[linux-audio-user] OT: FUD or not? what is your opinion?

Ivica Ico Bukvic ico at vt.edu
Wed Feb 21 23:46:36 EST 2007


It appears that there is a usual rich variety of opinions on this matter,
which pretty much reflects the general state of the Linux scene (for better
or worse). What follows is an *incredibly* long rant. Please do not take
this against me--I simply feel very strongly about this matter and as such
FWIW I feel compelled to share my thoughts with you.

Having used many different WMs over the years and eventually settling for
KDE which I used exclusively until very recently, I now prefer Gnome simply
because I am tired of endless customizations. I simply want things to look
moderately appealing out-of-box without having to mess with them too much
(perhaps not having an ability to customize too much in the first place was
the needed catalyst to prevent me from wasting my precious time on endless
tweaking of my desktop experience ;-). Yes, Gnome does have its own share of
stupid limitations and on top of that in many places looks inconsistent
and/or incomplete, but for me (as subjective as this sounds) it does what I
want it to do today. Call me a sucker for eye-candy or simply someone who
prefers to have well-rounded user experience. But the way I see it is that I
can always open a console full-screen, which more often than not I do (all
right, not in full-screen, but definitely in a window ;-), yet if the rest
of the UI sucks, then advocating Rat Poison et al is usually seen (at least
from the outside) as a poor attempt at rationalizing current deficiencies of
the Linux desktop rather than a choice made on a level playing field. Add to
this the fact that we live in a predominantly oculocentric culture (at least
the Western culture, including its ubiquitous sphere of influence), and
you'll quickly realize that no matter how good the backend of a framework
is, most "users" (and that still includes a vast majority of artists) will
opt for the UI which looks prettier, even if that is at the expense of
usability. This means that WMs and Linux audio are inseparable whether you
like it or not.

So, what of it? FWIW, I think Linux audio will remain a research/enthusiast
platform no matter how cool tools end-up being produced by the LAD
community, as long as we'll have incompatibilities (Robin mentioned one
feature that both XP and OSX are ahead in, and that is a drag-n-drop
feature) and inconsistencies plaguing the desktop experience, and perhaps
more importantly fragmentation due to endless pursuit of new approaches to
existing solutions (i.e. how much energy has been already wasted on
redundant sound servers?). Add to that the current dubious state of GTK+,
whose potential demise, regardless whether you like this toolkit or not
and/or whether such demise is imminent, will bring about tremendous overhead
upon us as well (think how many different audio apps rely upon this
framework and if it becomes obsolete how much time will it take to port
something like Ardour to a different toolkit).

I might be simply getting older and thus more conservative. Back when I
started using Linux, I probably would've found Rat Poison the coolest
concept on the planet. Today, I am a mouse/UI junkie who finds Rat Poison et
al to be a noteworthy experiment, with my hope and attention turned almost
exclusively on AIGLX and Xgl. But with this newfound conservative outlook on
technology and supporting philosophy of GNU/Linux, I am beginning to believe
that this endless choice (anarchy?) of solutions and/or options inherent to
our platform/doctrine of choice, as wondrous as it is, is now beyond its
zenith. I see us now speeding along on a downhill path with several
generations of enthusiasts/programmers/hackers who have gotten burned out by
this endless pursuit of superfluous yet for the most part incomplete
solutions (whose incompleteness was often a byproduct of our resources,
namely manpower, being spread too thin). The irony of this situation is
that, just like in the cutthroat corporate market, in part due to resulting
saturation not every time the best technology has prevailed (think JACK and
desktop sound servers).

So, now we have incomplete Gnome desktop (at least IMHO), over-customizable
KDE desktop (again IMHO), bunch of other WMs all of which have their
advantages and drawbacks, but more importantly most of which are so foreign
to a common user that their mainstream adoption seems very unlikely, dubious
future of GTK+ upon which a volume of audio apps depend, and overall a
fragmented community and resources.

Is it possible to unite the army of volunteers to focus their cumulative
energy on a unified framework even if that offends some, simply because
their preferred toolkit was not chosen for this purpose (can you imagine all
WM developers focusing their energy on one singular WM--now that would be a
miracle!)? Does the community have the courage to make that decision within
our community and take the plunge (of course many lingering questions
remain, i.e. who will be granted the jurisdiction to make such a decision on
behalf of the community, as well as many other logistical and/or
authority-based nightmares)? Can the community learn from the successes of
the companies and/or non-profit organizations who have ventured in the world
of open source and turned profitable, or if not profitable, then at least
leaders in their respective domains? IMHO all of those entities have shown a
strong sense of focus and direction which ultimately empowered them to
maintain the course and brave the narrow path of success.

My hope is that one day the community will reach such a consensus, although
a volume of history suggests unfortunately otherwise.

To offer you some perspective on this one, consider LADSPA plugin framework.
I might be spreading total FUD here, and if so, please do correct me as it
is not my intention to offend. My understanding is that LADSPA was
implemented as a GUI-less standard to offer flexibility for developers, to
some extent end-users, and perhaps also to attract developers from other
platforms. Yet, in retrospect, of all the apps that have resorted to using
LADSPA, we've probably encountered no more than 3-4 different toolkits being
utilized for the LADSPA gui (with GTK+ and Qt being two predominant
options). Needless to mention that reaching out to other platforms has
resulted in very little return (again, please correct me if I am wrong).

Although I am aware that it is easy to sell wisdom now that it has been all
said and done, FWIW wouldn't it have been better to simply bite the bullet
and endorse a single toolkit (perhaps also making sure that it is
open-source and cross-platform to minimize restrictions in its adoption,
even though I personally do not care much for cross-platform capability
since that has not gotten us much of anything even after all these years of
dealing with the additional overhead due to such architectural decisions)?
This way, today we would have had at least a small contingent of LADSPA
plugins which would have looked as cool if not cooler than today's VSTs,
AUs, etc. IMHO, we would've been much better off, at least as far as the
widespread adoption of the Linux audio is concerned. But, as I already said,
I am very much aware that selling wisdom after the fact is cheap. However,
what I hope can be learned from this is that perhaps now is a new chance to
revisit this issue and try to use this as a catalyst for a discussion whose
purpose would be to agree within our little community what will be endorsed
across the board, so that at least drag-n-drop, midi files, DAW session
files, etc., are compatible across different apps. We could simply do a
study of all available options and settle for that one which seems to have
the greatest longevity potential and go from there...

I am sure that some will feel an urge to reply to this rant with an "I don't
care about wider adoption, I only care about how Linux affects my own work."
And I say, all the power to you for making what is IMHO the right *personal*
decision. Yet, whether I like it or not, the power is in the numbers, and
Linux as an audio platform will only persist if there is support for audio
hardware, and if there are software packages which will continue to be
maintained. All this is a lot easier if there are more of us. So whether I
like it or not, I am convinced that the success of Linux audio by and large
resides in the widespread adoption.

At any rate, time to shut up and go back to work...

FWIW, I have no intentions on giving up on Linux audio, for as long as I am
able to run it. I am simply hoping that my rant (apart from wasting
everyone's bandwidth and for those few brave souls who opt for reading this
long e-mail, for wasting their time) may at least get us to start thinking
about these challenges (or perhaps should I say opportunities?).

Best wishes,

Ico



More information about the linux-audio-user mailing list