20 Feb
2006
20 Feb
'06
7:27 p.m.
I was an enthusiastic user of free software since the mid-80s, and
during a phase where I was actively porting software between about 10
platforms, I became a convert to gcc (you think C is standard now? try
the early 90s. Ugh!). A zealot-like conversion to all things GPL and
LGPL followed.
To digress briefly:
One of the subtle points of my song "It GPLS me", (which I recorded in
1999 with the aid of non-free software and rms has dinged me on that),
was that I reused a song written in 1968 that was recorded in 69.
Copyright regarding songs was
changed in 1971. To this day I'm confused as to whether I'm
legitimately able to reuse the chords from the song I cribbed from
(under anything other than the parody clause).
Free software gets it's strength from copyright law. For all of us to be more free,
copyright law must be returned to something closer to what the founding fathers intended.
The current length of the copyright term is absurd. It should be
returned to 14 years, renewable once. Once the proprietary copyright
owners realize that their hands are tied by all this "free software"
with restrictive licenses, perhaps there will be some pressure to
return the length of this artificial monopoly to something reasonable
(IMHO It should be returned to 14 years, renewable once, and the right
of reversion, lost in 1920, restored). But I'm not holding my breath.
Back from digression. The core analogy that I make regarding free or
open source software is not the one "would you buy a car with the hood
nailed shut?", but "Would you buy a house without the plans provided
by the architect?".
It was very tough, politically, within an organization to use or (god
forbid) help create "free software" up until the late 90s. Now it's my
day job - although I find myself spending 95% of my time coping with
the timesump of the proprietary software.
Three years ago, or so, I tried to make a record using purely open
source and free software. It wasn't ready. That didn't matter so much, it
turned out the proprietary software I was trying to use wasn't ready either, despite
marketing promises of wonderfulness, the products were too buggy to use for my purposes.
(I was trying for a 96khz concept record with surround). Today, after steady visible
progress I could do it with ardour, linuxsampler, and rosegarden - (but don't have the
time or the money for soundproofing)
The free/open source engineering ethos leads in the long run to less buggy, more
featureful, well supported, useful products, written by engineers, for users.
The proprietary ethos leads to products that require license keys, tech support, marketing
staff, requirements documents, language lawyers, and inevitable legions of bugs unfixed
because of all that deadweight a commercial programmer must support.
I'm damn happy with the current state of the Linux RT kernel, and ardour, et all, are
making good progress. With the availablity this year, of *good* profiling tools for linux
like oprofile, performance is making a comeback - see gnome 2.14 and the gslice allocator
for one major example.
On the free vs open source side I believe that if RMS didn't exist, we'd have to
invent him. While "open source" is a weaker concept, it IS a gateway drug. Most
people I know that have got into open sourced products (and out of a buggy proprietary
one) would never go back. The rest, so far as I can tell, are masocists, that LIKE
spending time with clueless support people or deciphering glossy brochures.
I would like it a lot if microtransactions were easier to implement, or there were a
better model than dual licensing to keep programmers paid for doing what they love, but I
just love watching the code get better, and get better faster, than the proprietary
alternatives. The quality curves vis proprietary software have already been crossed, and
features are not far behind.
--
Mike Taht
PostCards From the Bleeding Edge
http://the-edge.blogspot.com