[clue-tech] Mephis linux

[Roy J. Tellason]

On Wednesday 19 November 2008 05:24:45 pm Nate Duehr wrote:
> Problem is... many people have the talent and the ideas, but there are
> business reasons to "leave it all alone", mainly that companies make BIG
> money on patches and upgrades.  Writing software ONCE that works is not
> in ANYONE's business models anymore.  Maybe it once was, long ago... but
> it's not coming back.  Service contracts are big money.

Seems to me like the same holds true for hardware as well as software.

> The sad reality is... having to have a service contract means you KNOW the
> software won't work right.  Deep down, that's the only reason to buy one,
> isn't it?  The MAJORITY of "IT workers" rely on this fact that keeps them in
> their jobs (myself included), seemingly forever:  Software devs CAN'T --
> literally CAN'T -- write bug-free code for any reasonable price.

Or can,  but it's not considered economical.

(Snip)
> Computers become commoditized in the 90's and we still haven't gone
> through the phase of "Do we REALLY need a computer for this job, or is
> it actually more efficient, more secure, and smarter overall -- to use a
> piece of paper, a pen, and a filing cabinet for this particular task?"

Hehe.  :-)

> We haven't hit that point -- a um, for lack of a better word -- "luddite
> stage" yet, because we haven't had a complete meltdown of something very
> important to us all (say, the banking system) due to bad software...
> yet.  (We will, eventually.)

You might want to subscribe to RISKS Digest some time,  there are all sorts of 
interesting examples of things Not Working Right in there.

> Meanwhile the things we tout as "computers getting better" is just that
> we have faster hardware that can crash faster and reboot faster after
> the crash.

They're not getting better.  Sure there's faster hardware out there and bigger 
hard drives and all that,  but better?  No.  And software seems to expand to 
use up those resources as fast as it can.  Users are pushed along that path 
because of the available software needing newer hardware.

My first linux box was a K6-200 (which I still have!  I only stopped using 
that as a server because of HD issues),  and my first install was Slackware 
4.0 on a 1G or so drive and I installed *everything*,  so I could check it 
out.  This was back in 1999.  Now a complete install takes several CDs worth 
of data,  and takes longer,  and eats up one heck of a lot more HD space even 
though I am quite selective about the packages that get installed.  Better?  
Maybe a little,  but there's also a lot of eye candy and other junk in there 
I could just as easily do without

> Think about it this way, if software was getting THAT much better,
> wouldn't the Mainframe be completely dead by now?

Nope.

> Why isn't it?  Because it sits there and runs and runs and runs, and the
> software on it was developed under a much different mindset.  "Downtime not
> allowed". 

That's part of it,  but the requirements are different,  and the applications 
are different,  no downtime being only a small part of where they're 
different...

> That era is long gone, but the results of that work (quality work) are 
> still running key applications today.  We may have ported that code over
> to newer hardware, but it's still the code written in the era of
> "mistakes are costly", not the "aww, just release a patch, and blame it
> on the QA department".

You're talking about the difference between a slow and disciplined approach to 
programming on the one hand and the free-form anything-goes fiddle with it 
until it sorta works approach taken by many _individuals_ on the other.  
They're entirely different contexts.

> What I would HOPE Linux does... and I do think it can do it with
> leadership and vision, is become the OS that shoots for BETTER quality
> than the others.  Right now, it's shooting for quantity not quality.

Some folks are,  yeah,  but not everybody.  Some people like lots of eye candy 
on their desktops,  while others prefer more robust applications instead.  I 
don't think you should generalize like that.

> Now that seemingly "everyone" knows what Linux is, somehow, somewhere
> the quantity vs. quality pendulum MUST swing for Linux to survive
> long-term as anything more than a niche OS or a server-only OS.  The
> server application folks recognize this, but they also moved to running
> on ANY underlying OS long ago (Apache, for example - some of the highest
> quality code out there, but it'll run on anything... they don't care if
> you use Linux anymore, and never will again).
>
> Just some random thoughts... I think this is headed into the ditch again
> for "clue-tech", so we can go over to clue-talk, but it never leads
> anywhere.  The devs that have the talent to really pull this off, aren't
> hanging around here, AFAICT.  I'm certainly not one.

Nor am I,  but I can be picky about what software I choose to use.  :-)

> My dream/vision for the real growth (as in getting smarter) in computer
> use isn't all that realistic, I know.  It's an ideal.  Lots of people
> voted for ideals in the last election that aren't realistic either, so I
> guess I'm allowed my little fantasy too, aren't I?

Hey,  we all get to make our individual choices,  and unlike elections,  the 
ones we make for software stick.  :-)  Those choices can also come into play 
as recommendations,  too.

> Nothing for you to get upset or depressed about.  You like Linux on the
> desktop, I don't... so?

So what do you like on the desktop?

> I still use it on servers... and probably always will... until something
> better comes along.  (GRIN... darn those users, they're fickle, aren't they? 
> Especially the ones that have no emotional ties to their software like
> becoming part of an open-source cult or similar.)

Heh.  :-)



-- 
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space,  a critter that can
be killed but can't be tamed.  --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James 
M Dakin