Thoughts on the past of the PC and the future of Linux

Thoughts on the past of the PC and the future of Linux

Post by Ray Chaso » Wed, 15 Aug 2001 14:15:17



25 years ago, I could leaf through my dad's _Byte_ and _Popular
Electronics_ magazines and actually learn something.  _Byte_ was about
all the cool (in a budding geek's eye) things one could do even with
the limited hardware of the time.  _PE_ was about electronics, not
necessarily computers but anything built from silicon.

18 years ago or so, there was the Commodore 64.  How's this for an
open system?  For about $15 or so (I forget exactly) one could get
the "Programmer's Reference Guide".  This spiral-bound tome -- not
really a tome, it was about 5 inches by 8 inches and 1 inch thick --
had all the inner workings:  explanation of assembly language,
listing of the instruction set, all the "Kernal" calls, list of the
I/O ports and what they did, and even timing diagrams for the
interfaces and a complete schematic diagram of the computer.

Ten years ago, I could walk into Software Etc. and find books on
just about any programming topic imaginable, at least where PCs
were concerned.  One could find a good selection of software, too:
everything from MS-DOS to Windows to word processors to development
tools.  As it happens, I bought my first Windows programming book at
a Software Etc.

_Sic Transit Gloria Mundi_.  _Byte_ morphed into just one more bland
Ziff-Davis rag plugging the same products as all the others, and then
disappeared as a dead-tree publication.  _Popular Electronics_ has
gone through three or four incarnations, and disappeared for a while;
it's back again, but it's just the kit of the week now.  The C64 has
given way to the PC, a much more powerful machine of course, but
details of how the darn thing works are disappearing behind NDAs;
witness NVidia, Winprinters, and (ugh!) Winmodems.  Software Etc.
sells games and little else, just like Babbage's (which is the same
company) and Electronics Boutique.

Much of this has happened because business chases the lowest common
denominator.  Just *try* finding a book on writing a Windows device
driver among all the zillions of copies of _Office 2000 for Dummies_
and the like.  When the largest market was the geek, one could get
geek-grade software and literature.  Now the largest market is the
sort whose understanding goes little deeper than click-on-the-pretty-
icon, and geek-grade stuff is hard to find in the Windows world.

For a geek, much of the draw of Linux is that this dumbing down
hasn't happened yet.  One can still get under the hood and understand
what's going on.  One can still get in-depth literature; check the
shelves of a good bookstore.  Want to write a device driver?  You can
still learn how without signing away your firstborn.  There are books,
there are HOWTOs, and if that's not enough, there's the kernel source.
Linux still has the qualities that got many a geek into computing
in the first place.

This is the strength of Linux, and the dilemma that it faces.  Linux
needs market share to remain a viable platform -- not a monopoly,
just enough to assure decent hardware support and to force You Know
Who to play fair with interfaces and file formats.  But Linux will
not get that market share from geeks alone.  Linux must be accessible
to the masses, or it risks becoming a lab curiosity as You Know Who
embraces and extends everything else.  Linux must also be accessible
to the geek, as a geek defines accessible, if it is to get any further
development.

I do not think this problem is insoluble.  A solution may already lie
in the diversity of distributions.  Mandrake caters to the newbie,
Red Hat focuses on servers, SuSE ships everything including the
kitchen sink.  And your geek will go for Debian or Slackware.

I end with a call for sanity among the Penguinistas.  Every time some
nutcase yells at a newbie who's having problems, that's points for
You Know Who and its apologists and astroturfers.  Better to answer
the usual complaints.  When someone says it's too hard, point them
at your favorite newbie distro.  When someone says X "sucks", point
them at KDE.  When someone says they need to use Office formats, point
them at StarOffice and the various other open-source word processors.
Yes, there are gaps.  No, Linux isn't perfect.  Yes, we've got a good
thing going.  But we're not going to get that market share by being
elitist.

--
 --------------===============<[ Ray Chason ]>===============--------------
         PGP public key at http://www.smart.net/~rchason/pubkey.asc
                            Delenda est Windoze

 
 
 

Thoughts on the past of the PC and the future of Linux

Post by TomF » Fri, 17 Aug 2001 01:01:14



> No,?Linux?isn't?perfect.??Yes,?we've?got?a?good
> thing going.??But?we're?not?going?to?get?that?market?share?by?being
> elitist.

I enjoyed your post, Ray.

I think some of your points towards the end are very valid, but I also
think we need to get over this ridiculous self-righteous elitism that
suggests that creating a GUI of any kind is anathema to the entire
philosophy of Unix.

