[luau] Re: Crossover 1.1.0 Review

[Warren Togami]

This is the usual naysayer statement, and in theory I agree with you.  The
idealism of using a tool merely for moral purposes works for us hackers, but
unfortunately it doesn't cut it for normal end-users and the software
development houses that currently make software for those users.

Realistically, we cannot hope to _quickly_ convince both users and software
developers to use the Linux platform if we advocate for only pure native
ports.  Wine is important for three key reasons.

1) User Migration Path
Users like to feel safe.  Familiarity is good.  They don't like learning new
things.  If you don't agree with those statements, then you haven't dealt
with end users very much.  If these users see that their favorite or
necessary programs like Microsoft Office, Lotus Notes, Starcraft,
Counterstrike, and Windows Media Player work in Linux, they feel less afraid
of giving Linux a try.  They may then eventually realize that they no longer
need the expense of their old proprietary software once they see the free
and open alternatives. (We must continue to improve our free alternatives
for this to work.)
Also, many small proprietary software houses will never port their software
to be native Linux or even WINELIB native.  If Wine can handle these
applications, more power to the Linux platform.  (e.g. This ability is
crucial for a certain Library Catalogue client software used by my
workplace.)

2) Fast and Cheap WINELIB Native Porting
WINELIB facilitates extremely quick porting of Windows software to run
natively on Linux.  In order to convince software developers to make Linux
software, we must first show them economic reasons to do so.  This is the
usual chicken & egg problem where they don't see the demand for Linux
desktop software so they wont port their software (thus end-users wont use
the Linux desktop).
Wine and Winelib makes it very easy for these people to port their existing
software to native Linux executables.  Costs and complexity of maintaing the
cross-platform ports are also reduced because the delta between the port
source trees is reduced dramatically.  One example of this being extremely
important would be cross-platform MMORG games.  MMORG clients must be
updated quickly and cheaply in order for that company to be economically
viable.  A complete native port of such a client would be incredibly
difficult to keep in sync with the Windows client because of the large
source delta.  However, if their Linux port compiles easily from the same
tree, it is fairly easy.  I mention MMORG specifically because these are
some incredibly addictive games - possibily killer apps that could do a lot
toward promoting Linux desktop usage.  Gavriel State, the CEO of Transgaming
has mentioned his efforts in trying to convince Electronic Arts to release a
Linux client of their upcoming The Sims MMORG.  I hope he succeeds.  The
same could be said for the upcoming Final Fantasy XI MMORG and the exising
Everquest.  Diablo II already runs in Linux.

3) Developer Migration Path
When these software development companies begin to make Winelib ports, they
may for the first time be exposured to the other non-Microsoft alternatives.
They may see value in using QT, SDL or Mono in the future in order to make
completely cross platform software that easily compiles and runs on Windows,
Linux, MacOS 9, MacOS X, and more.


Yes I do acknowledge there is the risk of diluting the efforts of Linux
developers because working proprietary (especially free beer) software
already exists, but I assert that the many additional Linux users will
offset this loss.  Additionally, there will always be hackers interested in
open software and the intellectual challenge of coding it.  Open Source will
eventually win.

Now why don't we begin by winning the Windows desktop userbase?  I see no
other realistic migration path other than Wine.

If this upsets you, then fine, don't use it.  Linux has the luxury of
choice.  Windows doesn't.

----- Original Message -----
From: "Garrick Staples" <garrick at polop.usc.edu>
To: <usclug-chat at zig.usc.edu>
Sent: Saturday, March 09, 2002 9:24 AM
Subject: Re: [usclug-chat] Crossover 1.1.0 Review


> Call me a fanatic, but I think this type of software is an overall bad
> trend for Linux.  It's a message to software companies that Linux users
> don't need good native versions of software.  It shows that Linux users
> are willing to leap through hoops to use some applications.
>
> Linux needs OSS alternatives to proprietary software.
>