en.swpat.org is a wiki.   You can edit it.   May contain statements End Software Patents does not endorse.

November 2014: About Microsoft’s patent licence for .NET core

SitemapCountriesWhy abolish?Law proposalsStudiesCase lawPatent office case lawLawsuits


Talk:Publishing information is made dangerous

From en.swpat.org
Jump to: navigation, search

[edit] Evidence that burden of proof is unjust (useful for i4i/Microsoft case)

I suppose a patent has not been awarded in the x264 event, but if the USPTO does not look at open source, does this not suggest it can be very easy (since there is so much open source and things they don't look at) to get bogus patents (see the comment here by a presumed USPTO employee http://opensource.com/law/10/11/software-too-abstract-be-patented#comment-2807 ) to be enforced many years later against innocent parties and lead to likely success of the effort because of very high costs required to manage the burden of proof then placed on the defendants?

If so, this supports that a great problem exists with patents by nature (particularly perhaps with so much software having been developed) since, unlike with copyright, you aren't making the statement that X is violated but rather that no such X exists (which can be arbitrarily difficult to prove even when such X exist).

And I think this is an argument useful for the i4i/Microsoft case (recently accepted by the SCOTUS): that all parties have too high of a burden of proof automatically by stepping into the court room as defendants no matter that the patent was awarded. [This problem is known and probably has been covered various times in briefs, but the x264 case could be tangible evidence.]

The recent Red Hat victory (at $3 million cost to Red Hat for the successful defense) is another useful example of how the huge burden on defendants leads to a lack of justice. Jose X 03:19, 2 December 2010 (EST)

