Chris Metzler wrote:
As much more of a music maker than a developer, I'm not able to evaluate
whether this post from a current thread on Slashdot is a valid criticism
of JACK:
http://ask.slashdot.org/comments.pl?sid=217898&cid=17700570
I'm curious what people here think?
<Achtung! Grumpy Sunday morning rant!>
AFAIC, any mention of Linux audio on /. is a troll.
I wrote a reply to the original post but didn't send it (see below).
"Slashdot + Linux audio" is a non-starter. I couldn't even get the
service to announce the Linux Audio Conference decently (two years ago).
I'm still astonished by the utter lack of support for Linux audio from
within the Linux community itself. It's as though it never occurs to
users that their precious music is made with non-free tools on non-free
systems. I guess as long as they can play it on a free platform, it's
all okay.
Slashdot is not a good place to go for accurate information. Nice for
some things, sucks for others. Linux audio is one of its others.
The Coward writes :
"You *can* still get Jack to work perfectly with 2ms latency (and
hundreds of people do), but only with extreme care about what
applications you're running and with the use of fast top-end audio
hardware, no Internet access, and no random commands invoked from xterms."
Right, and pros routinely use less powerful hardware and are thoughtless
about their software selection, while they surf the Web and run random
commands during paid sessions. WTF?!
That post is filled with contradictions. This is ripe :
"In contrast, the thousands of other people running Jack with
realtime-lsm on their general-purpose Linux box still get the occasional
XRUN at the best of times despite running with a load average of 0.01,
and for professional work, even 1 XRUN per week is 1 XRUN too many."
Does anyone else notice how adroitly he confutes normal desktop use with
pro audio ? He acknowledges it's possible to get sub 2ms latency for pro
work, but normal users typically get xruns from general-purpose (not
pro-audio) systems, but pro-audio people (who we can safely assume are
*not* using unmodified systems) can't tolerate even a single xrun.
I'm sure Paul and the JACK crew will find this statement interesting:
"it requires *HARD* realtime to work"
His numbers are interestingly suspect:
"hundreds of people do"
"thousands of other people running Jack"
"a million Jack clients"
Even allowing for hyperbole the article reads like it's written by a
troll. According to his title, Jack is broken, but in his discourse he
makes it clear that people can and do successfully run it. So is this a
troll or just poor reasoning and writing ? You decide.
Here's my unposted response:
"I'm curius to read your explanation for the following scenario:
Yesterday I ran ImproSculpt, a massive Csound + FLTK program with
extensive GUI controls, for more than three hours, running under JACK
(set for sub 6ms latency), processing multiple input streams in
realtime, with not one xrun. I surfed the Web and wrote an article while
ImproSculpt did its thing. I do similar stuff with Ardour. Your
assertions don't hold up here at Studio Dave.
Machine is an AMD64 3200. Yes, it's fast, but none of my Mac and Win
audio buddies use anything less. You are bound to use power hardware if
you intend to do serious audio, as I'm sure you know. Btw, Linux distro
is 64Studio."
Also btw: None of my Windows-based music friends like the OS at all.
Some still run Cubase on Win98, citing better performance than with XP.
Those fellows have absolutely 0 intention to purchase Vista for their
music purposes. My friends running Pro Tools *never* run anything else
during a session with that software. They are also unhappy about being
so completely locked into a commercial scheme. Nevertheless, their
innate conservatism and their heavy investment into the Win/Mac world
saps their incentive to step out of it.
Anyone else remember the days when Mozilla received the same treatment
from naysayers ? Anyone using Firefox these days ?
I need more coffee.
</rant>
Best,
dp
