Ad networks

There are millions of websites on the Internet, operated by millions of publishers. A single advertiser cannot practically contact millions of publishers individually to inquire about placing ads. Nor can an advertiser necessarily trust reach metrics (views and clicks) provided by a publisher that isn't well known. So instead, advertisers and publishers use web advertising networks. These allow an advertiser to place an advertisement on websites operated by a wide variety of publishers.

Interest-based targeting

Advertisers don't want to waste their money buying ad space only to show ads that are completely irrelevant to a viewer's demographic. So they pay a premium for ad space more closely aligned to a particular interest. To better match ads to viewers, ad networks build interest profiles about the behavior of visitors to sites operated by publishers that use the ad network.

"Retargeting" occurs when these profiles become as narrow as "viewers who have visited a page about X product". Sellers of this product then place ads for this product through the network in order to encourage people who are interested in this product to complete the purchase. Some users consider retargeting creepy, as if the stores were "stalking" a particular user.

But as phozz bare pointed out in a comment:

Sure, "retargeting" [has] happened to me. As a Firefox user. I don't really care

So long as the ad network operator doesn't disclose an individual visitor's interest profile to the advertisers or publishers using the network, what's the downside of this surveillance?

It's possible to install a GNU/Linux environment on a Chromebook using Crouton. The first step is to put the Chromebook into developer mode, as described in the article "How to Enable Developer Mode on Your Chromebook" by Chris Hoffman. But every time you turn on a Chromebook in developer mode, the Chromebook firmware begs the user to wipe it. Here's the exact wording that the firmware displays:

The first screen (screenshot):

OS verification is OFF

Press SPACE to re-enable.

The second screen (screenshot):

OS verification is OFF

Press ENTER to confirm you wish to turn OS verification on.
Your system will reboot and local data will be cleared.

To go back, press ESC.

(The screenshots are from How-To Geek. Please ignore its pop-ups requesting that you add yourself to the site's mailing list.)

The primary user of a Chromebook in developer mode can press Ctrl+D to skip this prompt each time it appears. But someone else who turns on the device is unlikely to know Ctrl+D. Instead, she is likely to follow the prompts ("Press SPACE" then "Press ENTER") without fully understanding the hardship that following them could cause for the device's owner.

This means the device's owner will not be able to trust the device's internal storage for anything. He will have to operate the device with a USB flash drive sticking out of its USB port, so that all user data gets written to external storage, which is less vulnerable to this sort of data loss. He will also need to carry a second USB flash drive containing reinstallation media for Crouton so that he reinstall Crouton should someone wipe it.

How practical is it to ensure the data on your Chromebook doesn't get wiped without losing eligibility for fixes to, say, the screen hinge and power jack?

The problem is that our society has been systematically eliminating most of the occupations where an honest, hard-working, but not-especially-bright-nor-politically-savvy person can make a decent living.

Then I wrote:

Then perhaps we need to encourage people with those biological advantages to breed more.

Another Slashdot regular told me that comments like these are "dude, not funny". This is something that I occasionally need help to discern because of my mental condition. I think part of my problem comes from trying to fit in with other users on Slashdot who write comments suggesting similarly impractical workarounds out of hardcore laissez-faire ideology: "No jobs in your area? Just move." "No good ISP in your area? Just move." (1 | 2 | 3 | 4) Some such comments even get moderated up.

Sometimes I can get through to them: "Cost of living in some areas has become so high that an entry-level job doesn't pay a living wage." Or "Public high schools aren't doing a good job of teaching basic life skills such as how to relocate for a job. For example, about how much money should I have saved up before I move to, say, Austin?" Yet some posters can't even come up with a ballpark figure. Or a more tongue-in-cheek approach takes them up on their "offer": "How should I go about qualifying for even a temporary work visa in your country?"

But other times I've concluded that it's easier just to try to fit in. If it has in fact gone too far, perhaps I should take the advice of Jesus of Nazareth. To paraphrase Mark 9:45: "If your Slashperger buddies cause you to stumble, cut them off."