I want to use this argument for the i4i case, but I see one difficulty. The argument is that it's too difficult for examiners to find prior art. If that's true, then we should see a lot of patents being granted (by examiners with limited time) but then subsequently thrown out when they get to court (because the accused and their lawyers have more time and probably more knowledge of the software packages that exist in their domain). Is that what's happening? If so, we just need to prove it. If not, then it looks like the argument is flawed. Ciaran 07:45, 2 December 2010 (EST)
Red Hat... had virtual desktop patent ejected, but at a cost of $3 million and *with* help from the open source community. Do you think most companies or individuals will succeed? That troll went all the way even when it looked like the patents might get tossed. Is it justice that a person has to spend anywhere near $3 million to possibly win a case or else lose everything along the way? Of course not. Justice is if with minimal expenditures, the case gets tossed with summary judgment because the plaintiff did not provide enough evidence aside from the patent. Now, I don't know what evidence needs to be provided to be enough. This is why I stated that the patent game is broken from the get go (vs. copyright). At a minimum, the fair resolution would be to let small businesses off the hook or limit patents much more significantly.
Simply, you can't know what prior art exists in its totality. The Salin essay was arguing this in 1991 for software even before the Internet, Linux, and FOSS took off. Just imagine now. The argument is simple, there are hundreds of millions of lines of code for FOSS. That is all prior art and no examiner can cover that. I even gave a link to one patent examiner that stated very directly that he doesn't really even try. It's not law to try, and he stated the problems with searching for patent ideas (vs. for copyrighted text). At this point in time this system is unsolvable. If we allow software patents, regardless of other reasons not to, then we are putting defendants into the shoes of Red Hat once the technology is old enough.
Repeating.. even with access to FOSS, you would also have to study all of it to see if ideas of patents were implemented or suggested by the code or comments (in contrast to copyright). So it's not about doing a search, but about reading it all and studying each piece in relation to each patent [VCs have stated that patents come from out of the blue even after you search on the actual USPTO site, never mind the issue of searching for prior art in general.]
Let's not forget that Microsoft is one of the **wealthiest companies on the planet**.. and they are crying about a related issue of burden of proof costs. And a bunch of other companies much much wealthier than you or I are also crying. Now, let's ask again, is it fair to Ciaran's small company if a determined troll comes around? What are Ciaran's options? What if Ciaran thinks he found 3 cases of prior art (at I think Red Hat did). Ciaran really believes the patents are not valid and even shows the troll and the court the evidence. His reward: a $3 million debit to the bank account if he goes through to the conclusion through appeals. Imagine now if you think you might not be so lucky to already know prior art! How many man-years will it take to cover all the open source out there and all other prior art (much of which may have effectively disappeared or been closed source.. an issue related to Microsoft's complaint). In fact, it's not the lines of code that exist today in final shipping software. You'd have to dig through all lines of code lying inside version control systems (just for the open source cases). What is that, a billion today? And in 10-20 years when today's patents start getting used (eg, if they get issued in 5-10 years), the FOSS haystack will be much larger. .. Oh, and from a legal perspective, it might not even count as prior art! Yet as shown here, people can certainly take these FOSS devs ideas to get the patent. No justice!! Not even close.Jose X 14:15, 2 December 2010 (EST)
Update: A speculative argument against this would be that the FOSS community is so large in general that they would look for patents on an ongoing basis. However -- Simply, it's clear that abuse is easy to attempt. It's also clear that few will notice. The question is whether someone at all will notice and if so have an interest to say something. We know many profit software firms search for patent violations and do not discover them until the suit comes around. There is also the treble damages discouragement to looking at patents in case you find an older patent to your project. And I don't think for many FOSS coders (especially without the help of an attorney) looking for patent infringement is glamorous or scratches an itch. A FOSS coder can be as inventive as anyone else (this is a fair assumption though it can be argued either way).
Also, there is a report I recently heard about that states that patents that are more litigated have a greater probability of being tossed. This is consistent with this (ie, is supporting evidence). Prior art doesn't magically appear, but the report is consistent with the fact that as more people look, the odds grows a needle will be found and as will a defendant willing to pay the litigation costs.
(I've only read the first few sentences.) Did Red Hat have to pay the $3m or will the patent holder have to pay RH's legal fees? If RH paid, then it might be a good example. If RH got reimbursed, it's a weaker argument (but still has some value since RH still had to have $3m in liquid assets to finance the trial before the verdict). Ciaran 19:02, 2 December 2010 (EST)
And to come back to the question, one example (or 20) doesn't show whether this is something that's a recurring problem. To use this argument, we need to show that this is a pattern. We need to show that this is a defining characteristic of software. Ciaran 19:20, 2 December 2010 (EST)
Considering only this one point, it need not be shown that this is unique to software. If all patents are expensive to litigate AND software creation is practiced by a great number who cannot afford such suits as they engage in original work, then we have a very unjust system.Jose X 12:49, 3 December 2010 (EST)
I think we're agreeing, just using different words. For me, your second sentence is an argument for how this is (not unique but) "particularly problematic" for software. The system is unjust, but the injustice is usually mitigated in most sectors, but in software the injustice usually isn't mitigated, so software patents should be abolished. Ciaran 20:31, 3 December 2010 (EST)
There is another point missing in this picture, and it can be useful to use in the i4i/Microsoft case. [I recognize you may only read a few sentences from this comment.]
It's not just that software is different. The argument is being made that patents are unfair to be used against small folks, period.. no matter the patent. You probably don't need patents against peers in many cases. This is a general attack on patents. It matters who uses it and on whom. This can become a fair use question: it's fair use if you are small (to avoid stifling issues). Remember that the large have many levers. If patents are very expensive to defend, that is acceptable perhaps if it's used only against very large entities. This solves the patent cost problem. [The dual argument is that the burden of proof should change to make defenses much cheaper if only up against a patent since USPTO has much prior art they don't consider so have a very reasonable chance of missing key prior art.]
But we can use the anti-small-entity-attack or even anti-peer attack argument within the narrowed context of software. Apple is an example of a company that made profits left and right and still wants to use patents to end competition (and this after leveraging BSD FOSS). Patents are not intended to protect past success or help you be lazy and hold back progress. It is intended to promote progress, period. And progress is hurt every single time someone is hand-cuffed. Remember that Apple can turn around and copy Google or whomever just as easily. Fact was (for them anyway) that first mover, quality, and marketing was very useful sans patent monopolies.Jose X 23:59, 4 December 2010 (EST)
We need a study - the likes of which Patently-O would publish. Maybe searching that blog would turn up a useful proof. Ciaran 21:26, 2 December 2010 (EST)
I agree more formal studies and detailed coverage is useful.Jose X 12:49, 3 December 2010 (EST)