Patent of a Database Query Method

[flyingron]

I am sure if he is all that great, you are misinterpreting what he is saying. I've been involved (both as inventor and as licensor) in Software Patents for the past 32 years. Your assertions are WRONG in both theory and practice for computer algorithms.

[nadsab]

OK you are the expert. They were not my assertions though, the stuff I posted came right out of the book - so maybe if you are interested you might want to check out that book and let the author know he made some errors.

Again from the above referenced book: "In order to have your claims be hardware-based to comply with Bilsky, your process must actually be hardware (machine or apparatus) based and you should describe the hardware in that manner in that specification"

[flyingron]

Ok, see you are MISINTERPRETTING what he is saying. He is not saying you have to describe a particular computer (or that the patent is specific to a particuarl computer) but that it has to be defined with respect to hardware (generic). I can describe how my invention is implemented in hardware without tying it to any specific hardware implementation. In fact, this was the standard practice and understanding way before Bilksy. Have you read any software patents?

[nadsab]

Have you read any software patents?

Yes.

- - - Updated - - -

I can describe how my invention is implemented in hardware without tying it to any specific hardware implementation. In fact, this was the standard practice and understanding way before Bilksy.

...which is the point the author was making in the book.

Yes, it was standard practice before Bilsky. However, the point the author was making in the book is that since the above referenced US Supreme Court decision, this has changed, and because of such, that for process claims, software should be tied to specific hardware - that is if the inventor does not want a test case. In other words CYA.

I am not sure however how patents prior to Bilsky would be affected, I would not worry about it though. But for new applications all the author was saying was to CYA and tie new applications to hardware - for process claims.

Have you read the above referenced book?

[tempo_gls]

I’m going to throw in my 0 cents here. I am a DBA so when I hear the idea of copywriting a “query” , weird buzzers go off in my head which might not for the legal folk here. Queries are simply questions asked to a machine containing a plethora of mostly useless information. Finding the needle in a haystack efficiently is partly dependent on how the DB engine processes a request and how quickly hardware and network limitations can service it. That historically has been the patentable part, and I work with the new fun stuff all the time. Queries are “HIGH” level logical definitions for requests to be passed down to machines. I tend to be open, but a search on "sort of like" is nothing special. Implementing it efficiently probably has value but I have little interest in "sort of like"-. In other words ANSI-SQL99 covers all logical bases but permits additional filtering on predicates and joins along the way. That's what you seem to be describing. This seems to be a new “join” technique which will utilize some form of algorithm to parse associate indexed columns on ‘non-matching’ but qualifying fields. The technology that can facilitate that is absolutely patentable in my opinion its merit should be evaluated as similar results are obtained with less overhead by denormalization, good indexing, statistics, and other OS/DB specific methods. There are file based db's which are starting to bridge this gap which I won't list here. Anyway, was reading what seemed to be a bunch of lawyers try to discuss query marketability which quickly deteriorated into vocabulary I could not understand. Best Wishes.....And maybe consider Open Source. I tend to believe that's were all really good but unpatentable technologies end up even if they are temporarily hijacked by lawyers. Kind regards

- - - Updated - - -

- - - Updated - - -

flyingron -


You likely need therapy. I became involved in a previous discussion with nadsab and he/she recommended counseling when such disagreements occurred. I also saw posts immediately after our discussion where he/she was inquiring about the possibility of 'taking legal action against another individual for abuse in cases where both parties were operating under anonymous handles.... Given that we both created our handles withing 48 hrs of each other and whereas I asked for advice and he dolled it out...I'm not too worried. Database is my territory and I can definitively say I don't recommend wasting your time going back and forth with a non-professional...That is unless you're also looking to practice creative writing or vent steam prior to dealing with actual legal folk.

