See also: last week's Back page page.

Linux links of the week

Watch out, it's the Microsoft Matrix, a story comparing our favorite big software company with the current science fiction film. "Like Keanu Reeves, most people's eyes will hurt when they first look at the real world, because they've never used those eyes before. But I've chosen that real world, because while the Matrix of Linux has rules and regs every bit as stern -- and often sterner -- as the Matrix of Windows, that Big Difference pops up: unlike the Microsoft Matrix, you can hack the Linux Matrix from the inside..."

Warning: non-Linux item ahead. But this was too much fun...it turns out that the famous "blue screen of death" really is configurable. Now there's a handy tool out there which will let you tweak the colors to something more pleasing. If you can't install Linux, this is probably the easiest way to abolish the BSOD forever. (Found, like most of the really fun stuff, in NTK).

Section Editor: Jon Corbet

Letters to the editor

Date: Fri, 23 Jul 1999 09:51:58 -0700
From: Eric Lee Green <eric@estinc.com>
To: corbet@eklektix.com
Subject: Keeping system software up-to-date

While I've had my reservations regarding Penguin's "in your face" approach in
the past (as I noted in one LinuxToday commentary on a Penguin press release,
"aggressive is one thing, but you went over the top into sheer ridiculousness
with this press release"), I really can't fault them for shipping an
out-of-date CD-ROM. I was in charge of keeping those things updated at Linux
Hardware Solutions and I can attest that it was NOT an easy task. Often we'd
build a machine, then it could sit around for as much as two weeks waiting for
the one last part that was backordered at the supplier... given that the
CD-ROM generally was two weeks behind updates.redhat.com in the first place,
this means that we could be as much as a month behind Red Hat by the time the
machine shipped. 


-- 
Eric Lee Green                         eric@estinc.com
SysAdmin/Software Engineer             Visit our Web page:
Enhanced Software Technologies, Inc.   http://www.estinc.com/
(602) 470-1115 voice                   (602) 470-1116 fax

Date: Sat, 24 Jul 1999 00:18:59 +1000
From: Steven Hanley <sjh@wibble.net>
To: corbet@eklektix.com
Subject: your penguin computer review on lwn

I notice you mentioned doing a make -j2 for building the kernel source
tree on your new dual box.

If you read the linux/Documentation directory a bit more you will notice
some information about using SMP and building kernels.

Basically it is reccomended when using make -jN that N be number of
cpu's + 1

this allows one make to be a task controller or similar (like in the
linux pthreads implememtation with it starting an extra thread as a
controller sort of)

The reason this is recommended is because it does work. if you do some
timing tests of builds you iwll probably notice it is in fact faster on a
dual box for example to do make -j 3 whenever building any source tree not
just the kernel. (well any sufficently large tree)

Also if you are in the mood you can infact set global make options so
anytime you type make it will infact automatically do -j3 or whatever
among other things.

	See You
	    Steve

-- 
sjh@wibble.net http://wibble.net/~sjh/
Look Up In The Sky
   Is it a bird?  No
      Is it a plane?  No
         Is it a small blue banana?
YES

Date: Sun, 25 Jul 1999 19:17:36 +0100 (WET)
From: =?ISO-8859-1?Q?Jo=E3o_Miguel_Neves?= <jneves@rnl.ist.utl.pt>
To: lwn@lwn.net
Subject: IDC study - Attention


	About the study that predicts 40,000 users in western europe by
2003 if MS doesn't port Office to Linux, I would like to call you're
attention that counting the users in counter.li.org for some western
europe countries I got to the sum of 33070 registered users, and we all
know that's not by far all users.

	I'd just like to call your attention to this because this is
supposed to be a serious study, but looks like FUD.

	My data came from http://counter.li.org/reports/short.html and is
below:

es:2934
pt:639
ad:6
gr:475
it:2125
dk:2361
gi:8
nl:3047
at:1282
ch:939
lu:48
be:1162
de:8534
ie:361
mc:3
gb:4666
fr:4480

	Thanks for your time and keep up the great work,

		A daily and weekly reader,

						João Miguel Neves
				       http://camoes.rnl.ist.utl.pt/~jmne/

Date: Mon, 26 Jul 1999 11:50:20 -0400
From: walt smith <waltech@bcpl.net>
To: letters@lwn.net
Subject: quirk issues

Hello,

I am an "occassional" Linux user.  My length of time
is a bit over 3 years with several linux distributions.
Like everyone else, I visit many Linux oriented websites.
I'd like to raise three questions based on the above:

1. Most Linux "intro" zine articles and most website descriptions
seem to not mention the FSF and the GNU project.  While I
don't have first hand knowledge about the development effort,
it seems to me that FSF/GNU is getting the short end of the stick.
Am I right?  The "old-timers" will have to correct me if not.

