LWN: Comments on "LFCS 2012: The Linux System Definition"

Invalidating "bad patents" solves almost nothing

coriordan — Wed, 02 May 2012 23:24:40 +0000

> what's left after getting rid of the bad patents is "very limited".

Nonsense.

MPEG LA claims that H.264 is covered by about 50 patents in any given country.

Invalidating 45 out of 50 would change *nothing* (even with just one patent, you can't develop or distribute). And current invalidation efforts have invalidated zero out of 50.

The "bad patents" topic is a distraction used by pro-swpat companies and their sock puppets.

LFCS 2012: The Linux System Definition

armijn — Fri, 20 Apr 2012 15:59:45 +0000

Yes, I have. It is actually really simple to write a defensive publication.

You can see various examples of how defensive publications are written at ip.com. Basically it comes down to this: one or two pages, and at least one diagram to describe interaction between components, dataflow, and so on. It does not take that long to write it (about an hour or less for people with some experience).

Very soon Open Invention Network will travel to several conferences (Akademy, GUADEC, COSCUP are planned) to provide hands on assistance with writing defensive publications.

LFCS 2012: The Linux System Definition

wookey — Fri, 20 Apr 2012 10:56:52 +0000

Does anyone have more details on this defensive publication process? I'm (on behalf of emdebian) an OIN member and didn't know/had forgotten. In principle timely defensive publication is a complete antidote to patents, but of course many of us don't do it, or don't do it in a way the patent office is likely to notice. So this seems like a really useful idea if it was just possible to get people to pass on every trivial idea they had today, and OIN could keep up writing them all up. You almost want something like a bug reporting system.

The problem it has is that, like the 'peer to patent' system, there is little incentive for developers to actually spend time on it. Actually the incentive is huge of course - defeating the whole shitty-patents nonsense that is the bane of everyone's lives, but it's so diffuse that very few of us (apoproximately none so far as I can tell) are actually spending time on peer to patent or filing 'please write this up' requests with OIN.

Has anyone here actually participated?

LFCS 2012: The Linux System Definition

jengelh — Thu, 19 Apr 2012 18:39:20 +0000

>The number of packages covered rose from 1100 to 1800

If only they did maintain a list upstream names/packages rather than a list that is specific to the RHEL distro, their bikeshed color and their chop suey flavor, i.e. how they define their %files lists and sub%packages. (Referring to the "iptables-ipv6" BRPM here.)

LFCS 2012: The Linux System Definition

felixfix — Thu, 19 Apr 2012 15:16:43 +0000

One could be paranoid and wonder if Google merely turned off public access to code search. On the other hand, old Microsoft habits die hard, paranoia is fun in its own way, and I don't drink coffee. So maybe I am just looking for brain farts to wake up.

LFCS 2012: The Linux System Definition

Seegras — Thu, 19 Apr 2012 10:43:44 +0000

> those who still benefit from traditional patent licensing (Sony, Philips) > and thus have quite different priorities than the other members.

I wouldn't call it that. Rather "those who think they can quash competition with patents...".

LFCS 2012: The Linux System Definition

njd27 — Thu, 19 Apr 2012 07:43:39 +0000

Another initiative that OIN is working on with Google and IBM is something that Bergelt is (temporarily) calling "Code Search". It will be a way to search pre-existing code and "unstructured project data" for prior art. It make existing code searchable and could be used by the patent examiners as well as the community and OIN. There will be an announcement "soon" about that project.

It seems a bit strange to be reporting that Google are now working with OIN on a "code search" system, given that they turned off Google Code Search earlier this year.

LFCS 2012: The Linux System Definition

laf0rge — Thu, 19 Apr 2012 05:46:49 +0000

Plaese note that the expansion of the Linux System Definition was announced/implemented (March 2012) at the same time that Sony and Phillips decided to carve out a lot of their patented technologies from the OIN License grant.

For example, any patents related to any mobile devices (undoubtedly a major industry with a lot of Linux use) are no longer part of their grant.

I think it pretty clearly states that there are two kinds of members of OIN: Those who actually believe in the general virtue of an open patent pool, and those who still benefit from traditional patent licensing (Sony, Philips) and thus have quite different priorities than the other members.