Linux-Advocacy Digest #711, Volume #25 Mon, 20 Mar 00 14:13:06 EST
Contents:
Re: Windows 2000: nothing worse ("Christopher Smith")
Re: Windows 2000: nothing worse (Paul Jakma)
Re: Windows 2000: nothing worse (Matt Gaia)
SUUPER !! kostenlose 80 MB Speicherplatz + POP3 ("Peter Müller")
Re: Windows 2000: nothing worse (Paul Jakma)
Re: Windows 2000: nothing worse (Donovan Rebbechi)
Re: Windows 2000 - the latest from work.... (Donovan Rebbechi)
Re: Why not Darwin AND Linux rather than Darwin OR Linux? (was Re: Darwin or Linux
(Donovan Rebbechi)
Penquins Forever! Was (Re: A pox on the penguin?) ("2 + 2")
Linux on the Desktop...TODAY! (David Steinberg)
Re: Gnome/Gnu programmers Suck. -- Not a troll (Darren Winsper)
Re: Debian Potato release? (Darren Winsper)
Another Box Dominated by Linux! (Nico Coetzee)
Producing Quality Code (mr_organic)
What are the limitations of using Linux on your server (if there is one)?
([EMAIL PROTECTED])
why is it important that linux do not fragment (mark@spamfree)
----------------------------------------------------------------------------
From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Tue, 21 Mar 2000 02:11:37 +1000
"Matt Gaia" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> >wasn't there a leaked statement from MS themselves that said as much?
> >Please Drestin, tell us it wasn't so.
>
> Yes, it was so. A leak from inside MS stated that there was about 63,000+
> potential bugs in W2K.
"Issues", not bugs. That ranges from spelling errors in code *comments*
(phew, real showstopper there) to real bugs - and also includes errors in
the testing code used. That same memo counted the number of "real problems
potentially visible to the end user" (or words to that effect) at around
20,000, although there's no indication of what such "real problem" actually
is.
~20k bugs for 40 million LOC isn't too bad. Indeed, until anyone can
provide similar numbers for some other OSes, any claims made based on that
number are nothing but FUD. After all, the bug/patch lists for many
commercial Unices are pretty damn big.
> >a lie you say? Well, by golly, why hasn't /Microsoft/ Hotmail been moved
> >to WNT? Surely it would a matter of pride and principle for an OS vendor
> >to have such a high profile site run on their own OS, rather than that
> >of their competitors? (ie FreeBSD on x86, Sun Solaris on Sparc)
>
> <sarcasm> Maybe it's because of MS's willingness to give other operating
> systems a chance to survive </sarcasm> . . . before they attempt a hostile
> takeover.
You can provide a list of other vendors willing to give competing products
"a chance" ?
[chomp]
------------------------------
From: Paul Jakma <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Mon, 20 Mar 2000 16:17:03 +0000
Erik Funkenbusch wrote:
>
> They have full access, yes. What administrator has that root does not is
> the ability to remove permissions. Yes, the administrator can reclaim those
> permissions but you have to deliberately do so.
>
you can set things up on a unix machine so that root has to undo things
before they can modify something. So what's your point?
root on Unix = all-powerful
administrator on NT = all-powerful
what's your point eric?
> I've been writing Win32 software for over 5 years.
>
> Were you not aware that you can remove the SE_RESTORE_NAME right from the
> administrator group?
>
of course he knows that! (god damm it he probably knows more about NT
file permissions/domains/smb than MS). And he even said as much - ie you
can restrict administrator on NT, but administrator can undo those
restrictions at will.
> No, a user running as root need only type rm /* -rf to wipe out a system, no
> matter what the file system priveleges are set to.
So? they are root, by definition they have full access to every bit and
byte on the system. That is root.
And administrator can do the exact same thing.
Only difference is that the Unix tools don't molly coddle you as much as
the tools on NT. But if you wanted rm to assume you're a fool then use
rm -i.
