Slashdot Mirror
How Joel Spolsky Shot Down a Microsoft Patent In 15 Minutes
Thornburg contributes news of a story spotted on Techmeme, writing: "[Joel Spolsky of] Joel On Software has a story about how he found and submitted prior art for a Microsoft patent listed on Ask Patents in 15 minutes. The patent was rejected based largely on the document he submitted." Spolsky gives a very readable introduction to the patent system, and software patents in particular; I especially like this part: "Software patent applications are of uniformly poor quality. They are remarkably easy to find prior art for. Ask Patents can be used to block them with very little work. And this kind of individual destruction of one software patent application at a time might start to make a dent in the mountain of bad patents getting granted. ... How cool would it be if Apple, Samsung, Oracle and Google got into a Mexican Standoff on Ask Patents? If each of those companies had three or four engineers dedicating a few hours every day to picking off their competitors’ applications, the number of granted patents to those companies would grind to a halt."
The big boys build weaponry to keep each other in check, and to eliminate all the smaller boys.
Works nicely for them all.
Don't know why they'd rock the boat.
What's the correct term now? Home Depot standoff?
If each of those companies had three or four engineers dedicating a few hours every day to picking off their competitorsâ(TM) applications, the number of granted patents to those companies would grind to a halt.
Why would these arguably-sociopathic organizations engage in what amounts to mutually-assured destruction for the sake of leveing the playing field?! :p
A Tarantino standoff?
This outfit (previously covered on /. though I didn't find the link as quickly as I'd wanted it) does something similar, though with a different money model.
(Full disclosure: No connection to either, though I had email contact with article one at one time.)
Can you imagine how far back computing would be if we were all stuck with using bubble sort because all the other sorting algorithms were patented? Sure the quicksort patent would have been long expired by now, being developed in 1960, but it would have set us back quite a bit to not be able to use the more efficient sorting algorithms.
Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
hire some old burnt out developer who started back in cobol and assembly. maybe basic as well
feed him patents to read and shoot them down because he had probably done the same thing or read about it
the kids these days, all the know is dragging boxes in an IDE and typing a few words to connect them together
I Welched on my bet and it led to a Mexican standoff with another guy who was an Indian giver. In the end we settled it with a game of Russian Roulette. It was chaos, a real Polish Parliament. In the end, the gun didn't go off and we all felt like we were Gypped.
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
Here is some more prior art: http://en.wikipedia.org/wiki/Mipmap
It already is a standoff. The big companies have an unwritten agreement not to assault each other's patents. When one things it has the upper hand it might start a battle such as Apple vs Samsung, but these are rare. This allows them to use their patents to crush smaller companies without being in danger of having their own patents assaulted.
I'm a good cook. I'm a fantastic eater. - Steven Brust
I didn't bother reading past the point the author claims that pixel density and resolution are synonyms, when the patent at least in the summary appears to be using pixel density correctly.
Can you imagine how far back computing would be if we were all stuck with using bubble sort because all the other sorting algorithms were patented? Sure the quicksort patent would have been long expired by now, being developed in 1960, but it would have set us back quite a bit to not be able to use the more efficient sorting algorithms.
wat? The post you replied to pointed out that you should _not_ be able to patent algorithms... error, retry from start.
I still don't understand it, but there was a patent issue a few years back, where the smaller player put up a plea to the community for help invalidating a certain patent that the megalocorp was wielding against it.
Being curious, I did a quick Google Groups search (Splotsky's 15 minutes sounds about right) and submitted the prior art (a then-defunct software package that was announced on a Usenet group which had the same functionality years before).
A few months later, I got a note from council, asking if I had any contacts with that software company and that they were using my submission as the basis as their challenge, which they ultimately won a couple years later.
Anyway, the surprise was how easy it was for me to find that prior art when the company hadn't managed to. The work I do overlaps with what they do, so, yeah, I had some domain expertise, but so did their employees.
FWIW, they never offered me a token copy of their software or anything for my help. I wasn't expecting it (I'd have no use for it anyway - they make complex proprietary configurations of open source software, while I tend to use the simple-blocks model), but it was also surprising to me that there was no follow-up or loop-closing after the fact. So, if you get into this kind of hobby, do it for the knowledge that you're helping defeat a dangerous patent system.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Have you ever seen a patent application that appears ridiculously broad? (“Good lord, they’re trying to patent CARS!”). Here’s why. The applicant is deliberately overreaching, that is, striving to get the broadest possible patent knowing that the worst thing that can happen is that the patent examiner whittles their claims down to what they were entitled to patent anyway...
