Subject: Re: Open Source for Windows
From: "Stephen J. Turnbull" <>
Date: Mon, 18 Dec 2000 16:24:42 +0900

>>>>> "kms" == Karsten M Self <> writes:

    kms> A suggestion, however.  I'd borrow a page from Tim O'Reilly
    kms> and try to infect Legacy MS Windows users with the GNU/Linux
    kms> bug by actively advocating Windows suport for GNU/Linux
    kms> software.

At XEmacs, you'd stumble through the open door you just tried to shove

But once you pick yourself up off the floor, I suspect you'd be
disappointed.  Sure, you can port these projects to Windows, but
they're still Unix programs.  Even with an "Install Wizard" that
doesn't give them the choice to not install, you're going to find that
"Legacy Users" of MS Windows are highly resistent to even the most
virulent[sic] GNU/Linux strains.  And there's lots of stuff (eg, I18N)
that requires a lot of reimplementation to make compatible with the
"Windows Way."  Real support for Windows is hard work, with uncertain
payoff in the far distant future (ie, at least Q1 2002 ;-).

Maybe the next generation of KDE/GNOME will be close enough that it's
just a matter of translating APIs, but I wouldn't bet on it.  Even
there the attitude is just not "Designed for Microsoft Windows[tm]."

Instead, we're targeting people who say "I work in Windows, but I live
in Linux" or *BSD or whatever.  People who see ports of Unix tools to
the environment that they have to work in as productivity-enhancing
precisely because they _don't_ work like Windows; just _on_ Windows.
(But even this minimalist orientation still incites flame wars because
Windows advocates see Windows as the Right Stuff, and advocate pushing
further in that direction than the "old-timers" can stomach.)

Nor is it always obvious how to translate to the "Windows Way".  Eg,
many programs now are inspired by Windows equivalents, and use the Alt
keysym as the "natural" second modifier.  But older Unix programs use
Meta.  In order to reconcile these usages, some unpleasant decisions
have to be made for programs, such as Emacsen, that are able to map
and use multiple modifier keys.

Still, I think targeting Unix users transplanted to Windows is the
high-profit way to go, especially with highly extensible software like
Emacsen.  The Windows users will do a lot of the work for you, and
hopefully contribute it back, in the process of making themselves and
their coworkers comfortable.  But it will be a slow trend.

    kms> Interestingly, given concerns some have had that GNU/Linux
    kms> might be falling into the same "a better Windows than
    kms> Windows" trap of OS/2, it's arguable that Microsoft has
    kms> better support for GNU/Linux software than the other way
    kms> around.

Well, of course---just like everybody in the proprietary world, we are
now finding that we have to provide Microsoft compatibility at our own
expense.  Isn't that just what you are advocating?  (Stripped of the
intentionally provocative phrasing, of course.)  Whereas, again just
like proprietary vendors, we find this is a mug's game because
Microsoft will change the "standards" on us just as we think we have
them implemented.

I think this is the right way to go if we want to increase share on
the desktop.  But we should recognize who's in the driver's seat, and
that by itself it is a self-limiting approach.

University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
_________________  _________________  _________________  _________________
What are those straight lines for?  "XEmacs rules."