Too *ing bad.

First of all, WIMP interfaces are not all bad. Yes, they're wimpy. But they
can also be created in such a way that they are a pleasure for even a geek
to use. Witness Enlightenment, KDE, GNOME and Opera.

I believe it is possible to build GUI applications that are excellent for
the newbie, but flexible enough and accessible enough for the geek.

And I believe it is possible, with all the different distros out there, to
have Linux customized for anyone from the point-of-sale clerk, to the
accountant, the programmer or the nuclear physicist.

I also firmly believe Linux (and open source's) strength is in SERVICE, not
products. I believe the product-oriented state of the software market is a
fluke, a mistake and misguided. I don't believe software is a product. It's
a service. We, the geeks, need to be vigilant about SERVICING our
CUSTOMERS. Our customers are ALL PEOPLE who want to use a computer
reliably, easily, and powerfully.

--

Read about the carnage:
http://www.*edcode.com/

 
 
 

Thoughts on the past of the PC and the future of Linux

Post by mjcr » Tue, 21 Aug 2001 09:27:25




> I think some of your points towards the end are very valid, but I also
> think we need to get over this ridiculous self-righteous elitism that
> suggests that creating a GUI of any kind is anathema to the entire
> philosophy of Unix.

How can that be so, when unix had "GUI's" available long before Dos, Mac,
or Lisa?  It is not have GUI's that is anathema for unix, it is the call
for full GUI intergration into the kernel that is anathema.

--
I run Linux, no *y RedHat, Debian, Slackware, or Corel, just Linux.
Linux accepted my new hardware without any effort on my part.
Windows took one look at my new hardware and committed suicide.

 
 
 

Thoughts on the past of the PC and the future of Linux

Post by Ian Pege » Thu, 23 Aug 2001 05:46:54






>> I think some of your points towards the end are very valid, but I also
>> think we need to get over this ridiculous self-righteous elitism that
>> suggests that creating a GUI of any kind is anathema to the entire
>> philosophy of Unix.

> How can that be so, when unix had "GUI's" available long before Dos,
> Mac, or Lisa?  It is not have GUI's that is anathema for unix, it is the
> call for full GUI intergration into the kernel that is anathema.

Before I lost may way in the wilderness that is windows, my first
experience with a proper computer was with a UNIX machine the Torch XXX.
This ran UNIX system V and had a proper colour GUI. I think this was back in
1987, when apples were monochrome and windows 2 was a disappointment.

Ian

 
 
 

1. Thoughts on the future of the PC and Linux

In the past, DOS allowed application programs direct access to the
PC's hardware.  So, they used it.  Eventually, many programs began to
depend on direct access to the hardware.  Therefore hardware vendors
became reluctant to fix bugs or improve their hardware, since it would
break many programs that depended on those bugs.
  But now, we have Linux, where only one program (the kernel) has direct
hardware access (well, two if you count X) and we have full source code
to it.  We could now fix some old legacy bugs in the PC, and improve the
hardware, with only simple software rewrites.
  For example, we could eliminate those darned cascaded 8 channel PICs
with a single 16 channel PIC.  Or even expand to maybe 32 IRQs.  We
could add standards for the IRQs.  Like say, the first ISA slot always
gets IRQ4.  Or, the sound card always gets IRQ7.
  Lots of legacy functions in the BIOS could be deleted, and we could
fix the problems with booting above the 1024th cylinder.
  But, do you think this would be feasible?  Could we get any major
manufacturers to develop "Linux friendly" hardware?  Would there be a
market for it?  If you had a choice of buying an old style motherboard,
vs. a new faster, simpler (but possibly more expensive) "Linux Friendly"
motherboard, would you do it?
  I'd like to hear anyone's comments on this.

--
/* Daniel */

WWW: http://users.gurulink.com/drk

"Fear is only afraid of the absence of itself."  - Mediocrates

2. Delivery problems with your mail

3. Linux's future support for ATA/IDE development thoughts...

4. hack server??

5. Joystick troubles: "Not Connected" - thought this was a thing of the past...

6. Oops with 2.4.18

7. ??? UNIX tick time of future or past date/time

8. charset oriented GUI

9. The future is what just went whizzing past you.

10. NYC LOCAL: Wednesday 12 June 2002 NYSA: David Stenglein on Clustering, Past, Present, and Future

11. Keeping your system updated past future versions

12. Future DJ Technology with a Blast from the Past!

13. Assigning UID's past 65535, future troubles?