But many authors will not take the responsibility of owning their own opinions. Instead, they want somebody else to take the responsibility. We have seen this problem often, and we will point out one recent instance.

A blogger, purportedly writing about “the genre known as ballroom dance”, but really writing about a Broadway show, complains:¹

As a result it isn’t easy to say, “This degrades women.” Yet much of it does.

And again:

It does degrade women.

We at Finding Fault are not in the business of deciding what degrades somebody else. (If we find something that degrades us, we will let you know.) We are, however, in the business of finding fault, so we want to point out here that the Writer in question appears to be a guy, so let’s begin by assuming that the show does not degrade him.

Who, then, does it degrade? We would have liked to have seen him mention some interviews with women who felt degraded by that show. Before you claim that something is degrading to somebody, you need to find at least one person who actually feels degraded by it. And by one we do not mean zero. Was he unable to get in touch with any of the performers—who presumably would have been close to the action and subject to its full effect—to ask how degraded they felt? Was he not able to query a few random female attendees to get their opinions?

Could it be that, rather than ask women if they find the show degrading, he is doing their thinking for them and telling them that they ought to find it so?

We also want to know why, if he wanted to see or experience “the genre known as ballroom dance,” he went to see a Broadway show in the first place. Ballroom dance is, fundamentally, about leading and following.² We think stage shows like the one in question are fully choreographed many weeks ahead of time, and not led and followed. We doubt that there is any significant leading and following in the “the acrobatic manhandling” the Writer complains about or in the manoeuvre of the (allegedly degraded) woman “lifted with her legs in a 180-degree sideways split” that makes him so miserable. All this choreographed showmanship cannot possibly represent ballroom dance in any meaningful way to anybody who has the slightest idea of what “ballroom dancing” really means.

No dance studios or ballrooms nearby?

We are hoping, however, that the so-called computation engine Wolfram|Alpha will be around for a while, because its work has so much potential. So we decided to kill two birds with one stone by not only finding fault, but also performing a public service by helping improve Wolfram|Alpha’s output a little.

We found it surprising that Wolfram|Alpha, based on significant funding and research, could contain such a basic error. We looked in its main stylesheet at http://www.wolframalpha.com/css/basic.layout.css and found that it uses a powerful and dangerous design technique called absolute positioning.

  #sidebar {
    position: absolute;
    right: -157px;
   ...
  }

Finding Fault claims no special expertise in web design and CSS, but we do have a high level of skill in recognizing sloppy output. Our Firefox optimized rebuild, called Swiftfox, showed us this mess:

Fragment of Wolfram|Alpha home page viewed with Firefox optimized rebuild called Swiftfox

Opera rendered the output slightly differently but it was just as messy:

Fragment of Wolfram|Alpha home page viewed with Opera

On our Google G1 phone with the Android browser, the mess turned into a disaster.

Fragment of Wolfram|Alpha home page on Google G1 Android phone, magnified

Wolfram|Alpha home page on Google G1 Android phone, full-screen

The Android browser copes nicely with almost any other web site. Here, for example, is BBC News.

All we did that was different from the norm in each case, if you could really call it “different”, was to set our font size a little larger than usual. “Aha!” we imagine hearing Wolfram declare triumphantly. “Larger fonts!! That will do it every time.” Apparently it will.

Wolfram|Alpha has a lot going for it, and we think it shouldn’t put unnecessary barriers in its own way by making simple web design errors like this.

We haven’t mentioned our exact browser versions (you can get them from the filenames) or the screen resolution and font size settings we used. We don’t want Wolfram to just tweak absolute positioning values to make things look right on some screens while screwing them up on others. The right fix is to get rid of the absolute positioning and let the browser flow the output like browsers are designed to do. The browser knows how to make things fit on the screen without overlapping. You just have to let it.

Here is the first paragraph from that Writer’s article:

The New York Times has an interview out with Yahoo CEO Carol Bartz where she declares that Yahoo has “never been a search company.” Astounding, in that that[sic] this is not true.

The Writer actually cuts off the quote, thus changing its meaning. We will analyze this in more detail later.

But still, our own reaction is: What difference does it make?

The Writer seems to begin with the same sentiment, saying: ‘Part of me thinks, “Why bother arguing?”’

And then, as bloggers generally do,² he proceeds to argue nevertheless.

If argue was all he did, we would leave him alone.

In his argument, however, he commits some serious errors of reasoning, perhaps even deception, giving us the opportunity to find fault.

First, let’s briefly think about how you would go about making a reasoned argument that Bartz’s statement is true or false. We think it would take four steps.

1. Keeping in mind the context, figure out what her statement means.
2. Determine the truth.
3. Decide whether the meaning you determined in Step 1 conflicts with the truth you determined in Step 2.
4. And finally, conclude either that Bartz’s statement is true or it is false.

So far as we can tell, the Writer skips the first three steps and fast-forwards to Step 4.

We don’t demand that these steps be done in a specific order, or that they be labeled the way have labeled them. We do think, however, that you can’t omit the substantive thinking required by these steps.

We at Finding Fault will do the analysis that the Writer skipped.

We think that Bartz was saying that Yahoo’s focus has always been on things other than search. A company can do many things, but “being” a company that does something implies a focus on that thing.

For example, Microsoft has been selling the Microsoft Mouse³ for a quarter century,⁴ but we don’t think of Microsoft as being a mouse company. (OK, well, it’s possible that the Writer in question does—we don’t know. But most people don’t.)

