[Discuss] Essentially all free apps will soon be available
onWindows
Alan W. Irwin
irwin at beluga.phys.uvic.ca
Sun Feb 18 09:16:58 PST 2007
On 2007-02-17 22:30-0800 Noel Burton-Krahn wrote:
> This thread has gone off the rails too.
We all have the big delete key if we find some poster is consistently boring
on about windows too much. :-)
OTOH, it is nice to be on topic again so thanks for making that happen.
>
> Going back to the original post: I'm sure CMake is great. I'll definitely
> look into it when I have to do another build system. I can't see how a build
> system can help me port a Linux app to Windows though. Alan, how will CMake
> help me get strace or top or tcpdump running on Windows?
I tried to use the term free app rather than Linux app for partly this
reason. Obviously the small subset of free apps that are directly related to
the Linux kernel is an entirely different story and there is no point to
port them to windows. However, if they are complicated enough to build
with autotools, then I would advise switching to CMake simply because of
the complexity and lack of speed issues with autotools.
>
> As for getting open source apps on Windows, most are already there: Cygwin,
> Emacs, Perl, Python, Inkscape, Gimp, etc. You know, if they weren't already
> ported to Windows, I'd run in Linux (in a VM, thank you) much more often.
I agree many free software projects try to have windows ports. Some are
quite successful at it, but others are not. A good example is PLplot
itself. Prior to CMake we dealt with this issue by having an autotools
build system for everything but bare windows, and for that latter case we
had a home-brew build system that our one windows developer tried to keep
maintained. That home-brew windows build system was extraordinarily
limited in results and was difficult to maintain. Our windows port
was a second-class citizen, and our windows developer could get any help
to overcome that problem.
After our CMake revolution most of our language interfaces and device
drivers are available for windows and the port is now a first-class citizen.
It's also been very good for our group dynamics since our veteran bare
windows developer now feels much less isolated; every improvement we make to
the build system helps him and vice versa. Also as a direct result of our
change to CMake with a uniform build system we have recruited another
developer who has helped out tremendously with the Cygwin, MinGW, and bare
windows tweaks that were necessary for our CMake build system. He has also
contributed some great code to PLplot for a wxwidgets device driver and
binding so that _Linux_ users have greatly benefited from his windows
development work.
In sum, CMake allows convenient high-quality windows ports of free apps, and
just as importantly also helps Linux and Windows developers to work easily
together on such apps.
> [...] I
> don't understand how porting apps could make anyone switch from one OS to
> another.
High-quality ports of many free apps (e.g., KDE) are not yet available on
windows, but that is about to change because of CMake. Once users start to
use those apps I believe it should make the switch to GNU/Linux much easier.
Although I am confident of this result in the long run, it is admittedly
difficult to argue logically about that long-term forecast one way or the
other. The fundamental reason to port free apps to windows is to increase
the software freedom available on that OS. If that results in switches to
GNU/Linux in the long term, that is just gravy.
> [...] Now it's easier to get Open Source on
> Windows, so there's a bigger install base to run open source apps on.
> Remember, open source is bigger than just Linux.
And software freedom is bigger than open-source. Open-source is a useful
marketing spin thought up by ESR to help introduce businessmen to software
freedom. However, I am convinced that software freedom is actually the
fundamental force that is driving this software revolution so that's why I
emphasize it (although I acknowledge the benefits of the open-source aspects
of software freedom as well).
Alan
__________________________
Alan W. Irwin
Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).
Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the
Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________
Linux-powered Science
__________________________
More information about the Discuss
mailing list