance on this group, hence I have yet to
It seems it's also been accepted by USA law makers. Check out this
quote from the "definitions" part of USA copyright law (USC17):
The term ''financial gain'' includes receipt, or expectation of receipt,
of anything of value, including the receipt of other copyrighted works.
Licensors who write of "cost", "money", "non-commercial", etc, must hope
courts won't accept this concept when interpreting their license.
------------------------------
From: Fernando Ortega <[EMAIL PROTECTED]>
Subject: Bottom Half(NET_BH)
Date: Mon, 6 Sep 1999 17:35:11 +0200
When a bottom half routine is interrupted, what happens?
Will it attend the interrupt?, and then go back to the routine at the same
point it was left?
My question focus on the NET_BH bottom-half, I have following the code,
and I think the only point where the NET_BH is marked is the function
"netif_rx". Is that right?
If so, does it mean that to empty the drive queue the only way to do it is
by receiving a pakect , so that it makes "netif_rx" to mark NET_BH?
Do you know any docs, where all these matters are explained?
Thanks a lot in advance
--
Fernando Ortega Bellosta
[EMAIL PROTECTED]
------------------------------
From: "Bert Douglas" <[EMAIL PROTECTED]>
Subject: want to move my russian keyboard to linux
Date: Mon, 6 Sep 1999 09:43:44 -0500
Hi all,
I just got my first Linux box (mandrake) yesterday. I used to be a systems programmer
on windows about 10 years ago, working on
printer drivers.
A while back, I made a custom russian KBD file for win95. This KBD file uses (some
would say abuses) the "dead key" feature
(intended by msft for umlauts, accents, etc) so that you can type certain patterns of
two keys to generate just one russian letter.
You can see my keyboard at:
www.dibbs.net/~bertd/
This is necessary because russian has a few more letters, and I did not want to lose
the ability to type punctuation marks in
russian mode. Also I tried the various existing russian keyboards and could not get
my typing speed up. I ended up with my
arrangement after lots of experimentation to find what works best for me.
Anyway I am accustomed to this way of typing russian after much use, and I want to
bring it to my new environment, Linux.
So a few questions:
1) Where does this functionality live in linux?
2) Can I do this by editing some table, or do I have to write code?
3) Once done, how do I (or should I) offer this little snippet for inclusion in
distributions?
Thanks,
Bert Douglas
------------------------------
From: James Andrews <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: The conceptual sandbox?
Date: Mon, 06 Sep 1999 16:19:33 +0000
Several things:
Firstly, your post was unnecessarily agressive. I also don't see
anything in your previous posts that has disproved in ANY way, the
concept of a sandbox model, so there is no justification for your
attitude.
Secondly, your little idea of the fallibility is, quite frankly,
bollocks. Firstly, why would a compiler have total priveledges?
Secondly, why would a compiler on ANY system have access to startup
scripts? In a sensible system the "control process" should do just
that, nothing else. Any task which wishes to attempt to breach the
ordained procedure can be monitored. Maybe in an existing OS, obviously
of your choice, this would be the case, but we are talking theoretics.
As for a kernal compiler, if you want security, you dont go and use
someone elses compiler for the kernal. Thats not a security flaw, thats
stupidity. And it isnt technically a breach of the system either, as it
means the breach is there before the system. You wouldnt be so much
hacking the system, as providing a faulty installation. If you intend
to post another resoning why I am wrong, please make sure it actually
quotes some facts, rather than assuming design flaws of older systems (I
believe you refer to several cases of un*x being hacked via the kernal
compiler... why do you assume a compiler would be involved? And why
would it be changed after the OS had been written to this phantom rogue
compiler?)
Thirdly, this is not about sides, attempting to extend the scope of your
argument by using "we" doesn't make you right.
Fourth, I am not working for or with Vladimir directly, I am an
independant person engaged in this discussion. I came into this
discussion on the basis that I knew something of the subject. I wished
to discuss civilly. Since you are no longer civil, this will be my last
post in reply to you. The fact that you made such a ridiculous claim,
only a few sentances along from claiming *again* that you know nothing
of the subject somewhat voids your commentary.
And as for the code situation, this is not a discussion about
implementation. The purpose is to knock out a fair plan before starting
so that no time is wasted coding a flawed system.
We are discussing a concept. We put forward points. You tell us they
wont work. Why? No justification, you have just decided that in your
mind, the concept is infallible. The irony of the situation is that the
entire reasoning behind your last, highly flamable, post was that we
don't listen to your reasoning, but in the same instance you get highly
agitated and flame people for putting across a good point. I can't wait
to see the next one!
I find you objectionable and if you resent having to provide basis for
your unfounded claims then dont make them.
If you don't like what we discuss, or how we discuss it, then mail
filter us out. We'd try not to feel *TOO* hurt at losing you.
James
------------------------------
** 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
******************************
