Linux-Development-Sys Digest #862, Volume #6 Tue, 22 Jun 99 19:13:53 EDT
Contents:
Re: TAO: the ultimate OS (Vladimir Z. Nuri)
Re: TAO: the ultimate OS (Vladimir Z. Nuri)
X-Shell Development Environment ("Bob Bryla")
Re: TAO: the ultimate OS (Vladimir Z. Nuri)
Re: TAO: the ultimate OS (Vladimir Z. Nuri)
Mmap, make pagefault at every pointer access? (Niko jokinen)
Re: Guides to Linux Performance Tuning??? (James Stevenson)
Semaphores don't compile, please help (Stanislav Krasilovskiy)
Re: Ultimate OS (Konrad Mieredorff)
Re: TAO: the ultimate OS ("Anonymous")
Re: help with SAMBA pwds (Konrad Mieredorff)
Clist buffer in linux kernel? (Eric Fowler)
Re: gcc byte packing of inherited class data (Bruce Edge)
PCMCIA eth card does not work correctly, please help. (Mike Rabinovich)
apache problem: cant run c programs from scripts (peter marshall)
PCI device -- base address doubt (Rajarshi Bandyopadhyay)
Re: You can now use Winmodems in Linux!!!!!!! (James Stevenson)
----------------------------------------------------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: TAO: the ultimate OS
Date: Tue, 22 Jun 1999 19:55:35 GMT
void ([EMAIL PROTECTED]) wrote:
: On Mon, 21 Jun 1999 08:53:44 GMT, Vladimir Z. Nuri <[EMAIL PROTECTED]> wrote:
: >
: >yes, Unix clearly epitomizes the
: >sort of "real men don't design, they just start hacking"
: >charade you allude to above.
: As compared to what, hardware?
as Terry Murphy was saying, in good design
requirements -> design -> detailed design -> prototype -> beta -> release
or something like that..
now.. where is the unix requirements document? who wrote it?
if there was a unix requirements document or emphasis on
writing /maintaining it, there would be less hostility to
my own requirements document.. but unix programmers do not
think in terms of the above. instead we have, as ESR
glorifies,
code -> release -> code -> release -> code -> release
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: TAO: the ultimate OS
Date: Tue, 22 Jun 1999 20:04:02 GMT
Terry Murphy ([EMAIL PROTECTED]) wrote:
: The "rapid prototyping" philosphy is very tightly ingrained into
: Unix, into the doucmentation, into the way the tools are used,
: and into the tools itself. This is something, which, IMHO, we
: need to get very far from.
well, one way to look at it I hinted in the essay
is that the lower level components need to stabilize first.
so we have about 30 years of insight into the design of microprocessors.
one level above that is the interface of devices to the computer,
cards, etc.. even high levels, the highest of all, involve the OS
and applications that sit on top of it.
now I can see that all are converging toward perfection, but the higher
level components are more recently devised. it was really only about
5 years ago that the GUI OS+mouse was considered the standardized system.
i.e.-- the consensus is slowly building, as well as discipline !! once
you have a consensus/confidence about what an OS or chip really is, then you
can have much stronger discipline & standards.
so yes, I think as you move up the levels, you get more into the
"wild west" mentality.. which linux & win95 embody to varying degrees
(we would probably be aghast at what goes on behind the scenes at
MS relative to win95 development etc)
: But I guess that was my point, which I didn't state explicitly:
: DESPITE software engineering being more complex than chip design
: and many other engineering disciplines, it is treated in a much
: less professional/methodical manner than the other engineering
: disciplines.
agree, and I propose that a lot of this is centered around the microsoft
attitude toward OSes.. imho they have focused on churning out
products to maximize market share where the discipline is lacking
(although I admit historically their code discipline has been higher than
competitors, imho, such as netscape)
but as I state in other msgs: the bar is continually being raised.
we will see if the MS culture can continue to adapt to "more
discipline".. they have succeeded somewhat with NT, I acknowledge.
