On Jan 26, 2004, at 8:32 PM, Chris Huston wrote:
> Poll question:
>
> If you have installed fink which is true for you:
> [ ] Fink has caused SERIOUS problems building or using standard OS X
> software (difficult or impossible to work without uninstalling Fink)
> [ ] Fink has caused MINOR problems building or using standard OS X
> software (easy work around with command line switch or shell variable)
> [ x] Fink has enabled use of LOTS of useful open source software
> without interfering with standard OS X software
> [ ] Fink has enabled use of SOME open source software without
> interfering with standard OS X software
> [ ] Fink HASN'T HELPED OR HURT use of standard OS software.
> [ ] What the heck is Fink?
>
> Where do you prefer Fink install libs:
> [ ] /usr/local
> [x ] /sw
> [ ] Use the OS X "frameworks" mechanism
> [ ] Who is this Fink and what's a lib?
>
> How do you use your Fink?
> [ ] Aqua apps only
> [ ] Aqua and X11 apps
> [ ] Aqua and command line tools
> [ x] Aqua, X11, and command line tools*
> [ ] Now I'm really cheesed off - what the heck is Fink?
* Apart from Aquaterm, installed with (eg) gnuplot, I am unaware of any
aqua programs in fink.
Also, you forgot to ask some other pertinent questions, like these:
Fraction of unix scientific software on your computer installed via
fink (omitting dependencies like eg fftw):
[ x ] 100% - 75%
[ ] 75 %- 50%
[ ] 50% - 25 %
[ ] 25 %- 0.00001%
[ ] 0%
Fraction of fink installed unix scientific software on your computer
not easily installed without fink (due to major proting patches, etc):
[ x ] 100% - 75%
[ ] 75 % - 50%
[ ] 50% - 25 %
[ ] 25 % - 0.00001%
[ ] 0%
Estimated scientific productivity if fink did not exist (relative to
100% with fink installed, pick closest value) due to loss of time,
compiling frustrations, or unavailability of software (as above):
[ ] 100% - 75%
[ ] 75 % - 50%
[ x] 50% - 25 %
[ ] 25 % - 0.00001%
[ ] 0%
Fraction of Apple computers that I would have been purchased for my lab
for scientific computation if fink did not exist:
[ ] 100%
[ ] 75 %
[ ] 50%
[ ] 25 %
[ x] 0%
Now let's take a look at the questions:
>
> If you have installed fink which is true for you:
> [ ] Fink has caused SERIOUS problems ...
I have yet to have fink cause me any problems for building software,
but without fink I would find it essentially impossible
to do any crystallography on OS X. In the unlikely event that fink
causes a problem, you can simply start a new shell session
by not sourcing the /sw/bin/init.(c)sh file. If you don't like where
/sw/bin shows up in the path, redefine PATH. I have yet to have any
problems. Unless you explicitly tell your software during configure of
the existence of /sw, I fail to see how any interference could occur at
all.
> building or using standard OS X software (difficult or impossible to
> work without uninstalling Fink)
(see above) Also , you don't have to uninstall it. At most you simply
don't have to source the init file.
> Where do you prefer Fink install libs:
You can tell fink to install it anywhere you please, and if you ever
put together a fink package, you would
realize that fink absolutely insists that the user be able to define
whatever equivalent to /sw they want.
> [ ] /usr/local
> [x ] /sw
>
> How do you use your Fink?
> [ ] Aqua apps only
> [ ] Aqua and X11 apps
> [ ] Aqua and command line tools
> [ x] Aqua, X11, and command line tools
So why is "Aqua" present in every response? This, and especially the
first answer, betrays a fundamental ignorance of what fink software is
available. Apart from aquaterm, what aqua apps are available via fink?
>
> I'll tally and post the results.
> - Chris
Why? It's not like fink is going to dry up and die pending the
hoped-for response to a rather skewed set of questions.
The main problem with fink from what I can see is that sometimes the
latest versions of software aren't made available in a timely manner.
But that is only because it relies on the volunteer efforts of a lot of
busy people.
When my primary computational machines were SGIs, I found it almost
impossible to install a lot of software that is made easily installable
via fink, and found using computers very unpleasant as a consequence.
With fink (and OS X) that has changed completely. Having an automated
package installation system has made it possible for me to focus my
computational efforts on science rather that operating system and
software maintenance.
I'm sure there are many people who enjoy spending hours trying to port
unix software to a new operating system. For me this is basically a
waste of time, especially if someone else on the planet has already
done it. (When I do have to do this, fink is an easy way for me to
share the fruits of my labors with the user community.) Fink saved me
a grate deal of time, blood sweat and tears.
There are a handful of fink developers that focus on the scientific
programs for fink, and I think they have done a tremendous
job and have devoted long hours to making OS X a viable platform for
scientific computing. A lot of the fink packages that "duplicate"
Apple versions are in fact much more useful. Apple's default python
for example is lobotomized; many of the modules one expects with a
standard installation are missing. Where is Tkinter? Left out because
Apple doesn't need or use TK? Numeric? Scientific? Sure, I can
install these by hand, but it is nice not to have to hunt down and
install dependencies manually. Fink installs dependencies required by
all its packages, thus eliminating the horrible headaches I would get
installing stuff on my SGI the hard way.
Fink has probably helped to sell more Apple computers to scientists in
my building than all of the advertising, all of the stories about
supercomputers at Virginia Tech, and all of the visits by Apple sales
reps combined. I think for a lot of people, without fink, access to
Apple's unix underpinnings would have remained a partially-functional
but essentially useless novelty. I am sorry to see it get bashed in
this way. I hope Apple realizes this.
William G. Scott
Associate Professor
Department of Chemistry and Biochemistry
and The Center for the Molecular Biology of RNA
Sinsheimer Laboratories
University of California at Santa Cruz
Santa Cruz, California 95064
USA
_______________________________________________
scitech mailing list | email@hidden
Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/scitech
Do not post admin requests to the list. They will be ignored.