Mystikkman claims that I've been posting messages perceived as unjustly hateful toward Microsoft. I would prefer to express my feelings without hate, but sometimes I have the wrong idea of what is hate and what isn't. Please point out which of my comments are hateful and why so that I can understand how not to post next time.

"Let your speech always be with grace, seasoned with salt, that you may know how you ought to answer each one."--Colossians 4:6, NKJV.

In this post, h4rr4r wrote in a reply to a post by roman_mir:

It appears roman_mir is not a native English speaker. Through the "Homepage" link in his profile, I found what appears to be his user page on Mozdev. Roman Mironenko's native language appears not to even be written with Latin letters.

On Slashdot and other web forums, a lot of people reply to comments just to correct the grammar, usage, or mechanics. It's more polite to phrase your correction as a throwaway bit at the beginning of your comment and then, with that out of the way, proceed to make a thoughtful reply to the comment's topic. This way, your comment is more likely to be seen as a sincere attempt to build another user up, rather than the sort of abrasive and inconsiderate personal attack on one's intelligence that has caused people to associate corrections with National Socialism.

In this post, Immerman wrote:

I'm aware of the hyperbole in my illustration. They're probably not equally probable, but absent other evidence, one has to assume so. My point is that just because the probability isn't 100 percent doesn't mean it can always be treated as 0 percent. So if you want to plead false cause more effectively, explain why they're not equally probable. Be willing to discuss what further observations would be needed to show which of the four possibilities is most likely. But don't say "correlation does not imply causation" as if it were "correlation implies lack of causation" without providing evidence, as that's close to the fallacy fallacy and the black or white fallacy.

This discussion has been automatically archived. Discussion continues in Daniel Dvorkin's journal.

People tell me my Slashdot comments are repetitive. I'd appreciate some hints as to how to be less repetitive.

Sometimes it looks like I'm reminding other users of an unsolved problem, but the problem has in fact been solved. Perhaps the real problem is that the solution hasn't been well publicized. For example, one solution to a lot of problems with home entertainment is to put a PC in the living room, but almost nobody knows about this.

If it looks like I'm reminding other users of an edge case too often, consider that a solution that covers more edge cases will appear better thought out and more robust than a solution that covers only the common cases and leaves the edge cases unnoticed.

And sometimes I get confused as to which is the common case and which is the edge case. For example, h4rr4r has pointed out that whenever someone brings up Netflix as an alternative to cable television, I often bring up the fact that Netflix lacks sports. I try to phrase it like "Netflix is fine for people who aren't into sports", recognizing that both non-sports-fan and sports-fan markets exist but apparently putting undue emphasis on the sports-fan market. This goes back to discussions that I've had with heads of household in my survey sample. They tell me they don't see how Netflix would be worth an extra $7.99 per month on top of what they already pay for TV. So I try to make room for Netflix in their budget by suggesting how much they could save by switching from cable Internet+cable TV or fiber Internet+satellite TV to their current Internet+Netflix, and then they mention sports. I guess the survey sample of households in my extended family with broadband access must be a biased sample with more sports fans than the general population, and thus I have a biased view of the relative size of the sports-fan and non-sports-fan markets.

This discussion has been automatically archived by Slashdot. Please take your replies to this page.

USA already has other public options: public schools and USPS Priority Mail over private schools and UPS 3 Day Select.

My signature points out that the United States has a history of public and private sectors competing in a given sector. For example, an engineering student in Indiana can go to Rose-Hulman Institute of Technology (private), or she can go to Purdue University (public). An online hobby store can ship packages to customers with United Parcel Service (private), or it can ship them with United States Postal Service (semipublic, funded by an exclusive contract with the US Government for mailing letters).