: I do, of course, hope that this is the future of operating systems,
: and software in general. Taking all of the theoretical work on this
: and turning it into a product which you can build, however, is an
: extremely difficult task, but I wish you the best of luck.
its defn not a weekend code warrior project.. !! but thanks
for the encouragement. a few drops in a parched desert may be
enough to sustain me yet<g>
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
From: "Bob Bryla" <[EMAIL PROTECTED]>
Crossposted-To: linux.dev.x11,comp.os.linux.x,comp.os.linux.alpha
Subject: X-Shell Development Environment
Date: Tue, 22 Jun 1999 15:17:36 -0500
Hello,
I'm going to do some rapid application development for X-Servers, and don't
have the time (nor the skills) to do a lot of low level C++ calls to
X-Libraries.
I'm looking for a "shell" or "scripting language" of some type, that runs as
an X-Windows app, that will allow me to do things like: (1) do a browse of a
file system, and allow multiple select of filenames; (2) submit cron jobs
via running a c-shell, tcl or perl script; (3) receive e-mail or send e-mail
based on a simple evaluation of a text file contents; (4) create menu
systems for selecting other simple forms; and so on. Kind of like a high
level toolset for X. I can settle for one that can be called from a very
simple C++ program, but would be better with some kind of easy scripting
language, or from perl, or whatever.
I'm going to develop on Linux, and will eventually end up on a DEC Unix box,
for a while anyway, so source code for this stuff would be highly desirable
also. I can handle changing make files or looking for libraries or whatever.
So, any freeware or shareware, or cheapware out there that has at least some
of these things??? My searches have been fruitless up to this point...
For those of you who are familiar with WinBatch for Windoze, that's pretty
close to what I'm looking for in an X-Environment.
Bob Bryla
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: TAO: the ultimate OS
Date: Tue, 22 Jun 1999 20:06:11 GMT
I joked about LT's "authority & credentials" & then some humor-impaired
individual wrote:
: Not to take any respect away from Linus, who certainly deserves our
: respect...
: ...but you are not going to earn anyone's respect by insulting us like that.
: FYI, some of us were designing and writing operating systems, device drivers
: and programming languages before Linus was out of kindergarten.
the insult is in the eye of the beholder.. and who is insulting who?
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: TAO: the ultimate OS
Date: Tue, 22 Jun 1999 19:51:44 GMT
Konstantin Koll ([EMAIL PROTECTED]) wrote:
: That would be cool, but I don't think major companies will program drivers for
: a software that has only some hundred users.
well, lacking that (I wasn't expressly calling for that)... a good/clean
driver interface that has a strict/exact boundary with the OS is
a start... and some tolerance for buggy drivers in the OS.. those are
some of the goals I outlined that can be addressed without 3rd
party support.
: Sorry.
no problemo.. if you have written an OS and like the Tao essay,
then it is I who should be thanking you for your support, which
I did/do!!
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
From: Niko jokinen <[EMAIL PROTECTED]>
Subject: Mmap, make pagefault at every pointer access?
Date: Tue, 22 Jun 1999 22:47:48 +0300
Hi,
I am making a driver for hardware which used to run on proprietary
DOS-based OS. So this was handled by the OS.
Now I need this on linux. Hardware is visible at 0xb0000, 64KB window
divided in 4KB sections, but here is the catch: By writing to 0xf180
64KB window can be switched between 4 different cards.
Driver has major number 120 and minor numbers determine which is the
right 64KB window(minor numbers 1-16 should outb(0x1, 0xf180) 20-35
should outb(0x2, 0xf180) and so on...)
I have mmap and nopage functions working fine, but since user space
programs don't have any knowledge about each other, everytime one
accesses the pointer returned with mmap(), driver should execute
function which changes 64KB window according to minor number.
I guess nopage() is best possibility, but how could previous page be
invalidated, so that a pagefault would occur every time.
Any bright ideas anyone? I have worked with this a while, so any help
would be _greatly_ appreciated. If this is impossible I'll do it with
lseek/R/W and a library for userland programs (More work, not good :)
--
Niko Jokinen
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (James Stevenson)
Crossposted-To: comp.os.linux.misc
Subject: Re: Guides to Linux Performance Tuning???
