[Desktop_architects] Portland: How to fix this situation

[Brooks, Phil]

I love it - make the Linux desktop more attractive through the 
dilligent efforts of Windows users!

-----Original Message-----
From: desktop_architects-bounces at lists.osdl.org
[mailto:desktop_architects-bounces at lists.osdl.org] On Behalf Of Bryce
Harrington
Sent: Thursday, December 08, 2005 2:56 PM
To: Otto Wyss
Cc: desktop_architects at lists.osdl.org
Subject: Re: [Desktop_architects] Portland: How to fix this situation

On Thu, Dec 08, 2005 at 09:31:49PM +0100, Otto Wyss wrote:
> This discussion about "The Linux Desktop Integration Interface" is 
> really amazing to read, it looks so obvious to me that the top
inhibitor 
> for a Linux desktop adoption is the application shortage 
> (http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005.pdf). No user will 
> 
> Changing Linux-Only application to cross-platform has another
important 
> effect.
> I've heard the argument that users won't switch to Linux when
everything 
> is also available on Windows. Come on, if Linux doesn't have other 
> arguments and can't stand this, then it's not worth to discuss any
further.
> 
> As OpenOffice and Mozilla-Firefox proves cross-platform is possible
and 
> these two projects are highly successfully. Only cross-platform
projects 

Inkscape also is cross-platform, and I agree 100% with you here.  My
experience with making Inkscape cross-platform bears all of this out.

When the Inkscape project started, it was Linux-only.  The core
developers were all Linux users, but we agreed that having Windows
support was important for the overall success of the project.  Very few
of the core developers were interested in working on it, but I think
because users really liked the program and felt it had good usability
compared with the commercial alternatives, they were motivated to get it
working on Windows.

Thankfully, there was a port of Gtk to Windows, and this was the key
enabler for undertaking this.  Within a matter of months, a couple
developers and a whole bunch of users had solidified the port.  At that
point our download numbers went through the roof, and before long our
measurements showed more Windows-based Inkscape users than Linux ones.

In fact, the importance of Gtk here cannot be understated.  We've had to
drop support for Inkscape on Win 9x simply because Gtk hasn't had the
manpower to maintain the Win9x port.  (Since the Inkscapers who are
stuck on Win9x are often there due to lack of technical aptitude, we've
not been able to find any recruits to help with the Win9x ports.)

The OS X port was even more intriguing because it was done almost
entirely by motivated users.  It was interesting to watch because it was
done by a sequence of users each of whom spent a week or so, got
frustrated not being able to get it working, and gave up, but each time
whatever progress they made was incorporated back in, and after a number
of iterations, it actually worked.

Today, it's gratifying to note that a 3rd party company is providing
shrinkwrapped versions of Inkscape (and Gimp) for Windows and OSX.
(IIRC, Gimp goes for $49.95, and Inkscape around $170).


Anyway, having Inkscape available on all three platforms is itself an
enabler for a lot of people.  We have several users who can run Linux at
work or in their business, but be confident that they can share
documents with their print shop and co-workers.


I also agree that having all the app gaps filled is important, there are
a number of Linux-based artists who had previously been forced to resort
to booting Windows or OSX to use Illustrator or other drawing programs.
They had high expectations about the quality of the software and were
slow to convert, but are today happily 100% Linux.

> There are numerous ways how to achieve cross-platform better and
worse. 

Agreed.  As described above, perhaps the best strategy is to just make
the app so well loved by users that they port it themselves. ;-)


The survey mentions 4 apps:  Photoshop, PageMaker, AutoCAD, and
Quicken.  We already have some open source applications in each of those
spaces.  Gimp, Scribus, QCad, and Gnucash come to mind.

So, perhaps a strategy could be to a) get each of these ported and
working well on Windows (Gimp's already there), b) focus on making the
usability of these apps better than their commercial alternatives, and
c) help users get involved with adding more features and maintaining
ports, if the app is short on developer manpower.

I'm sure there's no end of apps that people would like to see on Linux,
but if those four are at the top of the list, they seem like logical
candidates to focus on.

Bryce