Hacker News Reader

Comment by PietdeVries | Hacker News Reader

And is there an alternative to software patents? One of the examples (linked to the case against TomTom) is related to Fat32. I can imagine that if you "invent" a filesystem and everyone is enthusiastic & starts using it - you have the right to be paid for your work?

View on HN

As far as I know, no one advocating abolishment of software patents sees the need for an alternative. Fat32 is not in common use because people are enthusiastic about it, but because it's easy to implement, and it's compaitble. At a technical level, it's no different from trying to patent a database schema.

Anyone who has ever ventured into home brew OS development will probably tell you that it is fairly hard not to invent a FAT-like file system (file allocation table). I'm strongly against software patents, period but even if you think they should be granted in some cases...I feel like FAT is actually too trivial. I guess one could argue that using a pair of tables is innovative etc. etc.

The thing about software is that it gets two bites at the IP cherry. If you write some filesystem code: you get copyright in it. If you take out a patent, then if I write a compatible implementation, you claim rights to my work.

> I can imagine that if you "invent" a filesystem and everyone is enthusiastic & starts using it - you have the right to be paid for your work?

We tried this. It led to the current insanity with software patents and software trolls. I also don't see any indication that innovation would stop if patents didn't exist (see: everything done in the Linux kernel). The trade-off is clear: patents do more harm than good. Get rid of them.

Taking a rights based approach to patents is not the right one, we should look at outcomes instead.

Independent inventors rarely seek patents and their best bet to contest the patent is to sell it to a patent troll, companies tell their employees to never read patent filings to not be willfully infringing and truly novel patented ideas are simply avoided until the patent expires or alternative methods are found (eg ECC crypto).

I'm sure you can find some cases where patents have served as a social good, but I expect them to be outweighed by all their negative outcomes.

Nobody is using FAT32 filesystem due to it being good.

It is used, because it is the only one widely compatible, especially with Windows.

If Windows supported F2FS for example, everyone would be happier.

But what would be a motivation for Microsoft to support a good, free filesystem? One, that it cannot ask a fees for?

The sad thing is that the patents are not even on the filesystem, but on the workaround they had to put it to support long file names. And even that wasn't a novel concept - OS/2 had already added long file names to FAT, just in a different way.

Aren't patents exactly for protecting a particular way of doing things, instead of applying to concepts?

Not to defend software patents, I still think they're a stupid idea, but for patents prior art applies to the exact workings of something, not the general idea and concept.

It's supposed to protect a particular way of solving a problem, but this is more a minor variation of the same way that others have already invented.

Reply on news.ycombinator.com