An example might help. Imagine a simple application with these three claims:
1. A method of transportation
2. The method of transportation in claim 1, wherein there is an engine connected to wheels
3. The method of transportation in claim 2, wherein the engine runs on water
What he's done is equivalent to finding "a method of transportation," knocking out the ridiculously over-broad claim, which Joel describes as "a long shot lottery ticket". The narrower, dependent claims may still be patentable.
Additionally, this was the first office action in this patent application (not an issued patent, contrary to the /. summary). Would the Examiner have found this piece of prior art or another piece of prior art that knocks out that over-broad claim? Almost certainly, again according to Joel's logic. In fact, the Examiner went on to using 5 other pieces of prior art to address the dependent claims. Any of those could well anticipate the independent claim too.
Is it a good thing to crowdsource prior art searches? Absolutely. But people doing the search can't stop at just a single piece of prior art to knock out the one over-broad long shot lottery ticket claim. In Joel's example claims, finding prior art describing a method of transportation may allow you to run a victory lap and get a Slashdot story, but it does nothing for invalidating the claim to an engine that runs on water, which is really what the patent is about.
"If each of those companies had three or four engineers dedicating a few hours every day to picking off their competitors’ applications, the number of granted patents to those companies would grind to a halt."
There is absolutely zero correlation between those two items. Invalidating a bunch later doesn't reduce the amount they file in the first place. If anything, they'd file more or re-file more specific versions.
I hope not, because in that case the vampires are more trouble than the standoff itself.
Get free satoshi (Bitcoin) and Dogecoins
I Welched on my bet and it led to a Mexican standoff with another guy who was an Indian giver. In the end we settled it with a game of Russian Roulette. It was chaos, a real Polish Parliament. In the end, the gun didn't go off and we all felt like we were Gypped and the Canadians were sorry about the whole mess even though they were not involved at all.
Get free satoshi (Bitcoin) and Dogecoins
What makes you think that the current patents will be allowed to run out? After all, copyrights have been extended forever; how long before patents follow? And can you imagine what computers would be like if Knuth's books were patented?
Don't stop where the ink does.
A patent application has been filed for "Single Seating Furniture Anger Relief System", submitted by, you guessed it, Steve Ballmer.
Donald Knuth once mentioned that there are maybe 500 really fundamental algorithms out there (and he mentioned the Bresenham algorithm as one example of a non-trivial one), and everything else is just derivative and nothing new. So this means that there should about 500 softwarepatents at a maximum (and most of them expired), and everything else is just invalid.
Just because it references an ethnicity, culture, or race doesn't mean it's a derogatory term.
How cool would it be if Apple, Samsung, Oracle and Google got into a Mexican Standoff on Ask Patents? If each of those companies had three or four engineers dedicating a few hours every day to picking off their competitors’ applications, the number of granted patents to those companies would grind to a halt."
What, if anything, is stopping other countries (Russia, China, or even Venezuela & Cuba) from sponsoring engineers to pick off patent applications from US companies?
Has anyone ever detailed the relationship between this guy (J.S.) and Slashdot (editors, readers, seemingly everyone). He (not necessarily undeservedly) has gotten a steady stream of love from this site for many years.
There's no relationship. Joel's just a well-known blogger that writes fairly intelligently about computing topics. A sort of micro-celebrity in the programming world. People like that end up being quoted and talked about in places like slashdot. Particularly when they do things that relate to the general categories that slashdotters love to get angry about (patents, microsoft, etc).
The fact that not one of them is doing that is evidence of collusion. They're using patents to protect their circle and keep lesser entities out.
I Welched on my bet and it led to a Mexican standoff with another guy who was an Indian giver. In the end we settled it with a game of Russian Roulette. It was chaos, a real Polish Parliament. In the end, the gun didn't go off and we all felt like we were Gypped and the Canadians were sorry about the whole mess even though they were not involved at all.
That whole post was Double-Dutch to me. As confusing as a game of Chinese whispers.
(Spudley Strikes Again!)
Everything offends someone, somewhere. You need to strike a balance between empathy and not going out of your freaking mind. :)
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
So Joel Spolsky writes about how efficient and wonderful his own website is and no one here notices ? OK, sterling work on shooting down a rubbish patent, but that is how the US patent system has been working for a long, long time. If any one had come after me with a claim of infringement on such a patent I would have laughed.
There's an easy example of this happening in recent history, where Unisys' assertion of the LZW patent as it applied to GIF spurred the development of the far superior PNG format.