As of the third quarter of 2009, health insurance for United States residents under age 65 is mostly provided by employers, who make insurance available to their employees. But not all employers are large enough to qualify for group insurance plans, and some employers even restrict employees to part-time hours so that they don't have to offer coverage. Some insurers offer individual plans, but these are known for refusing to cover people with any of several sorts of preexisting conditions. Estimates of the number of documented U.S. residents without health insurance range from 8 million to 47 million.

The legislature of the United States, called the Congress, has recognized that the lack of universal coverage is holding America back compared to other highly-developed countries. Its members have been debating whether to form a public health insurer to compete with private insurers; this hypothetical insurer has been nicknamed "Public Option" or "Obamacare" in the news media. Some more fiscally conservative members of the Congress argue that any public option would distort the market, and people would leave their current plans and end up on Obamacare. Yes, some people will switch from their current insurer to Obamacare, but that's to be expected: people switched from UPS 3-Day Select when USPS introduced flat-rate shipping boxes.

And the so-called "death panel" is actually called iMac.

The elements of copying are access and similarity. The plaintiff shows some similarity between the works. Then the plaintiff shows that the defendant should reasonably have had access to the work because the work was on the pop charts. This creates a rebuttable presumption of copying. My question: how would one rebut this presumption?

CheapBytes distributes copies of free operating systems for a fee.

What's the difference between an FTP mirror and an HTTP mirror in this case?

Google failed me on SUP, but it found Background Intelligent Transfer Service. That uses only 20 percent of bandwidth anyway, and the article is about throttling to 33 percent (or, alternatively, letting HTTP burst to 300 percent).

Any encoder for MPEG-4 Part 2 violates U.S. patents if not licensed by MPEG-LA, and as I understand it, MPEG-LA's standard license terms are incompatible with the four freedoms that define free software.

Yes, the ISP is trying to do so.

I have played at least three Nintendo DS games over Nintendo Wi-Fi Connection. Animal Crossing: Wild World copies the map from the server to any client that joins, but that's only 88 KB of data.

In this discussion, godrik and I were discussing the relative merits of web applications that use AJAX techniques compared to local applications. I brought up the advantage that web apps can run even on machines where the user isn't allowed to install new software, such as someone else's PC, a set-top web terminal, or a video game console.

Godrik countered that he'd never buy a machine that didn't let its owner install software, and that when he wanted a console to play games on, he bought a Wii and jailbroke it using Bannerbomb. He mentioned plenty of established PC titles that have been ported to libogc, the library used by Wii homebrew: source ports of Id Software's Doom and Quake, emulators such as FCE Ultra, ScummVM, and VisualBoyAdvance, and various Linux-original games that had been ported to about everything, such as SuperTux. These games had presumably recouped their costs of production entirely on the PC.

In general, there are four routes to being able to run code on a closed platform:

1. Make a web application that runs in the console's web browser. These browsers are usually severely limited in performance and in how much of the system's capability the browser exposes through the DOM. Some can't even read more than one gamepad at once, and they're impractical for playing handheld games away from Wi-Fi coverage.
2. Make a pay-per-download game and sell it through the console maker's online store. This is cost prohibitive due to various artificial overheads imposed by console makers such as Nintendo, such as the requirements of a separate office and a prior commercial title on another platform.
3. Make a native game that ships on a retail game disc. This is even more cost prohibitive than download.
4. Make a "homebrew" game that relies on a jailbreak. This is the solution that godrik appears to prefer, but it has problems.

First, jailbreaks break the console's warranty or worse. There are anecdotal reports that Nintendo charges more for out-of-warranty service, such as disc drive replacement after the first 12 months, if a jailbreak is detected than if not.

Second, Nintendo can break Bannerbomb at any time by fixing the defect in a new version of the Wii Menu and IOS. Nintendo would install the fix on newly manufactured consoles and require an update before people can connect to Wii Shop Channel (workaround: WiiSCU) or start newly manufactured Game Discs normally (workaround: Gecko 1.8+). It could take weeks for a new sploit to be developed and released on sites such as WiiBrew, just as it took weeks from Wii Menu 4.0 to Bannerbomb.