> You can't do something
> similar under NT if your priveledges are set correctly. Yes, the
> administrator can take ownership and change the priveledges then perform the
> operation, but that's very different from accidentally typing a space
> between the * and the dot.
>
> If you call that deliberate, i'd call that stupid.
it's a function of the tool... so what. If you want to write a MS-rm
that molly coddles you, go ahead.
As for capabilities, hey lot's of Unixen have this too, including linux
2.2+. Even better once you have removed a capability from root on linux
you can not restore it without rebooting. ie root can't get around it.
Which by your argument is a lot better than NT.
-paul jakma.
------------------------------
From: Matt Gaia <[EMAIL PROTECTED]>
Subject: Re: Windows 2000: nothing worse
Date: Mon, 20 Mar 2000 11:21:54 -0500
>You can provide a list of other vendors willing to give competing products
>"a chance" ?
Ok, I got one ironic one then. Microsoft. Seeing as they have yet to
port Hotmail onto a Windows box, I guess they really are giving Unices a
chance, huh? And I wasn't talking about "giving a chance" in the terms of
working with another vendor. I'm talking about setting up an illegal
monopoly in an area. Other vendors (Real, Netscape, AOL, etc...) would
fit into the first group, but only one that I can think of, M$, belong to
the latter.
------------------------------
From: "Peter Müller" <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.lisp,comp.lang.pascal.borland,comp.lang.pascal.delphi.misc,comp.lang.perl.misc,comp.lang.perl.modules,comp.lang.python,comp.lang.tcl,comp.mail.eudora.ms-windows,comp.mail.sendmail,comp.music.midi,comp.object
Subject: SUUPER !! kostenlose 80 MB Speicherplatz + POP3
Date: Mon, 20 Mar 2000 17:20:17 +0100
Hab gerade ein bisschen rumgesurft und bin auf die Seiten von
www.galaxy24.de gestossen.
Die sind echt gut aufgemacht, haben unter anderem eine Community, in der man
80 MB Speicherplatz und eine POP3 eMail-Adresse kostenlos bekommt.
Lohnt sich reinzuschauen !!
Gruss
Peter
------------------------------
From: Paul Jakma <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Mon, 20 Mar 2000 16:26:33 +0000
Christopher Smith wrote:
> > <sarcasm> Maybe it's because of MS's willingness to give other operating
> > systems a chance to survive </sarcasm> . . . before they attempt a hostile
> > takeover.
>
> You can provide a list of other vendors willing to give competing products
> "a chance" ?
>
what a potent piece of MS-advocacy!! I am astounded by the forthright
thrust and subtle twists that you so cleverly have injected into this
one line of rebuttal.
you want a list of other vendors that give competing products "a
chance"? do you have such a small forehead that the sarcasm just went
right over it? ????
> [chomp]
aye.. chomping on a doughnut, one hand on keyboard, other scratching
your arse i'm sure.
come on, you can do better...
-paul jakma
[hint: try and explain why MS has not converted one of the most
high-profile sites on the net over to their flagship enterprise OS]
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: 20 Mar 2000 16:33:34 GMT
On Sun, 19 Mar 2000 04:31:44 -0600, Erik Funkenbusch wrote:
>Jeremy Allison <[EMAIL PROTECTED]> wrote in message
>news:8b1uer$gqi$[EMAIL PROTECTED]...
>They have full access, yes. What administrator has that root does not is
>the ability to remove permissions. Yes, the administrator can reclaim those
Wrong.
>Who said anything about circumventing an administrator? Of course an
>administrator can do whatever they want to do. The difference is that you
I don't see how that's a "difference".
>No, a user running as root need only type rm /* -rf to wipe out a system, no
>matter what the file system priveleges are set to. You can't do something
Again, wrong. man chattr. man mount. Or even 'man chmod'
--
Donovan
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Subject: Re: Windows 2000 - the latest from work....
