Linus Torvalds: Long Filenames (1992) | Hacker News Reader

Linus Torvalds: Long Filenames (1992)

groups.google.com

70 pointsshawndumas16 comments

If Microsoft has been collecting billions in revenue with an invalid patent for years, could the court (or EU Commission?) force them to reimburse that money?

No. Basically if you fought and lost a patent dispute your dispute was still held. If you licensed a patent you licensed it, and it was believed to be valid.

If you dispute the patent, fight, escrow the license fees, or filed with the court, and the court ordered you to continue paying during the dispute, then you could get your money back.

Well they certainly should. It would also be great if a rule was introduced along the lines of "willfull patent abuse" so that since Microsoft was aware of the prior art while still attempting to collect fees they should pay back the money thrice over.

Is this "prior art", or evidence that the idea is "obvious" or "trivial"?

Well, it would be evidence that the idea was trivial, but they just assume that everyone magically knows about the patents we're generally forbidden to read and they won't hold against a patent the fact that it's trivial in hindsight because that's to vague... so they let through a lot of silly patents instead.

What do you mean "patents we're generally forbidden to read"?

Patents are public domain.

You don't work for a large company that does software development, do you?

I do and what's said above is not the case in my company. When we submit a patent we actually have to do our own search for prior art (either published scientific papers, or patents) before some patent lawyer does its own patent search.

That is a very dangerous way to act. If you glanced at another patent. Then 6 months later create a similar sounding patent. File that and the original patent holder can claim you stole their idea. Weblogs would should you looking at it and a lawyer would have no problem saying the idea was stolen.

You have no recourse to prove you did not copy the idea under the current system. The general rule for software developers/patents creators is NOT to look at someone else's work. Let a third party researcher do it and let them decide. As an inventor you have to have a clean room design.

Willful infringement carries a higher penalty than non-willful, so a lot of large companies ask their SW engineers to not read patents.

Many people are forbidden due to the threat of treble damages for willful infringement. Also, they're completely useless, so there's almost no incentive in the first place. I haven't learned a single useful thing from any of the patents I've read, except contempt for the people who think it worth thousands in lawyers fees to patent something like doing email on a cellphone.

When I first heard about this patent, the first thing I thought of was UMSDOS -- but this seems to be something else with the Minix filesystem.

I believe this "discovery" of prior art has already been shown in court a few years ago when the FAT patents were challenged. it has between some time, but this whole thread is very familiar.

I wonder why this patent has taken so long to invalidate.

Microsoft stands to lose a lot of revenue because of the near indelibility of the internet.

[edited]

The internet is very "delable". Usenet wasn't archived en masse until DejaNews. Even then it wasn't a complete archive. They also allowed posts to be deleted. DN sold the archive to Google in 2001, who ironically implemented a terrible search feature. Google also allowed deleting posts until 2009. http://en.wikipedia.org/wiki/Google_Groups#Deja_News

There is a lot of the internet that is missing.

Reply on news.ycombinator.com