Your idea was tested... and it failed [LWN.net]

Your idea was tested... and it failed

Posted May 22, 2009 6:32 UTC (Fri) by khim (subscriber, #9252)
In reply to: FSF settles with Cisco by trasz
Parent article: FSF settles with Cisco

So, when you develop something that is derived from GPL, your competitors will take all your changes for free and then sell the same product cheaper, because it was you who paid for all this.

Sorry, but this is the same flaved argument nVidia is pushing - and it was proved again and again that it's bogus. By the time competitors are getting access to your code it's already to late: your product is on sale and theirs is not. But the "lost productiovity" with BSD is much worse threat. Read this. Yes, it's true that FreeBSD and NetBSD have good kernels today - years after Linux. And they head-start! BSD was mature OS when Linux was a joke!

What does it mean for business? Not only must you write your top-sikrit proprietary part, you must spend enormous amount of money to play catch-up with GPLed offers! Not a good business investment at all.

(Log in to post comments)

Your idea was tested... and it failed

Posted May 22, 2009 9:24 UTC (Fri) by trasz (guest, #45786) [Link]

Yes, "my idea" was tested - and it was proven to be true, as demonstrated by Apple, Juniper or
Netapp. You don't have this kind of products based on Linux - simply because if something takes
years to develop, it doesn't matter that your competitors will start to sell it three months later than
you.

As for speed of development - the reason for this is very simple: marketing. Linux was - and still is
- all about guerilla marketing and making much noise about minor implementation details. BSD
focused on technical stuff instead. That's why you don't hear much about BSD, even in areas where
it's technically ahead of Linux.

Ah, the glory of things long gone

Posted May 22, 2009 11:25 UTC (Fri) by khim (subscriber, #9252) [Link]

You don't have this kind of products based on Linux - simply because if something takes years to develop, it doesn't matter that your competitors will start to sell it three months later than you.

Sorry, but no. We don't see these kind of products based on Linux because it takes decade or so to create polished product. When Apple's, Juniper's or Netapp's current offers were developed Linux was a toy system. Today Linux is more capable - and today Juniper and Netapp are playing with Linux because it makes no sense to continue with *BSD - but before they'll drop old *BSD-based systems they need to have replacements for them so *BSD-based stuff is not going away overnight. Apple is too tied to current platfrom to easily switch sides.

Linux was - and still is - all about guerilla marketing and making much noise about minor implementation details. BSD focused on technical stuff instead. That's why you don't hear much about BSD, even in areas where it's technically ahead of Linux.

Sorry, but it's just not true anymore. Was not true for many years in fact. When BSD were ahead they often trumped their advantage and were quite vocal - today the only reason to use *BSD is legacy: if you have products and personell trained for *BSD - it's probably Ok to continue to use it. But new projects? Linux with very few exceptions.

Yes, it's true that there exist small niches where *BSD still have advantage over Linux - but it becomes harder and harder to find them. And so even companies pretty dedicated to *BSD for years (aforementioned Juniper and Netapps) are starting to toy with Linux.

We see reinnasance of BSD licensing in the compilers space (rise of LLVM) but it remains to be seen how good this retailation will be - and all other free non-GPL compilers are history (or sideshow at best) today.