Re: [tlug] Stand Up for OpenOffice!!

["Stephen J. Turnbull" <stephen@example.com>]

Charles Muller writes:

 > So I don't mean to bash OO,

Is anybody bashing OOo?  All I did was point out that it's pretty
useless for grass-roots advocacy in my environment, and question by
implication how quickly it can propagate *today*.  It would seem my
standards are too high (I consider XEmacs's startup time of 15-20
seconds absurd---on modern machines it *should* be 5 seconds, but we
have about a dozen people reporting startup for OOo of 10-40 seconds
which they consider excellent to acceptable).

 > but I just can't see when it will ever truly become a realistic
 > option on a broad scale.

I'm nowhere near that pessimistic; my point is that realistically
*right now* it's a non-option on a broad scale, at least if my
experience is representative of its probable behavior in my
environment.  Leaving that aside, what are the prospects for the
future?

(1) Governments, especially local/regional governments in the U.S. and
Europe, know that an open document format is feasible, and that Open
Document Format is a feasible candidate.  They want this badly.
Others, especially those in so-called emerging markets (is Brazil
Gentoo Country? ;-), want cheaper, monopoly-free alternatives.
Equally badly.  [oops, that probably deserved a C&C warning.  and so
it goes.]

There will be supply to address this demand, and ODF means that small
and medium size consultants will often be able to supply.  Everybody
(except Microsoft) likes this.  This is going to drive a *big* market.
(Sorry, since my belief is that this market is going to be served by
small to medium size consultants based on commodity open source, I
have no idea which stock to buy to capitalize on it.)

(2) The Open Office technology is *almost* there.  My example of the
Ubuntu live CD taking 40 minutes to start OOo is an extreme;
unfortunately it is a strategically important extreme for getting the
diffusion process started.  A number of things could happen to make
that demo strategy feasible, some of which are dead certain (Moore's
Law of doubling chip density, for example, and the vulgar version
which says that clock speeds double every 18 months; cheaper bigger
memory) and others that seem pretty likely (OOo performance
optimizations).

(3) Related to (1), everybody wants formats that not only are portable
across vendors, but also across application domains.  Many of the
modern modular XML formats can be embedded in each other.  For
example, obviously we want to be able to embed SVG graphics in an HTML
document, but I believe that it's also possible to embed XHTML in an
SVG image!  Microsoft is going to have a hard time keeping up across
the board.  Of course it will have plenty of resources to convert OLE
objects from its traditional format to an XML-based representation
(presumably that's what OOXML is all about), but is it going to be
nimble enough to deal with thousands of "little object standards" all
of which are mutually embeddable via XML syntax and DOM semantics?

I think it's likely that somebody will get keystroke macros "very
right" (ie, a quantum improvement over what you can do with Wurd or
even Emacs) in the relevant time frame.  This is an area that's been
begging for attention for over a decade (see point 1 of Jamie's XEmacs
Wishlist http://www.jwz.org/doc/xemacs-wishlist.html).  We're starting
to see commercial entities *finally* do something with it (Apple has
some graphical stuff for assembling shell commands, although I think
the glue is AppleScript rather than something sane).  My guess is that
OSS will do a better job, though.

(4) I worry about documentation, as you do.  I suspect Tim O'Reilly
and friends will fix that, though.  Cf. "scads of SME consultants
needing docs." :-)

Note that the critical mass of brainwashed^Wtrained engineers has
already been achieved.  So my guess is that it will surely take more
than a year, but less than five, for all the prerequisites for
explosive growth of some open source office suite to get sorted out.
And my best guess is about two years.  Are there candidates other than
OOo?

(5) Note that your dictionary app (and Shannon's PIM database---dBase
II really was a revelation, wasn't it?) is pretty specialized.  It's
*important*, but it's not a great obstacle to general diffusion of
OOo.  Such apps are very widespread, of course, but they're becoming
a miniscule proportion of the total.  Cf. Shannon's own point about
having already converted everything else.