Slashdot Log In
Microsoft's Internal Advice About Patents
Posted by
timothy
on Thu Nov 06, 2008 02:34 PM
from the had-that-idea-first dept.
from the had-that-idea-first dept.
BigPoppaT writes "Eric Brechner writes a best practices blog called Hard Code for Microsoft under the name I.M. Wright. His most recent post sounds like an endorsement of open source development (and does end with a call for Microsoft developers to participate in the shared source community). But even better is his advice regarding patents: 'When using existing libraries, services, tools, and methods from outside Microsoft, we must be respectful of licenses, copyrights, and patents. Generally, you want to carefully research licenses and copyrights (your contact in Legal and Corporate Affairs can help), and never search, view, or speculate about patents. I was confused by this guidance till I wrote and reviewed one of my own patents. The legal claims section--the only section that counts--was indecipherable by anyone but a patent attorney. Ignorance is bliss and strongly recommended when it comes to patents.' Interesting advice from inside Microsoft. I wonder if Ballmer would agree that ignorance should be 'strongly recommended when it comes to patents'?"
Related Stories
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
never search (Score:5, Interesting)
The reason you never want to search, is that you could find prior art. If you find prior art, filing for the patent would place you in an inextricable position. If the patent is approved, and you subsequently try to litigate to "protect" this new patent of yours, the records of your fact-checking can (and will) be subpoenaed, at which point you will not only lose the case, but will likely be subject to a counter-suit.
If you don't search, you've got a lot more room to maneuver in court.
Re:never search (Score:5, Interesting)
The other issue is that, if you're creating a new product, and you know that its covered by a patent, you're willfully infringing on the patent, and are vulnerable to additional liabilities. If you can prove that you didn't know about the patent when you created the product, then you're not going to be hit as hard if you're found infringing.
Parent
Re: (Score:3, Interesting)
That is indeed a core reason why engineers are often discouraged from looking at patents. Another reason is that they have a tendency to document their thoughts with ema
Re: (Score:3, Interesting)
Often times, the possibility that you might infringe a patent is pretty clear from a cursory reading
Of course, with the quality of patents granted, the language and the overly broad claims means someone reading patents may very well conclude that there is no software they can possibly write without 'the possibility that they might be infringing a patent'.
When even the patent office can barely tell what it and what isn't valid, the average software engineer has no chance; if you have pockets deep enough, han
Re: (Score:2)
Methinks that I should have phrased it as "often times, the possibility that you might be infringing a particular patent is pretty clear from a cursory reading of that patent." It does make a difference whether you are assessing the risk of infringing any
Re: (Score:3, Insightful)
They're probably better at reading code than the programmers are at reading patent claims.
Who does best; programmers operating a dictionary, or lawyers operating a computer? :P
Re:never search (Score:5, Informative)
This is flat wrong. You have no duty to search for relevant patents. You can only be nailed for willful infringement if you actually knew of the patent you're infringing and did not reasonably believe you didn't infringe. It is up to the plaintiff to prove that you knew about the patent---usually by showing the letter he sent you that says, "You infringe my patent." You then show the letter you got from your patent attorney that says, "My reasonable opinion is you don't infringe this patent," or "My reasonable opinion is that this patent is invalid over X,Y,Z prior art."
Note that willful infringement has nothing to do with whether you infringe. You can infringe and get hit for very substantial damages on a patent you knew nothing about. In that case, searching might have saved you some pain, as you would be aware of what you were facing.
DISCLAIMER: I am a patent attorney, but I don't represent you. This is not legal advice and you don't get to rely on it. Yes, my .sig already says that, but some people turn them off.
Parent
Re: (Score:2, Insightful)
Maybe this is different in Pharma compared to the software industry but in Pharma we most definitely have an obligation establish freedom to patent (i.e., make sure the molecule we are patenting hasn't already been patented). In our case this is done via a series of searches in both text format and via chemical substructure searches of the patent literature.
The current standard is "best effort" which is generally interpreted as searching across all the databases your company subscribes to (chemical patent
Re: (Score:2)
The question wasn't about applying for a new patent. The grandparent said you needed to do a due diligence search just to make a product.
Re: (Score:3, Interesting)
It's always possible that I'm wrong (I don't usually prosecute biochemical stuff), but I think somebody lied to you. I've never seen or heard of a pharma exception where you're required to do a prior art search before you file a patent. I know that lots of them do, and they file IDS's with hundreds of references sometimes, but that's not because they have to. It's because they want all those references on the face of the patent to make it stronger. Has somebody ever shown you the law that says pharma pa
Re:never search (Score:5, Interesting)
As a developer, the reason you don't search for patents (unless directed to by your own lawyer) is that you can be liable for treble damages if you willfully infringe a patent. Also if it is shown that your knowledge of the patent somehow helped you in your R&D, it becomes much harder to construct certain kinds invalidity and non-infringement defenses.
