[linux-audio-user] Opening up the discussion -> why it is hard in the beginning

[Leonard "paniq" Ritter]

Per aspera ad astra, jono.

The phenomena you have with Linux audio software are not restricted to audio 
software. You have usually a longer time of getting to know the system, but 
once you're in it, its fairly easy to keep things running.

I work as a developer for a software company, and from my experience I can 
tell that stability is the least thing they care for. It needs to be 
marketable, it needs to have new features, technical bugs can be fixed later, 
or not at all, if they are not that serious. But wording, graphics, message 
boxes, novice stuff, all this should be tuned up to the max so that many 
people are being tricked into buying the product.

As a commercial software developer, you are the marketing departments tool.

I have been using audio software under Windows for a long time. It's symptoms 
match the way of development described above: the software is easy to install 
and learn, but it crashes often and wastes huge amounts of CPU.

Additionally, commercial competition creates an uncomfortable working 
environment for musicians, because things do not work together as tightly as 
they could. Any interface you get doesn't match the next one, because those 
companies do not communicate. There is no common underlying system wide 
library for a modular audio environment. You have to buy it indirectly 
through one of Steinbergs products. I  don't think I have to go on and on 
about the disadvantages of the above mentioned development paradigm.

As an open source developer, you are more or less your own boss. Therefore, 
what matters most to you is that the program is stable and efficient. That it 
has been designed to be interoperable. That it fits well into its 
environment. That a lot of things can be configured. Spare time developers do 
not have much time. And I guess they enjoy having the control. There is no 
marketing. A heavenly place.

The result is, that installation and learning can become a bit rough, as with 
open source, you are usually entering the dark caves of infinite development. 
Most things look barebone, but they work very well from a developers 
perspective.

GNU is heaven if you are a programmer or know something about the insides of 
development. There is lots of debug information. There is usually good 
documentation about the insides of a program. There are tons of API's. There 
is a bugtracker that responds to you. The popular GNU programs are a lot more 
stable than what I'm used to, once they run.

And there is of course a good reason not to make programs in development too 
easy to access: a point of view that I like to call "pragmatic elitism".

As it gets easier for untrained novices to enter the realms of an open source 
program, the quality of participation descends. Bug reports become unhelpful. 
Feature trackers become clogged with senseless ideas. Mailing lists burst 
with incompetence. It doesn't have to be like this, but from my experience I 
can tell that if you create software for simple minded people, you get simple 
minded feedback. As a developer, you feel a lot of frustration if this 
happens. It might stall or harm development tremendously.

Think of GNU software as a kind of cockaigne, where the entrance is blocked by 
a big riddle. Once you have solved the riddle (and I am sure that you will 
make a lot of friends this way, because you _must_ participiate in the 
community in order to learn about something), you can enter the land of 
plenty.

I hope that helps answering your questions.

-- 
-- leonard "paniq" ritter
-- http://www.paniq.org
-- http://www.mjoo.org