cduan

This will inevitably draw comparisons to Knuth's METAFONT and Computer Modern font family (as it should). It does suggest, to me, one particularly interesting use case: one could automatically measure parameters for an existing font, and generate similar-looking characters that might be missing from the font (e.g., small caps, Greek letters, Cyrillic, etc.). I programmed this once many years ago, using some ridiculously complicated PostScript and METAFONT programming, but I've since forgotten how that old code works.

The point of PG's essay is well-taken, but I'm not really sure how the linked article really applies other than at a superficial level. I can't find a recent press release from which this article was derived; the closest one is the cited Bain & Co. note from Oct. 2012, which is pretty old by current standards.

Additionally, the reporter here seems to have at least done a modicum of original research, given that he's quoting actual clothing makers. This is in contrast with the USA Today article cited in PG's essay, which quoted a PR guy, a GQ editor, and a consultant.

Thus, to the extent that the submarine theory applies here, it's probably just to show the long reach that a well-manufactured eight-year-old marketing campaign can have. But there were plenty of other marketing campaigns from that time that didn't stick around this long, so maybe that's an indicator that there is some real substance to this trend?

Of course, I've only done a little research into this. If someone can turn up the actual press release or other marketing push from which this article was derived, I will gladly reverse my opinion.

So here's what I gather, after having read the lower court opinion and several of the briefs.

Monsanto owns a patent on certain soybean seeds. They sell 1G seeds to farmers, allowing the farmers to grow them into 2G seeds. The farmers are not licensed to plant the 2G seeds. Bowman bought some 2G seeds and planted them, and Monsanto sued.

Bowman says that the planting of 2G seeds is permitted under the doctrine of "patent exhaustion." According to that doctrine, if a patented physical object is sold under proper license, then a patent lawsuit involving that same physical object is not permitted, even if the object is sold to someone else.

The lower court said that patent exhaustion doesn't apply to the 2G seeds, because Monsanto only granted a license on the 1G seeds. The 1G seeds are not the same physical object as the 2G seeds.

At first I thought this was a simple case, but Bowman is making a very interesting argument in the Supreme Court. It is based on an old case called Quanta.

In Quanta, the patent was directed to a certain computer process, and the patent owner sold computer chips with circuitry for performing that process. The chips themselves were useless, of course, but they just needed to be combined with some standard hardware and turned on to work. Did this mean that, by adding the extra hardware, a new physical object had been made that could be the subject of a lawsuit? The Supreme Court said no: because the chips "embodied" the patented invention and only required standard hardware to be added, the chips invoked patent exhaustion, so lawsuits based on their further use were barred.

Bowman's argument: the 1G seeds "embody" the invention (by having the DNA and biological machinery to produce 2G seeds), and only "standard hardware" (soil, watering, etc.) needs to be added to get the working invention (the 2G seeds), so therefore the 2G seeds fall under patent exhaustion.

The main counterargument is that in Quanta, the original computer chips were still present and intact, whereas the 2G seeds do not include the 1G seeds intact. This requires a narrower interpretation of Quanta, and I could see the Supreme Court going with either this narrower reading or Bowman's broader one.

(For fun, you can try to come up with hypothetical cases that are in between: what if the patentee in Quanta had sold semiconductor masks for making chips?)

I wonder if this algorithm would still present biases given the fact that large cities tend to vote differently from less densely populated areas. My hypothesis is that this algorithm would favor city voters over rural voters, for the following reasons:

1. Successfully gerrymandering in favor of one party generally involves stuffing the other party's voters into one district, while splitting your party's voters into multiple districts.

2. The proposed algorithm favors straight-line divisions through regions such that about half the population of the region is on each side of the division.

3. Thus, chances are, the straight lines will run through the middles of cities, cutting cities into multiple districts.

I'd be interested to see someone test this out--it's likely that my hypothesis is wrong, but it also seems pretty likely to me that there will be at least some urban/rural bias, one way or another.

Here's the patent (7,266,767):

http://www.google.com/patents?id=bHeBAAAAEBAJ

The reason for setting up a DMCA takedown policy is precisely so that the site doesn't have to deal with copyright infringement: by doing a fairly minimal task of responding to takedown notices, the site operator gets freedom from liability for pretty much any related copyright infringement lawsuit (which otherwise could cost a lot of time and money).

It's pretty easy to set yourself up with such a policy, and anyone running a site that includes user-provided content should have one. EFF used to run a startup boot camp that went through all the steps of how to do it, and if they still offer it I would recommend it.

Interesting--there is an email address listed. Wonder what happens if you email your notification there? Presumably, if you do, you are in compliance with the statute, regardless of the no-email policy, so they would have to do the takedown.

Is this new? Because I can't find a corresponding listing in the Copyright Office's list of service provider agents (http://www.copyright.gov/onlinesp/list/a_agents.html).

The reason I'm wondering is that the Copyright Office does require you to provide an email address for your designated DMCA agent, and I'd like to know how they got around that.