[Dogmatique]

I’m going to throw in my 0 cents here. I am a DBA so when I hear the idea of copywriting a “query” , weird buzzers go off in my head which might not for the legal folk here. Queries are simply questions asked to a machine containing a plethora of mostly useless information. Finding the needle in a haystack efficiently is partly dependent on how the DB engine processes a request and how quickly hardware and network limitations can service it. That historically has been the patentable part, and I work with the new fun stuff all the time. Queries are “HIGH” level logical definitions for requests to be passed down to machines. I tend to be open, but a search on "sort of like" is nothing special. Implementing it efficiently probably has value but I have little interest in "sort of like"-. In other words ANSI-SQL99 covers all logical bases but permits additional filtering on predicates and joins along the way. That's what you seem to be describing. This seems to be a new “join” technique which will utilize some form of algorithm to parse associate indexed columns on ‘non-matching’ but qualifying fields. The technology that can facilitate that is absolutely patentable in my opinion its merit should be evaluated as similar results are obtained with less overhead by denormalization, good indexing, statistics, and other OS/DB specific methods. There are file based db's which are starting to bridge this gap which I won't list here. Anyway, was reading what seemed to be a bunch of lawyers try to discuss query marketability which quickly deteriorated into vocabulary I could not understand. Best Wishes.....And maybe consider Open Source. I tend to believe that's were all really good but unpatentable technologies end up even if they are temporarily hijacked by lawyers. Kind regards

- - - Updated - - -

- - - Updated - - -

flyingron -


You likely need therapy. I became involved in a previous discussion with nadsab and he/she recommended counseling when such disagreements occurred. I also saw posts immediately after our discussion where he/she was inquiring about the possibility of 'taking legal action against another individual for abuse in cases where both parties were operating under anonymous handles.... Given that we both created our handles withing 48 hrs of each other and whereas I asked for advice and he dolled it out...I'm not too worried. Database is my territory and I can definitively say I don't recommend wasting your time going back and forth with a non-professional...That is unless you're also looking to practice creative writing or vent steam prior to dealing with actual legal folk.

So which screen-name were you using before you were banned?

[flyingron]

I don't know of your alleged qualifications, but nobody was talking about "copywriting" [SIC] or even "copyrighting" a query. We're talking patents not copyrights here. I didn't make any comments on the patentability of the database post and despite the fact that it could be implemented in SQL doesn't itself make it not patentable (though it would be highly unlikely to be able to do what the patent is talking about purely with SQL predicates because it involves scoring math rather than just an elaborate set of range selections you are imagining. The issue in the patent is it appears to be pretty common "fuzzy logic" and I'm sure there has to be some prior art on what the claims are.

By the way I've likely have more database experience than you. I was implementing (not just administering) ISAM systems back in 1979 and have implemented several databases for intelligence and medical imaging database.

My argument is that nasdab is misinterpreting both Bilsky and the book he is reading. In fact, Bilsky is actually INTENTIONALLY silent on the matter he is getting bent about (the decision specifically states that it is not making a ruling on that point). What the author states doesn't support what he is saying either.

Try please to avoid making baseless personal attacks. It contributes nothing to the discussion.

[tempo_gls]

well the 'scoring math' part is what I was getting at as the only potential proprietary element. I do know my DB's though and have helped integrate some of the latest technologies coming out of MIT including Vertical Partitioned Columnar Clustered DB's, In Memory Correlation Engines, and all the pieces required to architect Telcom systems processing 10B+ transactions per day with 5 9's SLA's. I sat on the sidelines as the Company I'm referencing was sued by Sybase over 'who owned what' and prevailed by demonstrating a unique methodology combined with proprietary algorithms. For hobby I have sites like this that run on networks you've probably never even heard of let alone gained access to and can code in what ever language the task required. Before that I did my share of VAX/ISAM migrations to newer platforms but find that sort of thing tedious...I'm not a lawyer nor have I stayed at a holiday inn express lately. The other guy read a book though so hey free advise is free advice. I found that after 2 pages of back and forth I was not necessarily speaking to an expert.