Date: 20 Mar 2000 16:43:41 GMT
On Sun, 19 Mar 2000 04:15:46 -0600, Erik Funkenbusch wrote:
>So, teaching a user to use TeX is much easier than teaching them to use
>Word, right?
Depends on whether or not you want them to use word properly. Sure, it's
easier to produce *something* in word, even if doublespacing is implemented
by hitting <enter> twice, etc.
Of course, if the document contains any math, they'll *never* get it to
look decent with word. ( argualy,it's impossible anyway. Hell, it doesn't
even kern or handle ligatures )
Contrary to what some would believe, TeX is not brain surgery. I was up
to speed with a week, and my documents use a lot of math ( since that's
what I do ) and hence I had more to learn than most users.
--
Donovan
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Crossposted-To: comp.sys.next.advocacy,comp.sys.mac.advocacy
Subject: Re: Why not Darwin AND Linux rather than Darwin OR Linux? (was Re: Darwin or
Linux
Date: 20 Mar 2000 16:45:59 GMT
On Sun, 19 Mar 2000 10:54:02 +0100, Matthias Warkus wrote:
>It was the Sun, 19 Mar 2000 00:58:18 GMT...
>...and Sal Denaro <[EMAIL PROTECTED]> wrote:
>[capitalism yada yada]
>> I wonder if that sound I hear is Ms. Rand doing back flips in her grave...
>
>References to Ayn Rand.
>Instant-O-Loss-O-Credibility.
You could do worse ( like the guy who cited Howard Stern in the affirmative
action thread )
--
Donovan
------------------------------
From: "2 + 2" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Penquins Forever! Was (Re: A pox on the penguin?)
Date: Mon, 20 Mar 2000 11:56:14 -0500
Thread = overblown hysteria.
2 + 2
Drestin Black wrote in message ...
>Ready or not, Linux viruses are coming, and no one is inoculated.
>
>http://www.securityfocus.com/commentary/2
>
>How to get infected using Linux...
>by ralmeida
>
>calvin:~$ wget http://somesite/happy99.tar.gz
>calvin:~$ tar zxf happy9.tar.gz
>calvin:~$ cd happy99
>calvin:~$ ./configure
>calvin:~$ make
>calvin:~$ su
>calvin:~$ make install
>calvin:~$ exit
>calvin:~$ happy99
>You must be root to run this program
>calvin:~$ su
>calvin:~$ happy99
>(ops!)
>
>
>
>Re:How to get infected using Linux...
>by QuantumG ([EMAIL PROTECTED])
>(User Info) http://www.virusexchange.org/vlad/
>
>hehe.. more like:
>
>calvin:~$ wget http://somesite/pointlessgadget.tgz
>calvin:~$ tar -xzvf pointlessgadget.tgz
>calvin:~$ cd pointlessgadget
>calvin:~$ ./configure
>calvin:~$ make
>calvin:~$ ./pointlessgadget
>
>"that was boring.. I'm gunna go shoot stuff"
>
>calvin:~$ su
>calvin:~$ /usr/leet/leetgame
>
>pointlessgadget was infected with a virus.. when you ran the virus it
>infected every one of your running processes, including your shell. You
su'd
>to root and it peaked at your psuedoterminal to snarf the root password. It
>then su'd to root and infected every running process on the machine. You
>then ran leetgame and the virus infected it. Next you'll probably run 'ls'
>and then it's all over.
>
>Fiction? You can do it using ptrace.
>You can read about it here(some linux viruses are at the bottom of this
>link - feel free to experiment):
>
>http://www.big.net.au/~silvio/
>
>
>
------------------------------
From: [EMAIL PROTECTED] (David Steinberg)
Subject: Linux on the Desktop...TODAY!
Date: 20 Mar 2000 17:23:35 GMT
Wow! These are exciting times to be using Linux!
