> Bob, I appreciate your very long and thoughtful response. I don't agree with all of
> it, but I appreciate someone making me thing on this thread for a change! One of
> the few I have printed out...
You're welcome, and thanks. If I start being critical because you don't agree with me, please let me know :-)
> Maybe I've missed something with respect to Linux because I've been surrounded by
> workstations and big iron UNIX for a while. Your article reminds me of the days
> when I played with publically available software for radio transmission of digital
> packet data. It was good fun. Never made money from it, but like you say, if I didn't
> like something I could fix the source. I think the "make money" component is a big
> issue however, both for investors and business like.
It is a big deal. Everyone has to eat and put a roof over
their head. The standard model for turning a profit with
open source is by selling services to help people use
it. Not everyone can hack the code themselves, but there
should be people available to do that under contract. There
is currently a cottage industry in this kind of support,
and it is just now growing and starting to go mainstream.
This model is, in a way, the logical extension and perhaps
culmination of the software price wars that have gone on
over the past few years. When Borland started attacking
Microsoft on price, Microsoft started dumping their stuff
and taking a loss on the initial sale. This is, of course,
when software vendors' tech support got so unbelievably
bad; the only way to get decent support was to pay for it.
The old model of charging for years of support as part of
the package price got broken, and the entire game became
market share. Now software sells only for a token price,
and then largely as a way to give the retaliers something
to mark up. With software distribution over the Internet,
even that starts to seem silly. So why not just give it
away?
You'll often see staunch open-source advocates be
offended at *any* software selling for *any* price,
and it is sometimes difficult to believe that they are
getting so worked up about it (and it is pretty easy to
conclude that they don't have a mortgage payment). Some
of this is religion and/or culture, but some of it is
just labeling. If you recall the initial skirmishes over
offering gasoline customers a "discount for cash", you
may recall that the problem was fixed when they started
charging a "premium for using a credit card". This,
in real terms, doesn't make a whole lot of difference,
but it was OK where the fomer wasn't.
If the software vendors simply let anyone copy their
software and the price tags on the boxes were billed as
being a charge for the CD-ROM media, documentation and
a short-term support contract, virtually no one in the
open source community would have any problem with it. The
industry could even go back to charging $500 for things
like Quattro Pro, as long as the the $500 got you the kind
of support you got in the late 1980's.
As much as Microsoft and others go on and on about software
piracy and how much it costs them, I really have a lot of
trouble believing that they don't, deep down, welcome it to
a limited degree, at least outside of places like Southeast
Asia, where abuse is rampant. This is because they know
that the easier it is to pirate their software, the greater
will be their market share. Market share creates more
market share, and support contracts to go along with it.
Open source/free software is largely just a way of admitting
this up front.
> If by kids you mean older secondary and post secondary level, I think you have a
> point, in that the cheapness of the platform makes it accessible (how I got my start
> in this business long ago actually, was accessibility to a computer) -- but I think it
> doesn't meet the vocational criteria of being a direction that is know to generate
> jobs yet. Today its very hard to find a good VB or C++ or Java programmer,
> available, at a decent rate... (we use the former two by 10 to 1 proportion right
> now).
Yes, that's what I meant. If you find a skilled Unix
sysadmin that wants to move to the east coast, let me
know :-)
Are you talking about generating jobs or bodies to be
hired? I can tell you that when a young person applies for
a job where I work, I take notice when they tell me they've
been hacking on Linux at home for three years. This is the
kind of person we need, someone with the motivation, the
attitude and the experience to be able to make things work.
A college degree in CS isn't worth half as much, because
from what I've seen a good share of what they teach in
college has nothing to do with the job of making computers
stay up and running.
> I see first hand in our document management practice, and specifically in the legal
> industry, how poorly Corel has acted with Word Perfect. Actually it has been going
> on since Novell acquired them. Large law firms all over the place (not all, but a lot)
> moved to MS Word over the past few years because of WP stability problems. WP
> seems to be one of the buggiest Windows applications around and even recently it
> is still a big problem.
You are probably right. We also still use Word Perfect as
our production word processor, and it is awful. I wish we
could just go back to 4.2 or something. (Personally, I
mostly use vi and LaTeX, and it just works). Corel also has a
terrible reputation for most of their product line. There
are still a lot of people who would like to try running
WP on their Linux boxen, however, and if that wounded
elephant can dance on this platform, it could mark a
turn-around. But it could well be too late, as you suggest.
> I don't think we can blame that on Microsoft (is there a kill WP code branch in the
> OS? am I being too pollyanna-ish to not believe that?). I can point to one firm that
> has stayed on WP and has been battling problems with WP for half a year. Corel is
> being pretty helpful in that they put people on the problem, but still, half a year?
There was an old saying about Microsoft and DOS: "The
OS isn't done until Lotus won't run". This is part and
parcel to the whole break-up-microsoft arguement, that
MS's applications are the only ones that work well (in a
manner of speaking) because they design the OS toward that
end. I too have intense doubts as to the liklihood that
they really did this deliberately. But still, it is easy to
doubt that Corel developers can find out everything
that MS developers can.
> In the rest of your article you suggest that writing to other platforms is an answer
> since Microsoft has dominated the x86 personal computer market.
>
> I don't agree that writing to other platforms will be more useful in competition.
> That divides up the development community more. Writing to commodity platforms
> is a sure way to success for most developers. So Intel is a commodity. Perhaps if the
> write once, run everywhere goal becomes a reality there can be more than one
> commodity platform. At least for "personal" applications.
I'm not sure that I agree with what you seem to think
that I said (or maybe what I *did* say, I don't know...),
either. :-) I think, rather, that writing *portable*
software -- software that can be made to run on just about
anything -- is useful in competition. I think that coming
at Microsoft head-on on their home turf is a mistake.
The one good thing that has come out of all the diversity
in the Unix marketplace is that people have figured out
very effective ways to write platform-independant code.
GNU autoconf is an amazing tool. Things like Perl,
Python and Tk are at the cutting edge of this. The Python
programming language is available for DOS, Windows NT
and just about every known version of Unix. I think that
the fact that Corel Computer can take something like the
StrongARM chip from hand-toggling a bootstrap loader to
a commercially viable product in just a few months is a
stunning demonstration of the cross-platform abilities of
Linux and all the free software that goes with it.
One thing to remember is that Linux was developed by people
who didn't have much money. My personal image of a Linux
developer is a 16 year-old kid finding an old ISA card
in his neighbor's trash. He doesn't know what it does,
but by God he's going to make a Linux driver for it so
he can find out. Then, chatting on the net he finds 27
other people who also have these things, and they all work
together to make the baddest driver for that card there
ever was for any OS. Then he tosses the thing back in the
trash because he found another card and needs the ISA slot.
Linux is insidious because it will run on just about
anything. The only OS that has device support that
comes close is Windows 95, and even that won't run on
half the things that Linux will (like Sun Workstations,
for example). In a way, Linux then turns the class of
all computers into its "commodity platform". It is
also insidious becuase it is free, and no one is going
to try to shove it down your throat. If you have some
old piece of junk in the back and you want to see if you
can do something with it, just try loading Linux have a
go at it. If it works, then great, if it doesn't, there's
nothing lost but your time. Linux traditionally comes into
organizations through the back door with great stealth,
to run on the old PCs and workstations that no one ever
uses anymore. The power of something that turns junk into
valuable assets cannot be ignored.
> Should Microsoft's OS and Applications divisions be split to permit fairness in
> competition? Perhaps that would help.
I don't know the answer but when I look at what has
happened since AT&T was split up, I'd say yes. |
