On Wed, Feb 16, 2005 at 04:28:21AM -0500, Norbert Omar Cartagena wrote:
> > Paul M Foster wrote:
>
> I was wondering how long it would take for my replies to get you talking
> ;-)
Hey, I had to take the bait! ;-}
>
> >On Tue, Feb 15, 2005 at 03:18:06PM -0500, Norbert Omar Cartagena wrote:
> >
> ><snip>
> >
> >
> >
> >>Yes, but you guys are talking almost specifically about problems *right
> >>now* that have been with Linux and open source software for a while,
> >>problems which are being addressed and fixed. What I'm talking about is
> >>the long term cycle and problems that may be (and likely are) with us
> >>now, but which we don't perceive as problems -- and might not until its
> >>too late. Bill Gates once pointed out (and I believe this was pointed
> >>out in the story that originated this conversation) that no one sits
> >>atop the world of technology forever. He's right. What I'm specifically
> >>talking about is this:
> >>
> >>In the long term, Unix's fall came because of (1) deregulation (when
> >>AT&T closed the source),
> >>
> >>
> >
> >Deregulation didn't cause this, though.
> >
> >
> It didn't cause it, but it was the first significant event, since it led
> to the splintering and the higher costs.
I'd dispute this, but I don't know the timeline. It could be. I just
didn't think deregulation happened, followed by the closing of Unix's
source by AT&T as a result.
>
> >Also Unix companies had a superior attitude, and didn't see the need to
> >put Unix in computer desktops. The AT&T PC was about the only one to put
> >Unix on a desktop box. Where was the software to run on it?
> >
> >
> Sorry. During my statements I kept a complete separation between the
> server software and the desktop software. AT&T's desktop Unix and Xenix
> (I believe) were the only desktop *nix's at the time, and treated the
> desktop as a "second class citizen."
>
IIRC, Xenix was the SCO product. Microsoft's name was on it at one time
as well, and I don't know if they licensed it from SCO, or sold it to
SCO.
No difference, though. Your statement is correct regardless.
> >Apple wasn't really a wunderkind. They had a superior technological
> >platform (68xx vs x86). And they had a killer app that captured one
> >specific market.
> >
> Uhmm... one?
>
PageMaker, as I recall. It started Apple on the road to dominance in the
graphics arena, where they still dominate. The Apple-dominated education
market came later, and has fallen off considerably. (It's interesting to
note that any time you see a computer in the movies or on TV, it's
either a Dell or Apple.) For the PC, the "killer app" was a spreadsheet
program whose name I don't recall (not 123, though, I don't think;
Multicalc?).
> >But they insisted on maintaining a closed architecture.
> >
> >
> Bingo.
>
> >That killed them, at a time when 1) IBM had been the go-to technology
> >company, and 2) IBM unwisely left the architecture open. IBM didn't
> >really think the PC would amount to much. When they figured it out, they
> >came out with microchannel and unsuccessfully tried to license it to
> >everyone.
> >
> >
> *heh* A wonderful little serendipity, eh?
>
Heh. And as I recall, anyone who signed up to manufacture microchannel
machines had to also pay licensing for the original bus architecture
(ISA?). That really killed it.
> >Microsoft cut a great deal with IBM, and so were on nearly every IBM PC
> >that shipped. When people made compatibles, compatibility dictated they
> >use MS-DOS.
> >
> 6.0, 6.1, 6.2, 6.21, 6.22...
>
And 1.0, 2.0, 3.1, 3.11, 5.0; all of 'em. Everyone bandwagoned the IBM
PC, because they were the "go to" company. That meant PC/MS-DOS.
> >Gates was always a master of marketing, never a true
> >innovator. They stole or copied much of the technology they use today.
> >
> >Linux and the Open Source movement suffer from not being paid to do what
> >they do, and no central authority to dictate what gets done.
> >
> Hmm... I'm not sure that the "central authority" part is all bad.
No, it has definite advantages. If Linus were in charge of hundreds of
programmers, you'd have fewer guys going off and doing the fifteenth
Linux PIM (scratching their own itch), and more straightforward progress
on the kernel and a single set of very polished apps for Linux.
(Particularly since Linus is a much better technologist than marketer.)
Don't get me wrong. I'm not advocating for a central authority. There
would be advantages to it. But there would be definite disadvantages.
The Linux software scene is one of the most vibrant and active software
scenes in the world. And you get a choice, instead of taking what
Microsoft feeds you.
It's a trade-off. Most people in Open Source prefer our side of the
trade-off. Newbies tend to complain about it, though. ("One-Click Linux
isn't one-click!")