Parent
Re: (Score:3, Interesting)
Re:never search (Score:4, Interesting)
To expand on the treble damages issue: The reason to have your legal counsel search is what the legal counsel finds out is protected by attorney / client confidentiality. Development teams should not search for or discuss patents outside of privileged discussions with counsel.
Parent
Adult literacy classes can help you too! (Score:3, Insightful)
The author is talking about using other peoples' code, which may be copyrighted. They are recommending that the developers should investigate IF the open source tool is protected, and if so that they should respect those limitations. But if they do not know if there are patents on the technology, they should contact the legal department, because the majority of USPTO listings are written like crap with legalise, jargon, and intentional vagueness written in.
What they are trying to prevent is some developer w
Re:never search (Score:5, Insightful)
True, but you also might find a patent that an existing product might violate. That could put your company in the position of having to pay triple damages for willful infringement.
This aspect of patent law shows how broken the system is. The concept of a patent is publication in exchange for a limited monopoly. If no-one can actually read the patent (because of the willful infringement problem), then publication is really a myth.
Parent
Re: (Score:3, Interesting)
So, what's applicable prior art? Anything an attorney can convince a judge/jury is. The only real defense is to ignore the literature and pursue a Kazinski-esque existence. On the other hand, finding evidence of the fraud can be t
Re: (Score:2)
"If you don't search, you've got a lot more room to maneuver in court."
More importantly: if you are sued for patent infringement, and it comes out that you did a patent search, you will likely be sued for triple damages. You MAY be able to survive a court loss for normal damages, but you're far less likely to be able to survive a court loss for triple damages.
Re: (Score:2)
That is part of it but the bigger part is you are an engineer. It is not your job to search for patents. Patents are a never ending black hole as far as time goes. You develop the product. Leave the legal matters to the company lawyers. That is what they get payed for.
Re: (Score:2)
That exact advice and reasoning were stated explicitly in an employee training I took at Microsoft. Incredibly, the words of Steve Martin: "I forgot armed robbery was illegal!" appear to have some actual weight.
Re: (Score:2)
Treble damages (Score:4, Informative)
Generally, I believe that if you are aware of prior art, you're liable for treble (triple) damages. If you're ignorant, you're only liable for normal damages.
MadCow.
Re:Treble damages (Score:5, Funny)
Well, that's the trouble with trebles.
Should you, by some chance, become aware of related patents, you can dismiss the treble problem by beaming them over to Redmond.
What? Like 90% of slashdot considers Microsoft to be far worse than the Klingons anyway.
Parent
Re: (Score:3, Informative)
Tribbles... the trouble with TRIBBLES. :)
Re: (Score:2)
The treble was punnier when you didn't explain it
Re: (Score:2, Funny)
One of things I have found in my career is that ignorance is an excellent strategy. Ignorant people aren't asked to work evenings and weekends because they fuck shit up. That means more free time to play FPS games, sniff petrol and argue on the Internet.
what exactly is strongly recommended? (Score:5, Funny)
I wonder if Ballmer would agree that ignorance should be 'strongly recommended when it comes to patents'?"
Makes you wonder if they think that constitutes "plausible deniability"?
Your honor, you see, we can't possibly have knowingly broken the law, we have a company policy of delibrate ignorance."
Re: (Score:2)
Re: (Score:3, Funny)
Your honor, you see, we can't possibly have knowingly broken the law, we have a company policy of delibrate ignorance."
That seems to sum up pretty much the entirety of their strategies in all of their markets, so yeah, I'd believe that.
Wow, what a messed up summary. (Score:3, Insightful)
Re: (Score:2)
Linux has come under quite a bit of criticism [1] for Linus' "don't look at patents" advice.
[1] As in, NBMers claiming that this proves that Linus and the gang know they're infringing MS patents.
Re: (Score:2)
Re: (Score:2)
Thats a blog about Microsoft culture written by a Microsoft employee. I commend his brutal honesty. If you had read the article, you would have remarked on how there is a strong aversion to using code from other Microsoft products. Developers at Microsoft, don't want to use
Never look at patents (Score:5, Interesting)
I have something like 20 patents in my own name, and if they could I think that our Legal department would have me forget them too.
One does wonder, though, how they are supposed to "advance science and the useful arts" by publication if the publication is supposed to be write-only.
Irony (Score:5, Insightful)
Ironically, the original purpose of patent was to get them published to help others learn new techniques and ideas and expand on them.
How Pointless is That? (Score:5, Insightful)
Then it shouldn't be a valid patent. Further evidence that the patent system in the US needs to be scrapped and rebuilt from scratch.
Patents are ensconced in the Constitution, "To promote the progress of science and useful arts." Even after a patent expires, it is supposed to contribute to humanity's corpus of knowledge and science. How does a patent, written in language that can't be understood by an actual person trying to actually get something useful done, serve to promote the progress of science and the useful arts?
Schwab
Re: (Score:2)
The claims may be hard to understand, but the part that precedes it--the disclosure--is less so. And it has to be detailed enough to enable a person having ordinary skill in the art to practice the invention. The connection between the two is that the claimed subject matter is supposed to be a subset of the subject matter disclosed elsewhere.
