[LAU] A year of Linux Audio revisited - would like to know your oppinion

[Pieter Palmers]

Chris Cannam wrote:
> On Wednesday 12 December 2007 05:50, Robert Persson wrote:
>> Here is a good explanation of how groove quantise works in protools:
>> http://www.audiomidi.com/classroom/protools_corner/ptcorner_63.cfm
>>
>> For some really fancy midi stuff going way further than groove quantise,
>> you could take a look at some of Ntonyx's products, such as StyleEnhancer
>> and StyleMorpher. If Rosegarden could implement some of those features that
>> would be very useful for composers.
> 
> Groove quantization is one of the oldest outstanding feature requests in the 
> Rosegarden tracker, submitted by me in 2002:
> 
> http://sourceforge.net/tracker/index.php?func=detail&aid=633259&group_id=4932&atid=354932
> 
> Unfortunately, as for so many features, this has simply never percolated to 
> the top of my priority queue, or stack, or whatever my brain uses, and nobody 
> else has ever shown much interest in coding it.  And sadly I only have eight 
> hands and three heads and there are only 132 hours in the day.

IMHO this remark gets to the heart of the problem. As a human being we 
have limited resources. I personally have to settle with one head, 2 
fingers and 24hrs in a day. The fact that humans also have to eat, and 
that food is seldom available 'for free' makes that some part of our 
resources are allocated to surviving. Oh, and having 'a life' also 
consumes an astonishing amount of these resources.

There is no way around the fact that one has to eat, and that that 
doesn't go well with writing free software in the 'free beer' 
interpretation. Which seems to be the main reason why people turn to 
Linux. I've done a quick check on the cheapest offer from Cakewalk 
(Project 5) that might have all functionality that came up in this 
thread, and it is 100euro here. If all Rosegarden users were to pay 
Chris 100euro, I think Rosegarden would have more features than 'Project 
5', or maybe even Sonar.

The catch 22 seems to be that we currently attract a lot of 'free beer' 
attention. But in order to get really professional software, you need 
time, and time = money. So in order for the programs to become more 
professional, we need people that are willing to pay for them. Which we 
don't seem to have.

You could think of it this way: suppose you have a Linux tool that has 
20% of the functionality of an the equivalent Win/Mac tool, you could 
argue that it's worth 20% of the money. Now add all tools you use on 
Linux and try to estimate this '20%' value of your software collection. 
IMHO that's what you would have to be prepared to pay. Of course I know 
that this is a rude extrapolation, and that 20% of the functionality 
usually doesn't get you anywhere. But to be honest, I think most of the 
tools are more near 80% of the 'competing software's functionality.

To give you another idea, from my personal pet project (FFADO):
we are registered on the ohloh site 
(http://www.ohloh.net/projects/8040?p=FFADO), and one of the things they 
do is scan your code repository and use some industry standard way of 
value-ing the code (COCOMO). In the FFADO case they end up with a value 
 > $1.000.000. In other words, if you were to have a commercial company 
develop the code, it would cost you 1M$. But hey, pick me, I'll do it 
'for free'... Rosegarden is also present on ohloh, and is valued more 
than $2M.

I admit that these numbers are large extrapolations and have limited 
applicability, but they do provide some reference.

The only project that seems to be able to break this circle is Ardour. 
I'd say that that is due to the fact that Paul didn't have to worry 
about his survival for the time it took to bring Ardour to a critical 
level. I.e. a level that was high enough for people to start paying for 
Ardour as soon as Paul's self-funding approached it's limits.

For myself I can say that I'm spending an incredible amount of time and 
energy into coding open source, and that there is not that much in 
return. Well, there is the respect from fellow coders, gratitude from 
users, even free hardware (lucky me). But that doesn't pay the bills. So 
I have to go out and spend time at 'something that someone actually pays 
for'. And hence it takes 3 years to reach the functionality that comes 
'out of the box' on another OS. If I would be sure I can earn a decent 
living with writing 'free' software, I would seriously consider it. But 
alas...

Note that this is not really a reaction to the original blog giving an 
overview of the current Linux audio status, but more an attempt at 
expressing my view on why this is as it is, and why it's IMHO not very 
likely to change soon. It's like the Ableton guy said at LAC07: "I'm 
pretty happy with the we-sell-shrinkwrapped-boxed-software model, and I 
don't see a reason to change that.". Read: "why would we give it away if 
people seem to be willing to pay for it?".

2 cents for discussion,

Pieter Palmers