Reply-To: [EMAIL PROTECTED]
Date: Tue, 22 Jun 1999 14:26:09 +0100
Hi
read the document in the kernel source
Documents/proc.txt
it will exzplain each of these files and parms
On Tue, 22 Jun 1999 10:10:50 +0100, Jon Skeet <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] wrote:
>
>> I'd like to locate any references or guides on Linux system
>> performance tuning. I'd assume, to receive acceptance as an
>> enterprise level solution, there should exist an easy infrastructure
>> for tuning the OS. For example, Solaris and IRIX have the most
>> important kernel tunable parameters located in a central file. Those
>> OS's also include utilities that help assess bottlenecks in the memory
>> and network subsystems. Obviously these capabilities are of prime
>> importance for the sale of enterprise systems and are often developed
>> by the platforms' vendors!
>
>Under Linux 2.2 and above (at least - probably earlier too), you can use
>/proc/sys/* to tune many parts of the kernel.
>
>For instance, to change the total number of file handles which can be
>open at a time, you might do:
>
>echo 16384 > /proc/sys/fs/file-max
>
>Note that if you want to do this on a permanent basis, you should put
>lines like that in rc.local - they aren't automatically preserved over
>reboots.
>
>> Although my understanding is that VA Research and others offer high
>> performance Linux systems with support, I was hoping there might be
>> freely available guides on the topic of system tuning for a freely
>> available OS. I tried doing some quick searches on performance tuning
>> for Linux and found that there is very little documentation on the
>> subject. And, the little information I find is rather outdated.
>> Also, it appears that kernal tunable parameters are located in header
>> files without any documentation. Is this correct?
>
>There are very few (that I've come across and really wanted to change)
>that are only in header files. The default values are probably in header
>files...
>
>The documentation for most of this is in
>/usr/src/linux/Documentation/sysctl
>
>Hope this helps - I've been mucking around with this a bit recently. I'm
>far from an expert, but it's not hard when you get into it...
>
>--
>Jon Skeet - [EMAIL PROTECTED]
>http://www.pobox.com/~skeet/
--
=============================================
Check Out: http://www.users.zetnet.co.uk/james/
E-Mail: [EMAIL PROTECTED]
2:00am up 1 day, 3:03, 0 users, load average: 0.08, 0.02, 0.01
------------------------------
From: Stanislav Krasilovskiy <[EMAIL PROTECTED]>
Subject: Semaphores don't compile, please help
Date: 22 Jun 1999 20:33:27 GMT
Hi,
I am trying to use the Linux semaphores as defined in <asm/semaphore.h>. Even
though the functions "up," "down," and "down_interruptible" are explicitly
defined in Assembler in the header file itself, the linker complains that
it cannot find these functions. Could you please help me out?
Thanks
--
+------------------------+----------------------------------------+
| Stanislav Krasilovskiy | http://mdsp.bu.edu/prgrssor_html |
| AKA | "Even the mosquitos don't bite me, |
| "The Progressor" | because they know I'll bite back." |
+------------------------+--------=---------------------=---------+
------------------------------
From: Konrad Mieredorff <[EMAIL PROTECTED]>
Subject: Re: Ultimate OS
Date: Tue, 22 Jun 1999 22:34:03 +0200
RaiX wrote:
>
> Linux with netscape as GUI.
> That's it..
>
> Regs RaiX
Should I laugh or cry?
------------------------------
From: "Anonymous" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: 22 Jun 1999 19:27:47 GMT
Vladimir Z. Nuri <[EMAIL PROTECTED]> wrote:
> most of above are subject to the evolutionary pressures of extinction.
> - programmers who do not understand how users think will go
> extinct.
Not every programmer needs to be a UI guru (I interpret your use of "user"
as end-user).
> well, I think you are throwing out red herring.. it is not responsibility
> of user to implement features, and a common canard of programmers to
> complain that *difficult* features are *impossible*-- exactly the
> reaction I've received to my original post.
Rather that those features themselves are not well thoughtout. A feature
list shouldn't cross the boundaries of features and mention such things as
"object-oriented system" without ever explicating their relationships to
the rest of the system. To me features mean capabilities, not an incomplete
design.