Corel has just announced that they are shipping WordPerfect Office Suite
for Linux tommorrow:
http://www.newsalert.com/bin/clipstry?StoryId=ConwWudCbmdm1tG
And then, there's Mozilla. The next-generation, open source browser from
Netscape that will soon be blowing the pants off of IE,
cross-platform. It's going beta in just 25 days:
http://biz.yahoo.com/prnews/000320/ca_netscap_1.html
Okay, so maybe Linux still isn't suited for everyone's home computer.
But, with these kinds of applications coming out, there is no reason that
it couldn't be the desktop of choice for many non-technical users in
businesses. Compared to Windows, it's cheaper, more stable, just as easy
to use, and now it's got the apps, too! It's also LESS trouble to
install and support, for a trained IT team that knows what it's doing.
And then, there's XFree 4.0 and KDE 2 coming very soon! It just keep s
getting better and beter for us!
--
David Steinberg -o) Boycott Amazon.com! Fight
Computer Engineering Undergrad, UBC / \ the "1-Click Order" patent:
[EMAIL PROTECTED] _\_v http://www.nowebpatents.org
------------------------------
From: [EMAIL PROTECTED] (Darren Winsper)
Subject: Re: Gnome/Gnu programmers Suck. -- Not a troll
Date: 21 Mar 2000 02:07:08 GMT
On Mon, 20 Mar 2000 01:31:42 GMT, [EMAIL PROTECTED],net
<[EMAIL PROTECTED],net> wrote:
> Unfortunately your description of sloppy, scripted application
> prototype so accurately describes the typical Linux program.
Oh look, the homophobe makes yet more assertions without proof.
> Linux users are willing to put up with crap like this. Others are not.
Oh bullshit. Just because you think you are some almighty authority
about what's good and what isn't, doesn't mean you are right.
Do us all a favour and fuck off.
--
Darren Winsper (El Capitano) - ICQ #8899775
Stellar Legacy project member - http://www.stellarlegacy.tsx.org
DVD boycotts. Are you doing your part?
------------------------------
From: [EMAIL PROTECTED] (Darren Winsper)
Subject: Re: Debian Potato release?
Date: 21 Mar 2000 02:07:11 GMT
On Fri, 17 Mar 2000 15:59:31 -0600, Tim Kelley <[EMAIL PROTECTED]> wrote:
> Actually it's "Woody" I think,
Woody is the new unstable branch, where everything goes (Well, more or
less). Potato is the old unstable branch which is now frozen for
release (i.e. bug fixes only).
--
Darren Winsper (El Capitano) - ICQ #8899775
Stellar Legacy project member - http://www.stellarlegacy.tsx.org
DVD boycotts. Are you doing your part?
------------------------------
Date: Mon, 20 Mar 2000 20:24:08 +0200
From: Nico Coetzee <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Another Box Dominated by Linux!
I just dumped my last NT box for Linux!
W H A T A R U S H ! !
NT crashed after the registry grew out os it's allotted 11MB! Can you
believe it... I thought the OS should at least warn before kicking the
bucket. Obviously that did not happen.
Luckily I had my Profile on a VFAT partition (I was dual booting with 95
previously) so I could backup everything and easily migrate to Linux.
Interesting facts:
How long did it take me to install NT complete with all my apps and do
all configurations? About two days (a weekend).
How long did a similar Linux install took? Under two hours!
I'm a bit scheptic, but M$'s W2K must really rock before I ever touch a
M$ box again (home environment).
Cheers all!
Nico
------------------------------
From: [EMAIL PROTECTED] (mr_organic)
Subject: Producing Quality Code
Date: 20 Mar 2000 17:27:18 GMT
Reply-To: [EMAIL PROTECTED]
THE PROBLEM
===========
As readers of my previous rants probably know by now, I have emarked
on something of a Jihad against sloppy programming. The target of my
ire at the moment is the seething mass of Windows "developers", most
of whom know next to nothing about good software design. However, my
previous posts are unfair in that they give the impression that
Windows programmers are stupid, lazy, and incompetent. This is unfair
and prejudicial, and I apologize.