(The requirement for registering an agent, by the way, is 17 U.S.C. 512(c)(2).)

Here is one summary I found:

http://uscfsales.wordpress.com/2011/07/01/an-“easy”-engine-f...

Here's the solution I came up with:

http://sbf5.com/~cduan/technical/turing/

tl;dr: Use \A and \z rather than ^ and $ in regexps, because the latter two match newlines in the middle of strings, whereas the former strictly only match the beginning and end of strings.

I thought this was common knowledge, but it's a good reminder for anyone who doesn't know.

Here's what I would like to know: has this lack of privacy resulted in the author not doing something he would have liked to do otherwise? Because the real problem with privacy invasion is not the information that is collected (which more often than not is mundane, like the author's), but rather the chilling effects on activities that, though legitimate, would raise suspicion, embarrassment, or other concern.

I feel like any regular SEO advice book has most of the steps of the claims of this patent.

1. A computer-implemented method comprising:

receiving a document containing content for distribution via one or more online media distribution channels;

You write a web page

analyzing a structure of the document; determining optimal modifications to the structure of the document;

You should put your navigation after your content

analyzing keyword frequency and keyword positions in the content of the document, the keyword frequency and keyword positions being analyzed relative to historical data obtained from one or more of the online media distribution channels;

Here are some buzzwords that have been picked up by search engines in the past ("keyword frequency," "historical data"). Search engines pick them up when they are in the meta tags ("keyword positions").

determining optimum modifications to the keywords in the content of the document based on the structured document;

Thus, you should put those keywords in the meta tags.

displaying recommended structural changes to the document based on the structured document modification and recommended keyword modifications to the content in the document to a user in a graphical user interface;

This is just automation of what an SEO book teaches. It is well-established that automation of a known process is obvious.

receiving input from the user adopting one or more of the structural changes or the keyword modifications to the content; and initiating distribution of the document via one or more of the online media distribution channels.

You modify your web page and put it online.

> A ridiculously broad first claim does no harm

That might be true other than the fact that each claim gives rise to an individual legal right. A ridiculously broad first claim can be asserted as in a cease-and-desist, it can be the basis for a lawsuit, and it can cost those who are threatened thousands of dollars to invalidate, if it even is invalidated. So they actually do cause a lot of harm.

So it seems that the "abstract idea" under Bilski has turned into "does it look complicated enough."

In Research Techs. Corp. v. Microsoft Corp., from last year, the court said that a method of halftoning was not abstract, because it involved a process of comparing pixels. In CyberSource, the court explained that comparing pixels was not abstract because "the method could not, as a practical matter, be performed entirely in a human mind."

Similarly, in the Ultramercial case, the court says that the method is not abstract because "Many of these steps are likely to require intricate and complex computer programming," and "certain of these steps clearly require specific application to the Internet."

These words, "as a practical matter," "intricate and complex computer programming," and "specific application to the Internet," are what throw me for a loop. As a practical matter, I can compare four numbers on a piece of paper, but I can't download a million numbers off of the Internet and compare them. So comparing four numbers on a piece of paper is an abstract idea, but comparing a million numbers on the Internet is not?

I didn't see any science-related words when I took the test. Not sure if this is because of the process by which they made the word lists, or because there truly are not many common science-related words.

I think that they mean copyright registrations, not patent applications:

http://www.copyright.gov/eco/help-deposit.html

"In any copy option where portions of the code are blocked out, the blocked-out portions must be proportionately less than the material remaining."

Though most people I know just use the option to submit object code, since (1) the object code is usually available anyway to anyone who buys the software and (2) since you only have to submit about 50 pages of object code, you're not giving much away anyway.

I've always thought that the best response to "Privacy is less important because I've got nothing to hide" is what I call the "privacy lets you pick your nose" argument. There are plenty of things that we like to do in private, that are not illegal, but which, if revealed publicly, would cause substantial embarrassment. For example, picking your nose, watching porn, lip-syncing to cheesy music in the bathtub.

All of these things create individual pleasure, and thus overall societal utility and value. (That is, they make people happy.) To effectively deny people these individual pleasures by forcing them out into the open would be to reduce overall societal utility. Thus, privacy benefits society by enabling people to engage in legal, but potentially embarrassing, behavior.

(I happen to like Dan Solove's work a lot, and I think he is ultimately coming to the same conclusion as me, but his argument seems to rely on accepting his definition of privacy, while I think that the above argument simply attacks the "I've got nothing to hide" argument head on.)

I've always thought it would be nice if you could cause Ruby to dump its heap and symbol table to a file, which could then be loaded by other instances of Ruby. So, for example, you could run Ruby, load all of the Rails files, and then save everything, so the next time you want to run Rails, you just have to load one file instead of hitting the file system 2000 times.

You could probably also do this by running a Rails instance as a service, and then every time you want to run Rails, you tell the service to fork, and use the forked thread. I think this is how Passenger works actually?