> but imho it is largely responsibility
> of programmer to anticipate what users want without requiring them
> to articulate it explicitly. true? the greatest designer gives you
> something you didn't even know you wanted!! and you realize, only
> upon seeing it, that you can't live without it<g>
What users want depends greatly upon their prior experience and this means
not everyone wants the same set of solutions. This of course means a
programmer, with his/her own experience, is very likely to err according to
your standards.
> but let's not get stuck. the bar will continually be raised. Linux
> is an interim solution, ultimately-- as is every other OS or program
> known to man. what are are these new currents? I attempted to
> articulate them concisely.
There is a continuum between implementation and goal (of which the ultimate
is human happiness, at least for me), and your articulation was too thinly
spread along this continuum, in my opinion, with dubious effects.
> but I assure you that ultimately Linux will eventually be replaced
> by something that does, if it does not adapt. (not very quickly,
> however). also-- how much of recent industry hype & excitement
> & media coverage is centered on the idea that Linux is becoming more
> accessable to the "dumb users"? imho, MOST of it. will the founders
> of Linux recognize this and surf with it?
The founders of Linux do not design the UI and I'm glad that they aren't
stuffing the OS core with such peripheral, however central to end-user
computing experience, parts.
> or will the linux community
> become more insular & disdainful of "dumb users"
> as a community of programmers? I'd say its a tossup
> right now ..
You cannot dumb down a computer below a certain level without sacrificing
its functionality. The reason why computer is generally hard to use is
because it represents such a huge number of functionalities, making it
extremely difficult to present one generic interface through which all
these features can be accessed. Thus thousands of different interfaces with
conflicting paradigms proliferated, generic books like "how to use a PC"
notwithstanding. Heck, when I first learned about computers I thought I had
to learn programming to do all these cool stuff.
> yep. I agree. but there is a whole new realm waiting to be explored.
> bringing an OS to the unwashed masses. at least you are beginning
> to acknowledge that such a priority or goal exists.. and that Linux
> does not fulfill it.
Either you give them a small set of functionalities to begin with, or make
them go through an extensive training.
------------------------------
From: Konrad Mieredorff <[EMAIL PROTECTED]>
Subject: Re: help with SAMBA pwds
Date: Tue, 22 Jun 1999 22:37:17 +0200
Jun Yang wrote:
>
> Are you using NT 4.0? If so, your password sent to Linux server is encrypted.
> You need to disable this option in order to send plain text passwds to your
> Linux server. To do this, in your NT, run regedt32 and go through all the
> following:
>
> HKEY_LOCAL_MACHINE
> SYSTEM
> CurrentControlSet
> Services
> Rdr
> Parameters
>
> and add the following line there:
>
> EnablePlainTextPasswork : REG_DWORD : 0x1
>
> I solved my a similarpasswd problem when I use SAMBA to connect my NT to a
> sgi server.
>
> "..Luca T.." wrote:
You can also enable password encryption in your smb.conf:
encrypt passwords = yes
------------------------------
From: Eric Fowler <[EMAIL PROTECTED]>
Subject: Clist buffer in linux kernel?
Date: Tue, 22 Jun 1999 16:47:45 +0000
I am reading Pajari's book (Writing UNIX Device Drivers). He uses a
thingie called a clist buffer, which is a kernel-defined and maintained
buffer list. I would like to use it on RedHat 5, but can't find it in
the header. Is such an animal defined for RedHat kernels?
--
Vivez sans temps mort!
(Live without dead time)
-Situationist International
sockeye [at] rmii [dot] com
------------------------------
From: Bruce Edge <[EMAIL PROTECTED]>
Crossposted-To: gnu.gcc,gnu.g++,comp.os.linux.development
Subject: Re: gcc byte packing of inherited class data
Date: Tue, 22 Jun 1999 21:53:03 +0000
Bruce Edge wrote:
>
> I can't get gcc to pack this data:
>
> class a
> {
> char c;
> };
>
> class b
> {
> long l;
> };
>
> class c : public a, public b
> {
>
> };
>
> Without going into why I need a misaligned long after a char,
> I need the alignment for class c to be:
>
> cc ll ll ll ll
>
> not:
>
> cc xx xx xx ll ll ll ll
>
> which it insists on doing regardless of which attribute/packed
> parameters I use.