However, I think many points I raised in my posts are valid. A recent
article in The Atlantic (find it at
http://www.theatlantic.com/unbound/digicult/dc2000-03-15.htm) detailed
some of the concerns cropping up about the quality of software in
general. It seems clear that software is dropping in quality overall,
not improving. And this is probably as true of Linux as of any other
mass-market operating system. By *why* is this happening?
It's easy to blame Microsoft (but accurate, since they espouse methods
and designs to contribute to bad code); they have the hightest profile
in the industry. More people come into contact with their software,
so more people are exposed to the bugs. The various Unixes have their
fair share of bugs, but they tend to escape wider notice because they
are not as widely deployed. However, this is changing as Linux
charges into the homes and offices of millions of people. Now, when
Linux has a significant bug, it bites a lot of people.
At base, the problem is a cultural one -- we are not training
programmers correctly, regardless of the platform they use. The
software industry today does not reward stable, carefully-crafted
programs; it rewards newness, whiz-bang features, and layers of
eye-candy. In short, *potential* is rewarded, not *execution*.
Programmers are mercilessly ridden to meet marketing-driven deadlines,
which means that they tend to focus on things like slick GUIs and less
on architecture and maintainability.
Unix tends to have less of a problem than Windows in this regard for a
couple of reasons: one, Unix programmers are generally more
technically adept than Windows programmers, and consequently produce
better code; two, the architecture of Unix tends to promote "correct"
code more than Windows does. Complexity leads to instability (in
computer systems as in so much else in life), so a good rule of thumb
is "simpler is better". However, modern operating systems are almost
unbelieveably complex. Windows 2000 is reputed to have between thirty
and sixty million lines of code. Linux has nowhere near that kind of
bloat, but it still weighs in at several million lines of code (if you
include things like the X Window system and essential system
utilities).
One place where Unix has a leg up on Windows is in terms of system
calls. Generally, the fewer system calls you expose, the more stable
the operating system will be (because simpler tends to be better).
The *BSD's and Linux expose a few hundred system calls; Windows 2000
exposes several *thousand*. It's not hard to see why this can lead to
trouble.
Compounding the problem is the proliferation of GUIs. It's no
accident that software quality has been declining in direct proportion
to the popularity of GUIs. Graphical environments are notoriously
hard to "get right" -- there is no commonly-agreed-upon basis for
their design or implementation. A graphical environment is by its
very nature less precise than a command-line interface -- when a user
sees an icon of a painter's palette, for example, he or she usually
assumes that this icon is in some way related to a graphics or drawing
program. While this works for most folks, other people in the world
would be baffled by such a thing -- say aborigines or Eskimos, who
have a totally different conception of artistic endeavor than do other
people. What would a meaningful icon be to them? Or is a GUI even
appropriate? The whole "desktop metaphor" makes a lot of assumptions
that aren't necessarily valid.
The problem is that programmers are trying to solve problems that are
not well-defined, using operating systems that themselves are overly
complex and ambiguous. This problem has been snowballing for years,
until today we have an enormous base of code that is neither stable,
robust, nor easily fixable. And this problem will continue to fester
until something dramatic is done to arrest it.
The solution must begin with us -- software engineers. Notice that I
use the word "engineer" rather than "programmer" or "developer". This
is deliberate. We must recognize that our craft requires the skills
of an engineer as well as an artist. We must realize that careful
design and rigorous attention to detail is as important (or more so!),
than working on "sexy" problems. But at the same time we need to take
pride in ourselves as artisans, and produce code that not only
*works*, but is as elegant, maintainable, and clean as we can make it.
This will not be an easy thing to do; most of us are employed by
companies that only pay lip-service to software quality. They will
sacrifice stability in a second if it means gaining market advantage.