But finally, the homebrew community frowns on charging for anything, especially the jailbreaks (Twilight Hack, Bannerbomb) and the launchers (BootMii, Homebrew Channel). That doesn't look good for somebody who wants to feed his family but isn't rich enough to afford the overhead of a license to develop on a closed platform, or even someone who just wants a little economic incentive not to abandon his projects.

One could develop for an open platform such as the PC, but as I mentioned in my last journal entry, not all genres fit well on such a personal computer. For example, a developer might want to make a social game designed to be played with gamepads, a big screen, a sofa, and three friends, such as Nintendo's Mario Party series or Super Smash Bros. series. But four adults can't easily fit around a PC's comparatively small monitor, and a lot of PC gamers seem to be keyboard-and-mouse fanboys who would make other players take turns if they're not old enough to work and buy their own PCs and their own copies of the game. One could go the "home theater PC" route, running gamepads through a USB hub and a VGA or DVI-to-HDMI cable to an HDTV, but two-thirds of U.S. households still have an SDTV in the living room, most PCs don't come with an S-Video output, and the PC to TV adapter isn't sold in stores. Likewise, music games with key sounds, such as Beatmania and Guitar Hero, can feel unresponsive on PC sound cards with their much higher audio latency.

But then, godrik wasn't referring to free as in free beer but instead to Free as in free speech. One way an author can rely on Free is to make the game a massively multiplayer online game based on subscriptions or micropayments. This has its drawbacks: more complexity, requirement for lag-tolerant game play design, cost of administering the game server, need for a separate PC per player, generally no opportunity for children to play due to COPPA and foreign counterparts, failure to reach people who regularly game away from a reliable Internet connection (such as laptop users or people living in the country), and the fact that a lot of people prefer to buy rather than effectively rent their games.

Another way is to make the game engine Free but to charge for the data files, much like Doom and Quake after their GPL release. But are there any success stories of shipping a retail or pay-per-download game whose engine is free software from day one?

Slashdot users: Feel free to Reply to This.
Anonymous readers: Feel free to Reply to This.

There are two kinds of real-time multiplayer video game. Some games require one machine and screen for each player; these are historically associated with personal computers controlled by a keyboard and mouse, connected in either a local-area network or through the Internet. Other games allow multiple players to share a screen. Incidentally, this can be done without splitting the screen, as seen in Midway's Gauntlet, Konami's Bomberman series and Teenage Mutant Ninja Turtles (arcade), and Nintendo's Super Smash Bros. series. These traditionally run on arcade cabinets or on video game consoles with multiple gamepads. The historical reasons for this platform divide include the difficulty in connecting multiple gamepads and the difficulty in fitting four players' bodies around one 14- to 17-inch monitor.

But in the late 1990s, the line began to blur. At first, only consoles had hubs called "multitaps" to connect four gamepads to one machine, but starting with the popularization of USB in 1998, the PC has also had hubs that take multiple gamepads. In the early 2000s, more and more PCs have included composite and S-video outputs for a standard-definition television, and high-definintion televisions have included VGA-style video inputs, solving the screen problem. The rise of home theater PCs has led to demand for multiplayer games designed to fit an HTPC.

Yet even in 2008, this demand has not been met, and the stigma of one PC per player remains. A minority of PC titles, such as Serious Sam, Lego Star Wars, and Midway Arcade Treasures, allow two players on one screen, but not much more. Even cross-platform games whose console version works with more than one gamepad tend to need one PC per player. The landscape of HTPC gaming is so barren that some people have recommended loading up an HTPC with emulators to run unauthorized copies of console game ROMs.