>
Just for the record, I'm constructing a packed to send out a serial
interface to a 16 bit controller, so the alignment is really important.
Adding packed attributes all over doesn't help either.
-Thanks, Bruce.
------------------------------
From: Mike Rabinovich <[EMAIL PROTECTED]>
Subject: PCMCIA eth card does not work correctly, please help.
Date: Tue, 22 Jun 1999 15:59:32 -0600
Hello, i have installed linux on IBM thinkpag 560E, i user 3com
3CCFEM556B
10/100 ethernet card, which works fine under windows, in linux link
light goes off,
but i can't ping anything but myself, i am pretty sure network is
configured correctly.
any suggestions?
Mike.
------------------------------
From: peter marshall <[EMAIL PROTECTED]>
Subject: apache problem: cant run c programs from scripts
Date: Tue, 22 Jun 1999 16:36:59 +0000
6/22/99
To whom it may concern:
I am having problems configuring my Apache-US-SSL-1.3.3-1 server that
came with my Redhad 5.2 box, running on
an intel platform (pentium pro 150, 125M Ram,two hardrives (hda and hdb
with lots of partitions), matrox millenium II, NE2000 ethernet card,
etc.
The problem:
when the apache server calls a perl script I am able to successfully
call
shell scripts from within the perl script, eg:
$returnme = [back tick]/home/user/bin/shellscript[back tick];
works fine.
But when I try to call a binary compiled in C I get nothing, eg:
$returnme = [back tick]/home/user/bin/cprogram[back tick];
fails.
Please help.
Peter
------------------------------
From: [EMAIL PROTECTED] (Rajarshi Bandyopadhyay)
Subject: PCI device -- base address doubt
Reply-To: [EMAIL PROTECTED]
Date: Tue, 22 Jun 1999 19:12:23 GMT
Hi all
I am trying to program this PCI card. I have accessed the card
params using pci_find_device() call. This returns the card
info in a pci_dev struct. Now in /proc/pci my card
shows up as
I/O at 0x6200 [0x6201]
I/O at 0x6300 [0x6301]
The addresses 0x6201 and 0x6301 show up as
base_address[0] and base_address[2] of the pci_dev struct.
What do I use as by base address of the card since
all registers etc of card are defined rel. to that?
Please cc ur replies to [EMAIL PROTECTED]
Sorry for the inconvenience.
Thanx for ur help,
--Raj
--
================================================================
Let me tonight look back at the span
'Twixt dawn and dark & to my conscience say:
"Because of some good act to beast or man,
The world is better that I lived today."
---Ella Wheeler Wilcox
=================================================================
------------------------------
From: [EMAIL PROTECTED] (James Stevenson)
Subject: Re: You can now use Winmodems in Linux!!!!!!!
Reply-To: [EMAIL PROTECTED]
Date: Tue, 22 Jun 1999 20:31:34 +0100
Hi
i am not invloded in this but has anyone asked the manufacuters yet??
On Tue, 22 Jun 1999 13:07:56 -0500, Medical Electronics Lab
<[EMAIL PROTECTED]> wrote:
>Frank v Waveren wrote:
>> Can we change the name to hoax? This is some guy who thinks he's funny,
>> there are currently no winmodem drivers for linux :-(
>
>So let's change that! I bet if we can help the modem manufacturers
>sell an extra 1000 modems a month they'd be happy to help us write
>the drivers. I ain't no salesman, but I can deal with modem tones
>no problem!
>
>The winmodem is just a digitizer that does DMA into a buffer. The
>processor has to convert the audio bits to multiple tone data, and
>then to data bits. It's straight forward data manipulation. There
>is no reason we can't write a driver for any winmodem.
>
>Who wants to go where no one has gone before?
>
>Patience, persistence, truth,
>Dr. mike
--
=============================================
Check Out: http://www.users.zetnet.co.uk/james/
E-Mail: [EMAIL PROTECTED]
2:00am up 1 day, 3:03, 0 users, load average: 0.08, 0.02, 0.01
------------------------------
** 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
******************************