Marketers and salespeople will continue to define software in terms of
feature-lists and gee-whiz graphics. What can we do, short of
revolting entirely? We have to make a living, don't we?
My suggestion is this: realize your power. Software engineers are the
most sought-after asset in the world today. Entire companies rise or
fall depending on the quality of their software engineers. If each of
us, individually, makes it a point to produce good, quality code, then
the effect will propagate quickly. But to do it ourselves is not
enough; we must insist on it from others as well. And this is the
difficult part. We must stop accepting lousy software and hold the
developers responsible. If a company produces software with bugs,
they should be held as accountable as any company that produces a
defective good. (Which is why UCITA is such an evil thing, but that's
a topic for another thread.)
The "Open Source" revolution of recent years has shown people that
peer-review is an excellent way to get quality software. To be sure,
much Open Source software is buggy; but relentless peer-review
produces software that is demonstrably less buggy than most commercial
alternatives. And don't be fooled by "performance tests" -- a faster
product is not necessarily better, if the slower one is of higher
quality. All pure speed means is that a buggy product will crash
faster than ever before.
The odd thing about all this is that you'd think companies like Apple
and Microsoft would *want* this focus on quality code regardless of
marketing or sales pressure. Better code means lower supoort costs
later (and better customer relations!). Still, vendors seem to be
stuck in a "get the code out the door at any cost" mentality. This is
most prevalent in the Windows space, where competition is very hot.
However, this problem is also emerging in the Internet space; it is
evident in the badly-written Javascript, VBScript, PHP, or CGI scripts
that infest the web. E-commerce sites assure us that our information
is private, then mutter red-faced when crackers crack their sites and
steal credit-card information.
In the coming century, the consequences of bad software design are
going to be more impactful than ever before. We bet our very lives on
software -- it controls airplanes (and air traffic control centers);
it controls nuclear reactors and utility companies; it controls water
delivery and waste-disposal; it controls all finance and securities
movement. In fact there is no major part of our life that software
does not have a critical part in. Technologies like Jini and
Bluetooth promise to make software an even-more-integral part of our
lives, so much that it will become essentially invisible and
ubiquitous.
But this scenario shows how calamitous bad software can be -- it's
only an annoyance when we lose a corporate memo or a PowerPoint
presentation; it means the loss of life if a medial system goes down
or two airplanes cannot determine each other's position in the skies.
THE ENGINEER'S PART
===================
The upshot is that we need to take our craft *seriously*. Hackers we
may be, and proud of it, but we need to embrace the *true* hackish
nature, not just the faux-geek trappings. A true hacker knows that
producing good code is often boring, repetitive, and unexciting; it
means hours and hours of groveling over code to find a misplaced comma
or missing semicolon; it means being open-minded enough to know when
your approach is wrong and allow others to help fix your code. It
means valuing correctness and stability over all other things, period.
One of the central tenets of the Hippocratic Oath that doctors must
take is, "First, do no harm." I think this is a superb rule for
software engineers to follow as well.
THE CORPORATE PART
==================
But we as engineers can only effect so much change; the corporations
which employ us have got to do their part. Unfortunately, the
situation today is not promising. Most corporations do not answer to
their customers at all, but rather their shareholders; and today's
shareholder is interested in only one thing: short-term monetary
gain. The quality of the product in question is only important
insofar as it furthers the goal of making money. Software is only as
good as it needs to be to make money (which unfortunately isn't very
good).
There are no easy answers for companies. The only way to produce
stable, robust software is to *take time* and do it right. But
companies do not want to spend the time -- lost time is lost money and
lost opportunity. Better to produce a shoddy hack quickly than a
robust solution later. Let the users be your beta-testers and make
them pay for the privilege. Patent everything and threaten to sue
anyone who treads upon your domain. Using this tactic, they ensure
that lousy software *stays* lousy because no one can legally fix it.
We have arrived at a time when a company must produce bad software to
prosper; writing good software takes too much time and effort.