Much innovation in software comes from microISVs, or small businesses that develop software and distribute it on the Internet. These are often home-based businesses and in some cases are run more as a hobby or moonlighting enterprise than as a profit-seeking day job. Some microISVs make their money by developing proprietary software, distributing a trial version at no charge, and selling copies of a version with more features. Others, especially developers of free software, just take donations and advertisements. But the console makers have consistently excluded microISVs from the market. For example, from Nintendo's developer qualifications for Wii and WiiWare: "In addition, an Authorized Developer will have a stable business organization with secure office facilities separate from a personal residence ( Home offices do not meet this requirement )".

Imagine that the head of a microISV has written a design document for a video game intended for two to six people in one room looking at one screen. His team has developed a playable prototype that runs on Windows. For which platform should he and the rest of his team develop and market the final version?

Slashdot users: Feel free to Reply to This.
Anonymous readers: Feel free to Reply to This.

Tip 1: Be sure to quote the parts of the comment you're replying to. Quote multiple levels to recap the discussion from the original article to Slashdot's summary through parent comments if you feel it necessary.

Tip 2: If a comment is far enough off topic, and you can't tie it back to the article somehow, put it in your journal. Then, under the original comment, reply "See my journal" without bonus so that it at least shows up in the other user's messages.pl.

Tonight, Nintendo's new video game console enters the hands of the obsessive-compulsive Americans who have queued up for over 12 hours. I am not one of them. So to pass the time, I put "Wii" into Google image search. Twice. But one result disturbed me (see image): Why is Noddy with Shigeru Miyamoto?

Blyton will probably take me to court for this given an incident from 1999 where I compared Noddy (pics) to Pinocchio (article) on a web page and got a cease-and-desist.

Bayesian text classification is a statistical method of determining the probability that a message is in a given category. It works by making a database of how often each word occurs in messages from a corpus that are or aren't in the category, looking up this probability for each word in a given new message, and then using Bayes' theorem on the probabilities to predict how likely the message is to be in the category.

When applied to the corpus "e-mail" and the category "unsolicited bulk e-mail", the method is called Bayesian spam filtering. For example, words such as "Viagra", "mortgage", "Rolex", "Nigeria", and the like are likely to occur in spam, but some other words are more likely not to occur in spam. Many e-mail service providers applied Bayesian filtering to their customers' incoming e-mail and moved likely spam into a separate folder. This worked ... for a while.

After several months, spammers discovered ingenious techniques to defeat filters. First they disguised the operative words by "creatively" spelling them, Some spammers just misspelled key words: "Ciallis", "mortagee". Others randomly replaced letters with near-homoglyphs from l33tsp34k or from foreign languages: "Viagra" became "Wla9ra" or "\/ 1 A G R @", or "porno" might use a Greek omicron or Cyrillic o or replace the 'p' with the Greek rho or Cyrillic er, both of which look like a Latin 'p'. Anti-spam filters eventually began to check for such techniques and flag them specifically.

Later, spammers attacked the method by using innocuous words in e-mail in order to fool the filter into thinking that a message is not spam. First they used random sequences of letters. Filters blocked words with too many consonants for the target language. Then they used random dictionary words. Filters blocked too many long words in a row. Then they used sentences from literature, as seen in so-called Gutenberg spam and Hobbit spam. These techniques are intended to increase the spam probability of innocuous words, introducing noise into the database and causing the filter to misclassify messages.

However, not all people have the same words marked as not-spam. For instance, people on a constructed language mailing list are more likely to have linguistic jargon marked as not-spam, while people on a video game mailing list may have video game terminology marked as not-spam. Thus, a spammer could collect addresses from a newsgroup, a public web board, a public mailing list, or the contact page of a public web site, and associate each address with words that appear on the same page as the address. How will Bayesian filters block this? Can it be blocked at all?

Warning: Flash links

The old

Zero Wing + Engrish + Photoshop = All Your Base Are Belong To Us

Badgers + Mushrooms + Snake = Badgerx3

The new

From the creator of Badgerx3:
Weebles + Bob + 355/113 = Wobbl and Bob, now on DVD.