I do believe that Linux would not be what it is without Linus T.
being the right person in the right place at the right time.
Couldn't equal billing be given by authors?  Perhaps mentioning
L.T's name first because he was the catalyst?

2. I browsed some of the code.  Even some of the kernel code.
No, I don't pretend to know what it's doing. I've touched a bit
of assembler and "C" in the past.  In addition, I've talked with
really senior software people.  What's the point?

/*   comments.  */

I see very few.  I'd like to put aside the fact that coders 'read' the
code and the statements are self-documenting. Horsefeathers!  I'll bet
each each of the coders involved have a zillion comments (perhaps not
written).  If not, then they are really geniuses.  But that leaves out
thousands of more normal folk who would like to be involved.  OK, so
normal folk won't write kernels...but - involvement is necessary for
the rest of the community. Period.

2a.  Which brings up comment point 2a: There's a lot of 2.11.x.ac5
type patches.  As an example. There is an extreme history of patches
for obscure reasons.  Yes, the patches are necessary.  Are the
comments regarding why's and wherefores in the code?  With details?

2b.  Which brings up comment point 2b: Are comments covered under the
GPL?  Enforcement should be interesting. Maybe comments could be made
compileable to insure they are there (YES! - I say this with
tongue-in-cheek!).

3.  Why don't Linux product sites say what the product is?  Neither do
the ezines!  OK, so Python x.x was released. It's snaked code, right?
I think I've seen maybe two sites out of, what, 20, 30...  that had a
brief description...in technobabble.  At least give a one-liner in a
ezine announcement!!!!  How about a short descriptive paragraph on the
home site?  I've often gone 3,4,5 levels of mouse clicks and can't
figure out what the Hades they're selling!  Marketing Managers take
note: what do I do?  I give up and go elsewhere.  These sites look
like many web pages put up by local governments.  OK, so I'm in
Rockingham County..Virginia or NewYork or Ohio?

I'm looking forward to the continued improvement of the Linux community.

regards,

Walt Smith, Baltimore

Date: Mon, 26 Jul 1999 12:20:57 -0700
From: Anand Srivastava <anand@nmi.stpn.soft.net>
To: metcalfe@idg.net, letters@lwn.net
Subject: Crafty benchmarks and no Unix certification undercut open-sourcerers

Hi,

Couldn't get to understand the point of the article. What were you
trying to say. It says on the top that it was an opinion article, but I
couldn't get any opinions. There are a lot of things that you have said,
in the article. I don't think that I can disagree with anything which
you have said or even agree, because I don't see an opinion. Only it
feels that you don't have anything good to say about us, maybe I am
biased because of your previous articles.

You talked about the mindcraft debacle and linked with the netcraft. You
seem to have praised netcraft, but could as well be a satire. Opinion
about mindcraft is not clear at all. What did you want to say, are we
hypocrites, I couldn't it really.

Then you point about Gurley, and then talk about his investment in
Redhat. Are you saying that he is using his article to inflate Redhats
stock to get better money, but really he doesn't believe in Redhat.

Then you talk about ESR's response to TOG. The general feeling among us
is that TOG is an organisation which doesn't care about the software it
maintains, only about the companies it represents. If X is not improving
as fast as it should, they are to blame. They even tried to stiffle
Linux by trying to change the license, because the other Unix
manufacturers are feeling the pressure of Linux. But the move backfired
terribly, because XFree86 people decided to go alone. Actually that move
would have been better. Later they saw the light, and came back. This is
what is happening with Unix certification. Although linux is more Posix
compliant than some of the other Unices, TOG will not grant the
certification to linux, because the paying Unix manufacturers don't want
Linux becoming legitimate. If that happens they will be dead in the
lower server end, which has a much bigger and more lucrative base. These
manufacturers are getting squeezed towards the top. As linux gets better
they will only be squeezed further. And anyway Linux distributers
already enough on their hands to really think about the unix
certification. Actually its more important for them to do linux
standardization than doing unix certification.

There are a couple of next generation OSs, heard of the HURD. Of course
its taking forever to work because Linux is taking away the resources.
But then since the source is free somebody can take it up later when
Linux is really finished with the most interesting parts, and it has
really taken over the world ;-). Then there is Freedows. It started with
a good idea, but seems like it has gone wayward. But then there will be
people later for that too. For now linux is the most interesting thing.

You are right that those questions are haunting and they are haunting to
the TOG, they are going towards obsolesence, and to the proprietory
critical infrastructure software.

You are also right about the stock bubble burst, and it will probably be
triggered by the MS stock crash. Only it won't happen till the mid next
year. You are probably predicting it too early. You may be basing it on
something else triggering it, thats why it is that early. Is that thing
the release of Windows2000. You know that has a real chance of taking MS
stock along with it. I just think that MS won't release it till mid next
year. Then too they probably won't release it, the stock will go because
public simply got fedup waiting for it.

-anand