Why You Won't See Traktor on Linux

[tekki]

Guys, please keep yourself in check here. We are a nice community, so let's play nice.

No need to have bad vibes here!

[Jack Bastard]

I'm one of those people who would ditch windows completely if Traktor ran on it, but I can understand why NI can't afford to have a linux development team.

For those who are interested in high end audio work in Linux, my friend who works for a large community project has put together a custom, low latency distribution of Debian for their recording studio. I'll see if I can get a torrent link or something for anyone who'd be interested.

[tekki]

Is that the Dyna:Bolic distro of Debian, Jack?

Or am I way off?


And I will soon be ditching windows, just for two reasons:
1. Midi over Ethernet
2. I have seen it being more stable while performing.

[Jack Bastard]

I think it's an ubuntu variant that's been tweaked for low latency straight out of the box.

[tekki]

Cool!

I also heared about one being a Live distro? (Pop a cd/USB stick in and ready. )

[ctran]

Even if there's no live distro you can use Wubi to safely test it out alongside Windows:

http://wubi-installer.org/

[tekki]

Cool, thanks!

[DennisJ]

so, using linux over windows is better latency wise?

greets

[Monika.mhz]

As often as I argue with Nemonic. I gotta say he's right on this one.

Support is key here. The amount of dollars they'd have to drop into support outweigh the potential dollars they'd make from marketing the product.

Let's say it costs NI: 800,000 to develop Traktor Pro for Windows and another 400,000 to port (is it ported? or redrawn from scratch?) to osx. (Based on about 20 developers payed 30,000 a year and only one year of development.)

Estimating EXTREMELY conservatively, they have to invest 300,000 into getting a *nix version running well enough to distribute.

The general rule of thumb is you need to invest 60% of the development costs for each year over the lifecycle of the product into SUPPORT/IT. If the life cycle of Traktor Pro is 2 years. (with a support carry-over of another 2 years as most companies provide) You're looking at a total cost of 1,002,000 for a linux version of the software.

Keep in mind that all these costs are ULTRA conservative. I guarantee development and support costs from NI are muuuuch higher.

As you can see the cost of development and support skyrockets quickly. Not to mention that when developing linux support costs would most likely be higher (as linux experts just cost more. simply put) and having to develop new docs and troubleshooting methods (some of the costs of developing fixes etc are minimal because NI already has them from years ago!)

I'm just not sure NI is willing to take a 1-6 Million dollar risk to please an uber geeky minority.

[empolo]

As often as I argue with Nemonic. I gotta say he's right on this one.

Support is key here. The amount of dollars they'd have to drop into support outweigh the potential dollars they'd make from marketing the product.

Let's say it costs NI: 800,000 to develop Traktor Pro for Windows and another 400,000 to port (is it ported? or redrawn from scratch?) to osx. (Based on about 20 developers payed 30,000 a year and only one year of development.)

Estimating EXTREMELY conservatively, they have to invest 300,000 into getting a *nix version running well enough to distribute.

The general rule of thumb is you need to invest 60% of the development costs for each year over the lifecycle of the product into SUPPORT/IT. If the life cycle of Traktor Pro is 2 years. (with a support carry-over of another 2 years as most companies provide) You're looking at a total cost of 1,002,000 for a linux version of the software.

Keep in mind that all these costs are ULTRA conservative. I guarantee development and support costs from NI are muuuuch higher.

As you can see the cost of development and support skyrockets quickly. Not to mention that when developing linux support costs would most likely be higher (as linux experts just cost more. simply put) and having to develop new docs and troubleshooting methods (some of the costs of developing fixes etc are minimal because NI already has them from years ago!)

I'm just not sure NI is willing to take a 1-6 Million dollar risk to please an uber geeky minority.

Totally agreed on the ROI part.

The thrust of the debate from my point of view is that it is silly to think that only so-called geeks can navigate their way around and troubleshoot a Linux system. This was very true a short time ago but IMHO, the playing field has leveled quite a bit. Of course there can be some really involved command-line jockeying need to get something to work but you have the same situation in a lot of cases on the Windows platform.

For example, on what other platform would you have to disable a core service (which results in a loss of processing power) in order to avoid latency spikes that result in glitchy audio -->

http://www.native-instruments.com/fo...highlight=BIOS



Again to point out, my original question to the brethren was would they consider moving to the Linux platform ***if*** the software was available for it.