Eugene wrote:
> Emlyn wrote:
>
> > The cruddy "standard" will always win. While you are trying to hold your
>
> Not necessarily. It is much easier and quicker to release a shoddy
> standard
> providing a minimum functionality and get people to use it. See Napster.
>
> However, people will pick up and use a good standard just as
> indiscriminately. The question is rather: can you be early enough?
> Mainstream is a decade away before picking up AR. I'd say we're early
> enough.
>
> We're not gas molecules. You can shape the larger scale phenomena by
> knowing the rules, and being early.
>
> > (indubitably) superior standard together, in agreements with a zillion
> > variously competing interests, someone who doesn't give two hoots about
>
> Good standards are written by a single person. Maybe a team of two or
> three.
>
I agree with you. However, writing the standard isn't the issue; it's more
usually gaining the agreements of all major competing players to use it.
> > technological superiority will sneak in a minimally interoperable
> > alternative. The politics is so much easier, and it gives everyone a lot
>
> So you keep the politics out of it. There's not much politics going on
> within a single person's head, unless she's got MP syndrome.
>
That's debatable (politics in one person's head), but that's not what I
meant.
The politics are in getting various important groups to support your
standard. Standards don't just get adopted by accident; people choose to
adopt them. Usually a handful of groups (companies? committees? whatever)
are crucial to it's success. And often, these groups compete in some way.
Assuming some groups have a competitive agenda with respect to some
other(s), each will develop it's own unique position on standards - maybe
they make their own "standard" (ie: proprietary solution), maybe they adopt
the standard but subvert it, or maybe they are so weak that they willingly
sublimate themselves to it, hoping thereby to get a piece of a pie from
which they would otherwise be denied a slice.
> > more room to move. For instance, the big orgs will want to subvert the
> > standard and own it; a minimal, cruddy standard requires (proprietary)
>
> So you make 1.0 fit the bill, get people to use it in an avalanche
> process
> and brand it. Sign the source tarballs and the binaries. Patent it to
> you,
> only allowing noncorporate use. There are probably other mechanisms
>
Let's take the case of Gill. He runs a big company; lets call it Sicromoft,
for hypothetical purposes. I have a large marketshare of related markets,
and a new tech area pops up in which I am a natural player. Maybe it's VR
over the internet. We are placed in a future a few years from now, where
broadband is fairly widely adopted, and people are starting to drop support
for any connection speeds below say 128kbps.
Gill's decided it's time to muscle in on this area and own it. He knows that
a few of the other big boys are thinking the same way. He's missed the boat
before (although overall he's very good at this game). He's getting in
relatively early this time around. It's too late to develop his own product
in this area now (actually, it was tried a few years back, but fell on it's
bum; he wasn't really serious about it then anyway). So he's going to need
to coopt someone else's technology, and run with that.
The state of play... basically at this point, there are a few companies in
the arena, and a couple of arms of the open source movement. The leading
"standard" (A) is the free standard. It is by far the technologically
superior standard, but unfortunately is for non-corporate use only, and is
guarded by a pack of rabid open-source guard dogs. Some of the proprietary
solutions have some of the ground (the most successful we'll call B), and
there is still a really old, crappy "open" standard, a predecessor of the
current players, which has a little support (C). The people using this
technology at the moment, regardless of the specific form, are in general
geekdom and other technologically literate, but the masses are starting to
sniff around.
What does Gill do? Well, he can't touch (A), he's been locked out. Even if
he could, it probably wouldn't be the right path. No room to move. He'd only
choose it if he had been beaten to the punch by one or more of the other big
orgs, which he hasn't been. One of them will probably do it, however, poor
bastards. Trying to manage a relationship with the free software movement;
like getting into bed with a porcupine. Urrggh.
Gill could buy out (B), and push that. As a bonus, it is quite good
technically. The company that owns it is worth one day's lunch money to
Gill; price is not an issue. This might be a goer, but he has the problems
of selling a closed solution. A good fix would be to release it as an open
standard, minus a few important details, then bombard the market with his
proprietary, almost standard implementations. He controls the playing
field, he decides on what is in the standard and what is not, he can stay
ahead of the pack. That's worked quite well for him before!
Another choice would be to grab the obselete open standard, (C), and become
it's champion. Of course, it's got problems, and would require some
proprietary extensions. It would be a legitimate looking way of fighting off
(A), however, which is currently his biggest threat. Then he can pillory (B)
and all it's cohort for being evil closed standards, and (A) he can rubbish
as being a geeks' toy, with no moral weight greater than (C), and with no
industry support. If they maintain their no-corporations approach, this will
be ridiculously easy; he can probably get some of the other big guys behind
(C), on the understanding that this is to kill (A), and that everyone will
fork off their own "standards" in the near future, and fight over the ground
left by (A)'s demise. If (A), on the other hand, decides to allow corporate
involvement, then a company coopting them might be a threat, if it isn't cut
to pieces trying to work with free software people.
Whatever the choice, there'll be a massive fud campaign against A. He can
probably get a lot of industry support for the fud; maybe together they can
form an independent special interest "user group" or some such to do the
job. Also, Gill will probably play one of his trump cards; provide optional
"support" for A in his products, which is mostly broken, and performs
abysmally.
> > extensions to be workable. A solid, locked in, proper standard just
makes
> > them look like bastards for stuffing around with it.
> >
> > Good standards for interoperability only seem to be useful to the weaker
> > players. I think that's why they fail.
>
> I think the weaker players would dominate the place if they'd realize
> how
> much power concerted action wields. The problem is they're not only
> weak,
> they're stupid (uncaring).
Not stupid; self interested. In the classic sense, that often looks a lot
like stupidity; working out how to group together to maximise each
individual's self interest seems to be a weakness.
>
>Masses produce nothing, it's the individuals.
Yep. A few rich ones, with a bit of drive.
> Has everybody smart sold
> out?
>
By definition?
Emlyn
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:50:37 MDT