The flaw here is that PNG didnt use any new algorithms. PNG was superior in that is allowed more than just 8-bit images, and supported alpha channels, but PNG simply used the same DEFLATE (LZ77 + Huffman) compression algorithm as already well known and implemented in PKZIP v2.0 archives.
"His name was James Damore."
If I remember correctly, however, there is a connection between Joel and Microsoft.
I Welched on my bet and it led to a Mexican standoff with another guy who was an Indian giver. In the end we settled it with a game of Russian Roulette. It was chaos, a real Polish Parliament. In the end, the gun didn't go off and we all felt like we were Gypped and the Canadians were sorry about the whole mess even though they were not involved at all.
That whole post was Double-Dutch to me. As confusing as a game of Chinese whispers.
That's because the game being played by the big boys is like Chicago style politics. They give you the illusion of choice but in reality, it is a Kansas City Shuffle. You work hard, you work the system, you work the work, if you catch my drift, and you feel like you have a leg up on everyone else. You hear everyone laughing, you laugh along making fun of the other suckers, but the reality is that you are the one that was getting conned all this time.
Nonsense; innovative new sorting methods come out all the time!
I imagine with a little extra funding the USPTO could hire folks like Joel in order to find prior art for these patents.
The issue, as always with government services, is funding.
Perhaps we should consider a frequent filer fine. That would avoid increasing the costs for small businesses or independent engineers.
Alternately we could decide as a society that all of us chipping in some more money each year to improve the USPTO is a reasonable sacrifice to make, but that strikes me as less likely and it punishes the many for the sins of the few.
So, they just add the phrase "on programmable hardware" which encompasses modern computer systems, big deal.
Change is certain; progress is not obligatory.
Set up a crowd sourcing site dedicated to this patent war.
"If any question why we died, Tell them because our fathers lied."
It's a simple balance for me.. if they can beat me up, their opinion matters.
Mexican's don't explode when sunlight hits them.
What makes you think that the current patents will be allowed to run out? After all, copyrights have been extended forever
The fact that 17 years after grant (or 20 years after filing, which is in practice the same thing given how long a patent takes to issue) has stood for several decades is a large part of why there isn't likely to be a Cher Act any time soon.
Are you just shilling? A quick 10 second Wiki search shows that MS now owns less than 40% of the server market share, down from 80% in their prime. Desktops, it depends on who's stats you believe. Most rate Windows in the high 70% range%, but there is a rating of over 90. Since I see how many people are using MAC now days, I tend to disbelieve the 90%. I won't even get into the amount of PCs as a whole declining so causing MS to lose tons of market share to IOS and Android.
-The wise argue that there are few absolutes, the fool argues that there are no probabilities.
There's an easy example of this happening in recent history, where Unisys' assertion of the LZW patent as it applied to GIF spurred the development of the far superior PNG format.
The flaw here is that PNG didnt use any new algorithms. PNG was superior in that is allowed more than just 8-bit images, and supported alpha channels, but PNG simply used the same DEFLATE (LZ77 + Huffman) compression algorithm as already well known and implemented in PKZIP v2.0 archives.
Not only this, there's a counter-example: JPEG vs JPEG2000 -- nost people have never heard of JPEG2000 because the way the patent was enforced meant that it wasn't worth licensing it over JPEG, despite the superior algorithms. And when the Unisys patent expired, suddenly all sorts of software started using LZW, where it couldn't before without licensing, which it wouldn't because the technological increase wasn't THAT much better to be worth yet another license.
More interesting info on Unisys: http://en.wikipedia.org/wiki/Unisys#Controversies
There's also a concept of angular density: the pixels per radian at the expected viewing distance. For example, a 7" Nexus 7 tablet with a 720p-class display has the same angular density as a 28" 720p-class TV or a 42" 1080p TV at four times the seating distance. The CSS 'px' unit, for example, is defined as a unit of angular density: 1/2688 of the viewing distance.
Its not just licensing. The reason that most things use these older compression algorithms is that they are very near optimal for the memory requirements that they demand. If you want your data format to be decompressible on an embedded device with 1 meg of memory then clearly the decompression algorithm cant demand a 2 megabyte entropy model in practice.
Lots of compressors are much better than ZIP (with regard to compressed file size), and in fact lots are also much better than RAR which the pirate community so often uses. Nobody is using any of the top 10 methodologies.