Re:How Pointless is That? (Score:5, Funny)
Parent
Re: (Score:2)
Then it shouldn't be a valid patent. Further evidence that the patent system in the US needs to be scrapped and rebuilt from scratch.
The GP probably means that the legal claims section is the only part that matters in court. Patents also contain detailed descriptions and illustrations which you don't always need to be a lawyer to understand.
Re:How Pointless is That? (Score:4, Informative)
Then it shouldn't be a valid patent. Further evidence that the patent system in the US needs to be scrapped and rebuilt from scratch.
If I were to look at your code (presuming you're a coder [for the sake of argument], and presuming I'm *not* a coder [which I am, actually]) and say, "I can't understand this - you must be a terrible coder and your programming language must be broken," would that be fair?
The patent claim attempts to do something unique in the domain of language: it attempts to describe, in the simplest and logically clearest possible manner, the bare minimum requirements of an enabled invention. That's not exactly easy... in fact, it's been called the most difficult type of legal drafting.
The complexity of claim language arises from 170 years of claim-oriented patent drafting. (That's not an exaggeration - the concept of a patent claim arose in the Patent Act of 1836.)
The language of the patent claim has evolved dozens of times as practices have changed (peripheral claiming, acceptable claim styles, rules of interpretation derived from case law.) It may be incomprehensible to you, but those trained in patent law understand not only what a claim covers, but [i]why[/i] it's written that way.
Patents are ensconced in the Constitution, "To promote the progress of science and useful arts." Even after a patent expires, it is supposed to contribute to humanity's corpus of knowledge and science. How does a patent, written in language that can't be understood by an actual person trying to actually get something useful done, serve to promote the progress of science and the useful arts?
A person having "ordinary skill in the art" (i.e., in the technical field into which the patent falls) should be able to read the specification (and review the figures), understand the invention, and be able to practice it without undue experimentation. That's a legal requirement of a patent, and if it's not met, the patent examiner may reject it as insufficiently disclosed. The claims, on the other hand, define what's covered by the patent... in light of the specification.
- David Stein
Parent
35 USC 284 & 35 USC 154.(d)(1)(B) & 35 USC (Score:5, Informative)
Willful violation invokes statutory damages:
35 USC 284
When that doesn't apply, then you have to consider provisional damages:
35 USC 154.(d)(1)(B) Provisional damages:
The intent of willful ignorance is to insulate the company against statutory and provisional damages. As an engineer, this protects you, as well:
35 USC 271.(b):
-- Terry
Happy Belated Hallowe'en ! (Score:2)
I'm struck by the stark contrast between TFA's ideas and the Hallowe'en Documents [wikipedia.org].
With Patents, Ignorance is Bliss (Score:2, Redundant)
Re: (Score:2)
If the law isn't comprehensible by the people governed by it, is it really a just law?
Re: (Score:3, Insightful)
I. M. Wright? (Score:3, Funny)
Re: (Score:2)
It's a pen name. I'm pretty sure he's aware of what it says.
Learned from Stacker (Score:4, Insightful)
Microsoft paid Stac Electronics [wikipedia.org] triple damages for willful infringement in the Stacker case.
Unwinding with some Microsofties at a conference, one of the folk claimed to be fairly close to the technical end of it and said: "we thought we were okay because the patent was very specific about the data structures and the algorithms applied to them. We didn't replicate the data structures at all because there better alternatives. The court didn't care". (Actually Microsoft's will-we-won't-we license it dance was pretty despicable, got a deserved kicking IMHO).
The advice Microsoft is giving here is pretty standard in large engineering companies. Yes, this behavior means patents are largely a self-serving scheme that gives employees a small bonus and makes jobs for corporate legal departments. Yaboo, patents suck...
To compound my cynical view of patents, I recently discovered a colleague had been granted a patent in an area I'd been working on independently and I could not make head nor tail of it because of the lawyer translation process. I went and spoke with my colleague and we had no problem having a technical discussion about the problem and our not so different approaches. Why, why, why, do we put up with the pretense?
Re:Learned from Stacker (Score:4, Insightful)
Because we're engineers and just want to get things done. Everyone else is much more worried about how to make money off of our work, and how to lock it up so that no one else can use it.
Parent
Oh NO you don't! (Score:2)
Richard Stallman already said about Patents, to not look at them....
So again MS Does NOT get the credit for that advice.
Funny if there is anything they do get credit for its trying to take the credit for the works of others.
Quick someone patent "taking the credit for the works of others"....
Re: (Score:3, Informative)
The reason why patent attorneys recommend against searching for patents is that once you know about a patent that is relevant to the one that you are filing, you have a duty to disclose it.
Hmm... your statement is correct, but it's not really "the reason why patent attorneys recommend against searching." I certainly wouldn't recommend a client not search for this reason, and I've never read that motivation expressed.
If prior art is out there that might render your invention non-novel or obvious, it's