Re: Will Monolithic Apps Dominate?

[Robert Hettinga <rah@shipwright.com>]

At 1:33 pm -0400 on 7/18/97, Tim May wrote:

<an excellent description of the industrial strength software bloat which
now afflicts the browser market>

"The Geodesic Network, OpendDoc, and CyberDog"
http://www.shipwright.com/rants/rant_03.html is the first rant I wrote on
<ducking> geodesic </d> software. It ended up in a much(!) shorter form as
a full-page opinion piece in InfoWorld two years ago this October.
Actually, it's about what happened to me at MacWorld almost exactly two
years ago.

Anyway, OpenDoc is now dead, probably because it wasn't geodesic enough, I
think. After all, it called itself a compound document archichitecture, and
people have figured out by now that software won't be about documents at
all.

OpenDoc, as implemented, layered too much onto an increasingly bloated
MacOS. It was awfully slow and required many twidgits and tweaks to make it
all go, and there wasn't enough stuff there when you were finished to want
to use it all. It became the bloatware it was trying to replace.

There will be sucessors to OpenDoc, but I also agree with Dave W(h)iner
that Java is probably more brand name than anything else. Like Chauncey
Gardner, Java is what we project onto it.

But, sooner or later, probably when we figure out how to make bits of code
handle money in somekind of micromoney ecology, with bits of cash as
software and processor 'food', we'll have geodesic software. Software which
scales and upgrades itself to the processor it's using, and which is
infinitely configurable and extremely efficient in its use of computation
resources. Software which gets smaller in order to solve smaller specific
problems instead of getting larger in never-ending cycles of feature-creep.
Software which auctions its services to the highest bidder, and which when
outcompeted by a rival, dies, just like any form of life.

As far as hardware goes, the world will not be populated with the kind of
top-down network-as-mainframe network computer of Larry Ellison's wet
dreams. Microprocessor prices will continue to collapse at least for the
next 10 years, and computers will trend toward ubiquity, moving into
practically every artifact of civilization, and all those computers will
talk to each other on a ubiquitous global internet.

The software those computers use will not be hard-wired, it will be
flexible and upgradable. It will be 'out of control'. It will be geodesic,
like the network itself.

Cheers,
Bob Hettinga







-----------------
Robert Hettinga (rah@shipwright.com), Philodox
e$, 44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'
The e$ Home Page: http://www.shipwright.com/