The best compressor for raw bitmaps is currently PAQ8PX which is benchmarked at 1.0392 bits per byte while WINRAR and WINZIP are benchmarked at 1.5194 b/B and 2.4185 b/B respectively. To be clear what I am saying here, that PAQ8PX offers the same level of improvement over WINRAR than WINRAR offers over WINZIP yet still people mainly use ZIP, rarely RAR, and never PAQ.
If you have ever played with the PAQ family of compressors.. they are dog slow (less than 1 megabyte per second) and use lots of memory (often a gigabyte or more of memory.)
"His name was James Damore."
If what he says is true that prior art is so easy to find then we have a corruption problem not a patent problem. Its so painfully obvious that the patent office is Corrupt and the US Attorney General needs to investigate. Period end of story.
Jack of all trades,master of none
I believe the poster was supporting the idea that software patents are a bad idea. If the sorting algorithms had been patented things would suck.
Since patenting of software didn't occur until after many of them had been written, they remain unpatented and this is a good thing.
Reading comprehension is hard.
- No Bounce, No Play -
Description:
1. Follow a well-fed dog around for a while, carrying a plastic baggie.
2. Use your imagination.
There. I have both characterized and shown how to implement the U.S. software patent system.
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
Ask patents is a step in the right direction, but the USPTO still assumes that prior art is the only thing that cna make something non patentable. The non-obviousness criterion is never taken into account.
I remember running into a patent that Google owned on something my software company was implementing. It probably took me 5 minutes to find extremely well documented prior art. Not that this stopped the already awarded patent.
Troll is not a replacement for I disagree.
Since 99% of all patents are insignificant to find prior art for, I think that is an obviously wrong statement. If you can Google search the title of your patent and find prior art on the first page, obviously you wanted your patent to slip through the cracks.
Troll is not a replacement for I disagree.
I have a hard time reading through Nokia's patents that VP8 supposedly infringes on. I thought it was just my inability to read patents, but as no-one on ask patents has been able to help...
Any slashdoters want to give it a try?
Is there anything better than clicking through Microsoft ads on Slashdot?
When it comes to software only copy right laws should apply. No patent on any software should ever be passed.
And wtf, Spolsky has a " very readable introduction to the patent system"? Really?? Come on. OK, I guess maybe it is readable, but that's because it's so condescending & silly. This is particularly true of the cited quote: "Software patent applications are of uniformly poor quality. They are remarkably easy to find prior art for." Yeah, OK, I've seen some poorly written software (and other types of) patents, but many were hard to understand because they were auto-translated from another language. I've also seen some brilliantly written software patents, and not just a few. A sweeping, conclusory statement like this one belongs in a dorm room, not in what purports to be a serious discussion of legal issues.
And sure, anybody who doesn't understand what they're reading can propose prior art and may even generate an interim rejection from a harried examiner who wants an applicant to assume the burden of distinguishing the prior art. But making a rejection stick is another story, and I speak from experience as an attorney who has worked long and hard to "shoot down" overly broad patents in reexam proceedings. If a patent comprises a well-written specification, it takes a lot of work. And you've got to know what you're doing. I'm working on one reexam now, e.g., in which I've probably logged 150 hours drafting leakproof arguments to invalidate 70 claims -- and I'm maybe halfway there. I can tell you right off the bat, Joel's silly efforts -- which apparently amount to little more than Googling prior art that sounds like similar to the subject matter of an application -- won't take much work for a knowledgeable patent agent to blow off. "Friggin' mosquitos!"
But wait, it gets worse! Joel's unsupported claim that patents are deliberately written obtusely to make them difficult to overcome?? HOLY SHIT! Think about that for 30 seconds. Patents are legal documents that concern technical subject matter, not technical documents. If you're an engineer, you're probably not going to be able to understand the language, or even understand why the language exists in the form it does, unless you make some effort to understand the character of what you're reading. I mean, imagine an attorney writing that IEEE engineering papers are deliberately written with unnecessary calculus because the authors want them to be hard to understand, so that they'll make it through the peer-review process. That's how dumb Joel's statement is.
Okay, Bob Lablaw, I know my word count already exceeds the attention span of the average reader here, so let me just close with two concrete statements that may have more probative value than a "he said - he said":
i) six months from now, check Joel's PAIR cite and read Microsoft's response to his prior art, and the PTO's analysis of Micosoft's arguments or amendments. Betcha donuts to torii that Joel's stupid prior art is effortlessly "shot down."
ii) If Spolsky thinks that patents are deliberately written to be obfuscatory, heck, I'm willing to spend a few minutes educating him. I'll let him p