Shareholder value is not enhanced by producing robust code; therefore
they will not produce robust code. Money, in this as in so much else
in the modern world, is king.
WHERE DOES THIS LEAVE US?
=========================
Were it not for projects like Linux and the *BSDs, the situation would
be grim indeed. The Windows platform is for all intents and purposes
a lost cause -- Microsoft has no intention of opening the source, and
no real incentive to fix the numerous problems themselves. Apple is
in a similar situation, but they benefit somewhat from the fact that
their new OS is based on FreeBSD, and can benefit from review of that
code. (Whether they will give back to the community in the same
measure they have taken from it remains to be seen.) The only other
viable commerical OS -- BeOS -- is just as closed as Windows is and
presents the same essential problems.
We, as software engineers, must commit to making the Open Source
operating systems -- whether Linux, FreeBSD, NetBSD, OpenBSD, or other
-- as robust and well-designed as we possibly can. We must not let
money rule our release schedules or feature-sets. We must not get
drawn off into feature-wars with commercial packages. In the best of
all possible worlds, there would be a Good Engineering Seal of
Approval and no piece of software would make it into a distribution
without it. The GESA (as I call it) would encompass areas like
stability (no buffer overruns!), security, and "correctness" of code.
In many ways we are the architects of the new century. The decisions
we make -- and do not make -- will profoundly impact the society of
which we are a part. We need to realize this and act accordingly.
Submitted for your consideration.
Regards,
mr_organic
------------------------------
From: [EMAIL PROTECTED]
Subject: What are the limitations of using Linux on your server (if there is one)?
Date: Mon, 20 Mar 2000 18:15:35 GMT
My company is planning on hosting roughly 200 web sites on a single
Linux box (I am unsure as to which flavor), using Apache server. The
server will have roughly between 500 megs ~ 1 gig of memory. These
sites will by dynamic and primarily database driven on a separate
server which will be using MYSQL as the back end and Perl to access the
data. Is this a feasible notion, can a single Linux box coupled with a
database server with the previous stats be capable of hosting and
handling approximately 200 dynamic web sites?
Thanks in advance
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: mark@spamfree
Subject: why is it important that linux do not fragment
Date: 20 Mar 2000 10:06:24 -0800
with so many distributions, how do we make sure
software written to one, will work on another?
Is linux headed to the same path that Unix in the 80's
went to, with so many incompatable versions?
We must have a Unified Linux. see this article. when when LSB
be done?
http://dailynews.yahoo.com/h/zd/20000320/tc/20000320722.html
some quotes:
The technical diagnosis: "Debian must have some
libraries in a different location than the others,"
he said, meaning that some functions added by Debian
to Linux cannot be retrieved in the same fashion as
those added by its rivals.
"Business will want to know if it's commercially viable over
time. There's very much a command and control element to it,"
"If I got a Linux system that was totally different, it would
cause problems," said Jeff Davis, senior systems programmer at
oil company Amerada Hess
One discipline preventing that is a 2-year-old agreement by
the major vendors to establish a Linux Standard Base, or a
description of the essential elements of the kernel and
supporting components that should be in any distributor's
Linux. Caldera's President Ransom Love has been an advocate
of such a move. Red Hat, SuSE and TurboLinux are also all
members of the committee. After two years of talk, no defining
document has yet emerged.
Proprietary additions to Linux keep creeping into the individual
distributions, differentiating a vendor from the crowd.
Now that they are publicly held, the managements at Caldera,
Red Hat and VA Linux must be responsible to a constituency -
profit-minded shareholders - that may find itself at odds with
the "let's share"-minded open source movement.
"When the power of money is at stake, it changes the way people act,"
said Hurwitz, who remains a skeptic that the open source code
movement will maintain the cohesion necessary to allay IT managers'
concerns.
------------------------------
** 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.advocacy) 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-Advocacy Digest
******************************