> Doesn't this kind of get into the "cathedral/bazaar" debate? After all,
> you could argue that decentralization leads to higher response time for
> a larger number of needs. A bit like "narrowcasting" instead of
> "broadcasting."
>
Absolutely. ESR lays out the pluses and minuses in his original essay.
> >Microsoft
> >has $30 billion in _cash_ to pay programmers, and a head lock on
> >hardware and software vendors.
> >
> Who will only support Linux when they see a market demand. However, the
> market will only demand it when the hardware is capable of interfacing
> cleanly with the OS. Catch 22.
>
That's where you get companies assisting with kernel development. They
get the demand from customers, and so they pour assistance into
assisting us to build a better kernel. It's not as much as we'd like,
and the kernel developers are pretty persnickety about trusting the
suits. But every little bit helps.
> >Easy. For us, we have to wait until some
> >programmer in Slovenia gets around to doing the drivers for X hardware.
> >And we have to wait until someone figures out how the make the pretty
> >dialog boxes seamlessly move across the screen without leaving artifacts
> >and killing the X server.
> >
> >
> And this right here might be the long term killer. Will anyone dedicate
> themselves to creating LINUX drivers, provided that doing so will likely
> force them to reveal code and even their proprietary hardware specs?
>
It's been a tough battle. Witness nvidia et al. That's where guys like
ESR, RMS and Perens come in. Arm-twisting and negotiation. And then
there are the guys in Lower Slobodonia quietly reverse engineering
hardware. Every uptick in market share helps. Every Open Source opinion
leader whispering in the ear of folks like HP, IBM, et al helps. Again,
it's not as fast as we'd like.
I think the hardware guys are overly paranoid. I'm guessing that video
cards are relatively simple to reverse-engineer, if you have the proper
facilities. In other words, I don't think their secrets are really all
that secret. If it weren't for patents and the DMCA, I suspect there
would be many knock-offs of successful graphics cards, sound cards, etc.
(That's not an indictment of patents; for hardware, they're fine.)
<snip>
> >Linux continues to gather market share because we continue to improve
> >it. Remember installers back in 1995?
> >
> Even in 1999...
>
> >And remember, Linux is still
> >dwarfed on the desktop, partially because hardware vendors don't
> >typically promote and pre-install it. But with the tech folks, it's
> >doing great in the server arena. Which gives it momentum on the desktop.
> >
> >Microsoft's got $30B in the bank. They're going to be around a loooong
> >time.
> >
> Yes, much like AT&T at one point in time. Oh, wait... SBC, right...
>
Notice the re-consolidation of the telecommunications industry. Somehow,
governments have never figured out a way to properly administer
utility-class entities. So their answer is often something like
deregulation. Unfortunately, this often causes as many problems as it
solves. The baby bells still own the lines, but now they have to lease
them to CLECs. For the last few years, I had our business phone service
through a CLEC called Allegiance Telecom. When it came time to move my
business, I requested that Allegiance move the lines. They pretended
they didn't have the "facilities" to do so (whatever that means). I
ended up having to go back to Verizon. Why? Allegiance has no service
trucks. So when you want your lines physically moved, they have to pay
Verizon techs to do it. And for a small account like mine, it's too
expensive. But they'll never tell you that. So I got Verizon, a
contract, and more expensive phone bills (but better service).
Because of its market share, Microsoft is virtually a utility. But the
government repeatedly fumbled its opportunity to do something
significant about them.
> >Their deficiencies and shortcomings are becoming apparent. Since
> >the anti-trust trial, people have begun to speak out and look for
> >alternatives. But Microsoft's got patents, money, influence. Their stock
> >still looks good, and so they're not all _that_ worried about us yet.
> >They're trying to diversify to cover their bets.
> >
> >
> I guess I'm still looking at the anti-trust case statements where MS
> mentioned Sony as their biggest competitor.
Yeah, and the GPL is communist. Microsoft says a lot of things for PR
value and to get their butts out of the fire.
<snip>
> I don't think it's a matter of anyone ultimately winning. The players
> will likely all change before all's said and done.
>
Agreed. Back in the Model-T days, who could have predicted anti-lock
brakes, automatic transmissions and power steering?
Paul
-----------------------------------------------------------------------
This list is provided as an unmoderated internet service by Networked
Knowledge Systems (NKS). Views and opinions expressed in messages
posted are those of the author and do not necessarily reflect the
official policy or position of NKS or any of its employees.
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:12:47 EDT