A lesson to learn [LWN.net]

A lesson to learn

Posted Sep 5, 2006 23:50 UTC (Tue) by Erich_J_Ritzmann (guest, #39670)
In reply to: A lesson to learn by bojan
Parent article: SanDisk faces MP3 licence dispute (BBC)

Probably right. Most FOSS business models operate on a shoestring.

However, most "for profit" companies tend to look at a patent claim less on the merits of the
patent, and more in terms of the business case. If the cost of the legal fees exceed the patent
royalties claimed, most businesses would use the bottom line to decide what to do.

One of the refreshing things about the RIM case is that RIM started out fighting the patent claim
because in their view it was unwarranted. But, eventually the economics of the broken patent
system caused even RIM to cave.

My understanding of how the laws work is imperfect, though would note that people make no
shortage of claims about what their patents cover. Sometimes the courts disagree with the
patent holder, so the patent holder has a hollow claim before the courts decide in his favour.
Perhaps SanDisk made a conscious decision to challenge it in court based on their understanding
of the patents involved?

It might become a very interesting show yet.

(Log in to post comments)

A lesson to learn

Posted Sep 6, 2006 2:39 UTC (Wed) by bojan (subscriber, #14302) [Link]

> Perhaps SanDisk made a conscious decision to challenge it in court based on their understanding of the patents involved?

Yeah, who knows. Anyhow, I'd say the publicity stunt alone is sufficient to scare off some SanDisk customers - at least until the whole thing is resolved. But who knows, maybe SanDisk are confident they will win this so they don't care about some revenue loss and legal fees. And you're right about the bottom line reasoning by most companies - whatever turns out to be cheaper goes.

In any event, I don't think distros like Debian, Fedora etc., that are based on 100% open source software should get involved with MP3 stuff at all. It's a hornet's nest, apparently. I think everyone here at LWN knows what I was referring to with that comment - some people's desire to include MP3 players in those distros. From my perspective, it's a big no-no.

We need a countdown

Posted Sep 7, 2006 2:08 UTC (Thu) by jmorris42 (subscriber, #2203) [Link]

Yes, avoiding mp3 support is essential for any US based distro... now. But we also need to be keeping an eye out for when these patents go away and be ready to drop packages the same day.

Take the Unisys patent and the whole libgif => libungif problem. The patent expired in the US, but it was still thought to be in effect in a couple of countries.... But that was several years ago now and I know I still have libungif on my machine and I'll bet everyone reading this does also. rpmfind no longer even goes back far enough to find a copy of libgif.

We need a project to locate the patents blocking important features and show when they expire in each of the major countries where legal problems would be likely. We have to do it because none of the holders want to even admit they have an expiration date, heck it is hard enough to get them to specify exactly which patents they claim over a topic.

We need this database because over the next five to ten years some really important patents are going to be expiring. MPEG1 Video and MPEG1 layer 1 audio patents should already be starting to expire and MPEG1 layer 3 audio should be clear around the end of the decade. Anyone know when the last MPEG 2 video patent was granted?

We need a countdown

Posted Sep 7, 2006 22:01 UTC (Thu) by dwheeler (guest, #1216) [Link]

A general countdown would be a good idea.

Tracking down the patent databases is really difficult, and it's often difficult to determine if a standard actually REQUIRES a given patent. There IS a simpler way to get an upper bound and probable estimate for standards. As noted here, MP3 was publicly spec'ed in 1991; patents last 20 years, so no patent could be valid after 2011. It's morely likely that patents were revealed at least a year before (because otherwise public discussions would have invalidated the patent filing), so it's probable that any VALID patents would expire by 2010. It's 2006, so it's probably about 4 years to MP3 freedom. It may even be less. So the upper bound would be 2011, and likely less. At least it's a partial answer.

We need a countdown

Posted Sep 12, 2006 3:13 UTC (Tue) by roelofs (guest, #2599) [Link]

That's not a clean example insofar as IBM managed to patent the same thing. Yes, yes, we all know you can't patent the same thing, but the examiners didn't notice, and are you going to pony up the couple of megabucks should IBM suffer a change in management and decide to get feisty? IIRC, there was some complication over who filed first; for all I know, IBM owned the real patent and let Unisys take the heat in exchange for some of the cash. Unlikely, yes, but pretty much any level of patent risk is too much if you're a FLOSS developer. And IBM's beneficence never extended far enough to clarify their opinion on the matter.

Anyway, the point is that IBM's version issued a few years later than Unisys's, so there's this ongoing gray area that may actually extend to the present day. (1989 sort of runs in mind as the year of issuance, but I'd have to go look it up to be sure. Check the comp.compression FAQ if you're curious.)

Greg