Here is a link to an article with links to the agreement, the order unsealing it, etc.

http://paidcontent.org/article/419-righthavens-secret-contra...

Upon a quick read, I notice that the Defendants' brief argues, on page 9, section III, that "it appears indisputable that Righthaven's assignment of the cause of action is unvalid under Silvers, rendering Righthaven's claim objectively unreasonable," thus opening the door Righthaven having to pay attorneys fees.

EDIT: Here is a quote from the order, regarding Righthaven's arguments:

The use of phrases, in the Motion to Strike, such as “underhanded,” “a ruse,” “blatantly ignored,” “brazen attempt,” “fumbling attempt,” “purposefully muddle,” and “Defendants’ complaint reeks of hypocrisy,” is a very unprofessional attempt to attack counsel rather than address the issues. There is an old adage in the law that, if the facts are on your side, you pound on the facts. If the law is on your side, you pound on the law. If neither the facts nor the law is on your side, you pound on the table. It appears there is a lot of table pounding going on here.

Technically, in the days when programs were written directly in machine code, weren't they all "open source"?

I find it interesting that now both the FTC and a good chunk of the Supreme Court* have warned against vague patents. So far it has been generally hard to challenge patents on grounds of vagueness, as it can only be done in court during an infringement lawsuit. The new patent reform bill provides a mechanism for doing so through the Patent Office, and if it is enacted, it will be interesting to see how it plays out.

*In the recent Bilski case, four justices noted the "potential vagueness of business method patents." Four more justices thought that business methods were unpatentable outright.

There's actually a whole lot more to this patent reform bill than the article lets on. Here is an article listing the proposals:

http://www.patentlyo.com/patent/2011/02/patent-reform-act-of...

I'm particularly interested in the new third-party challenge procedures, which would let people submit arguments during patent examination and then oppose patents on any grounds after they issue (for example, you could contest the patent at the Patent Office for being vague; right now you can only do that in court).

PostScript is another stack-based language based on Forth, but with specialized graphics commands. It also has the delightful property that many printers can interpret it. I always wished I had known this while taking my intro CS courses, so that I could have submitted my problem sets as PostScript files, to be computed by the printer.

From my brief reading, this paper essentially deals with solving an information asymmetry. The males don't know if the females are willing to mate, and the females don't know if the males intend to stay around. Thus, gifts are used as signals that indicate, through a game-theoretic mechanism, the unknown values.

It's strange, then, that the paper doesn't look to the field of information asymmetry in economics, which is a fairly well-developed field at this point and more or less addresses all of the issues considered in this paper. Signaling is well-known as an approach to solving information asymmetries (a second approach is warranties).

Those would be sunglasses that filter out green, otherwise known as red sunglasses.

There was another recent case that touched on the related question of whether violation of a license agreement attached to software could give rise to copyright infringement:

http://www.eff.org/deeplinks/2010/12/mixed-ninth-circuit-rul...

Interestingly, the same three Ninth Circuit judges heard both this case and the OP case.

EDIT: Also, there was the Vernor v. Autodesk case a few months ago:

http://en.wikipedia.org/wiki/Vernor_v._Autodesk,_Inc.

Again, the exact same three judges! They seem quite interested in these first-sale/license issues (or else they wouldn't have published all three opinions).

Ah, this is helpful, as I think it highlights the point of disagreement. If you are of the view that an MLM scheme is legal if there is a "real product" that "might be useful for some people," even if that product is a zero-cost ebook, then I think that this is oversimplistic and you would have to consult a lawyer on it. As I mentioned above, the mere fact that you have a copyright backing a pyramid scheme would not necessarily absolve that scheme of responsibility. My guess is that other factors, such as how you advertise the scheme, would be relevant.

Here is a definition of a pyramid scheme:

pyramid schemes — also referred to as franchise fraud or chain referral schemes — are marketing and investment frauds in which an individual is offered a distributorship or franchise to market a particular product. The real profit is earned, not by the sale of the product, but by the sale of new distributorships. Emphasis on selling franchises rather than the product eventually leads to a point where the supply of potential investors is exhausted and the pyramid collapses.

http://en.wikipedia.org/wiki/Franchise_fraud

Seems like, under that definition, the question is whether people are paying you for the book, as opposed to paying you for the right to engage in the pyramid scheme.

Remember, copyright law gives you a right to exclude, but it does not absolve you of other legal obligations. For example, you can commit antitrust-like violations using a copyright (this is called "copyright misuse"). So if your scheme is indeed a pyramid scheme that violates the law, then copyright will not protect you.

That said, I'm just hypothesizing on facts I'm imagining, so obviously consult a lawyer with any real situations.

So if I'm understanding this correctly:

1. You sell books, and make some money off of the sales.

2. A commission is paid for each sale, to between one and four "referral sources."

Isn't this what every author does, except #2 is called an advertising fee?