>>
Doesn't really matter if the legal issues are ultimately in one's favor, if one's erstwhile opponent has enough lawyer time ... even if you can survive the lengthy battle, it may well be a pyrrhic victory.
<<
Well, I would assume two things would happen in a case like that. The first is that any patents we are alleged (unless they are truly rediculous) to infringe upon would be coded around very quickly. Secondly, I am sure that there would be a fund-raising campaign for legal fees.
The patent issue is actually very dangerous. I would be particularly worried about issues such as collateral estoppel where the moment we as the community lose one suit, it makes it harder on every other member of the community to fight similar suits. Therefore I would hightly recommend that in a case like this we really avoid letting something go to trial unless things are really clearly in our favor. IANAL though.
Not having specific assets to be tied up helps, but Oracle could then generate enough FUD it would hamper the use (and spread) of PostgreSQL. Oracle would probably not do so directly but through some 3rd party (or parties).
<<
The danger is great enough that I think it is worth really thinking about this well in advance. OTOH, it would mean a lot of publicity for us and if we are well prepared it could even be a net gain. In my opinion (IANAL, etc) our response should be:
1) Coding around alleged patents
2) Releasing press releases saying that in keeping with our no patents policy, we have coded around patents as soon as they were brought to our attention.
3) Look for legal allies (EFF, FSF, etc).
4) Raise funds for the legal fees of the project.
But I guess most of this should be self-evident.
OTH, has PostgreSQL cost Oracle enough, or does it threaten to cost enough, to make such a venture worthwhile? In the short run it would generate a lot of unsympathetic press and some support for the project.
<<
And.... Patent litigation is pretty risky in another way too. The USPTO will issue a patent for things they have no clue whether they are patentable. If you try to enforce a patent and lose, your patent may be either limited or even invalidated by the courts. This creates a huge gray area where nobody knows for sure whether a patent will be enforceable until it is tried in court (and even then, it could be invalidated later-- see the Lamelson case involving barcode scanners, where the plaintiff argued successfully that Lamelson had invalidated their own patents by keeping them in a pending state for an unreasonable amount of time-- well after they had been enforced against others). This is why if Oracle or another third party were to attempt such an attack, we would probably see a large number of patents being mentioned. This means huge expenses on our part and on the plaintiff's part. But we might have the expertise available to do some real damage to any future plaintiff if they try. Software patents in particular are so dangerous in the US because the FUD factor is so high. All I need say is "Remember RamBus?"
So any suit would also have to take into account the real risk of patent invalidation. This makes it less likely for a smaller company to try to sue us but much more likely for a large company with a large number of patents. So if patents are going to be alleged, I think they would come from Oracle or even MS before some other third party. Here are a list of possibilities and my assessment of likelihood of being a part of an attack if such were to occur:
1) Oracle. Reasonably likely. I am sure they hold a large number of patents and might have the will to use them.
2) Microsoft. Reasonably likely but why would they want to generate publicity for stealth competitors? Long-term possibility. Mid-range, I don't see it. We are probably a pretty low-priority competitior for Microsoft and will probably stay that way for the foreseeable future.
3) IBM. Possible but unlikely. They have been trying really hard for a number of years to help deal with patent issues and FOSS software, first via the often criticized IBM Public License and later the Apache License 2.0. I don't think they want to alienate all those who believe them when they say that IBM sees Open Source as the future.
4) Individual inventor. Possible but unlikely. No possibility of large quantities of damages. High risk (in terms of patent invalidation). I don't see it. I guess Oracle could pay the inventor a few million to sue, but that seems pretty dangerous from their perspecive too as the bad press would almost inevitably come back to them.
5) Independant patent license firms. I guess it is a possibility, but in the end, companies that mostly manufacture lawsuits usually go broke. Why would you sue a non-profit if you were mostly trying to make a buck with the lawsuit?
Best Wishes,
Chris Travers
Metatron Technology Consulting
begin:vcard
fn:Chris Travers
n:Travers;Chris
email;internet:chris@xxxxxxxxxxxxxxxx
x-mozilla-html:FALSE
version:2.1
end:vcard
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