Before we ask whether Yahoo’s focus was search, let’s digress a moment and look back at the glorious days, long past, of a project called Dmoz.⁵ It you look at the Wikipedia entry for this project,⁶ you will note that this project essentially comprised a hand-generated index containing a large number of URLs that grew in size parallel to Yahoo’s index and comparable to it in size.

Hold that thought for a minute—we will come back to it—and let’s now ask what “search” means.

When you look up a word in a dictionary, are you doing a search? You look at the guide words printed at the top of each page, and you end up finding a word after flipping pages just a few times.

Now you could call this a search, and computer science people will tell you that you did something similar to what they call “binary search”,⁷ but most people would call it looking it up in the dictionary, rather than searching for it in a dictionary. There is good reason for this. When you look something up, you have an idea where it will be found, and you know there is some rule for finding it. Dictionaries are in alphabetic order, so you know that a word beginning with “c” will be found after the “b” words and before the “d’ words. And you know that a word like “blogger” will come somewhere after words beginning with “ba” and somewhere before words beginning with “bo”. And so on. And that’s why we generally say that we look up a word in a dictionary rather than search for it.

Back now to Dmoz. It always was, and we assume still is, hand-generated. So this allowed you to look things up under categories and subcategories. But it’s quite hard to drill down through many categories to find the references one is looking for. To make things easier, Dmoz provided, and still does, a search feature. A visitor could simply enter some words into the search box, and almost instantly get a list of matching Dmoz entries.

And that’s how most people used it, so far as we can tell. As a search engine with its own hand-crafted search database.

But did that make Dmoz into a search engine? Let’s investigate a little.

Here is how the Merriam-Webster dictionary defines “search engine”: “: computer software used to search data (as text or a database) for specified information; also : a site on the World Wide Web that uses such software to locate key words in other sites.”⁸

There are two slightly different meanings here. First, a search engine searches data. Second, a search engine searches other sites.

The Random House Dictionary tells us that “search engine” means: “a computer program that searches documents, esp. on the World Wide Web, for a specified word or words and provides a list of documents in which they are found.”⁹

So again, we have two meanings. In the general case, the search engine just searches documents. In the specific case, it searches documents on the web.

Do you see a pattern here? There are two types of search engines: the first type that just searches somewhere, and the second type that searches places on the World Wide Web.

Going back to Dmoz, it’s pretty clear that Dmoz provided the first type of search engine, the type that searched Dmoz’s own data, but not the second type, the type that searches the web.

And this is why, if you do some web searches, you will find that Dmoz was always talked about as a directory, not as a search engine, even though for a period of time Dmoz’s index contained as many as three million URLs.¹⁰

So what have discovered so far in our analysis? Let’s summarize:

1. Finding a word in a dictionary is called looking it up, not searching for it, because we have rules that tell us where we will find it.
2. A search engine that searches a specific set of data isn’t the same thing as a search engine that searches the web.

There is an important third thing we should note, too. Take a look at Yahoo’s home pages through the years.¹¹ Compare these pages with Google’s home pages.¹² We also found a video that plays Google’s home page day by day.¹³ We have retrieved the respective home pages for Yahoo and Google from The Internet Archive¹⁴ for May 10, 2000, and we present them below as thumbnails linked to large images.

Take a look at both, Dear Reader, and tell us:

• Which one screams “Search!”?
• Which one yells “Some sort of weird kitchen-sink portal thing maybe you can also do search on it but I think they are like some type of directory where you can look things up and oh hey they’re also doing auctions and mail and shopping and stuff!”?

Google home page 2000-05-10

Yahoo home page 2000-05-10

So now, let’s go back to our four-step recipe.

What did Bartz’s statement mean when she said that Yahoo had never been a search company? We think she meant that to the extent that Yahoo had done search, it had always concentrated more on searching its hand-crafted index, analogous to the Dmoz index, than on searching the web at large. Thus, “search engine” with more of the first meaning and less of the second. And we’re sure she was also referring to the hotch-potch of other non-search stuff that Yahoo has always featured on its home page.

What’s the truth? The truth appears to be that Yahoo at at times certainly tried to match Google, but it has never focused solely or even primarily on the type of search that Google does. When it has done a Google-like search of the web, it has often relied on outside companies to do some or most of the hard work. Its identity as a company always has been that of a catalog or directory, and a portal, not as simple search engine. In recent years, emphasis has shifted from the catalog or directory to the portal.

Does Bartz’s statement conflict with the truth? We don’t think it directly conflicts, but it does stretch it. Although Yahoo has never focused primarily on being a Google-like company with search as its identity Yahoo has certainly tried to chase Google’s search traffic.

So is Bartz’s statement false? We think not, but again, we think she’s trying to stretch the truth.

Does it really matter? We don’t think our conclusions matter very much. But in determining our conclusions, we did manage to find fault quite a bit—always a fun thing to do and our raison d’être—and we also found a few useful and entertaining things. We hope you enjoyed reading about them.

And oh, we almost forgot to mention this. Going back to the valiant Writer who is so indignant at Bartz. Our revisionist Writer does correctly quote Bartz as saying Yahoo has “never been a search company.” But here was the complete quote in the New York Times:

Ms. Bartz places Yahoo’s position in a rather different light. “We have never been a search company,” she said. “It is: ‘I am on Yahoo. I am going to do a search.’ ”

So she essentially said that Yahoo has indeed provided searches, but they were just one of the things that people did on Yahoo. And this matches closely with the images of Yahoo’s home page through the years that we pointed you to above.

But our Writer, in addition to disingenuously quoting half of the thought from Bartz, also proceeds to pretend (starting right at the beginning, in his inaccurate headline “Bartz Claims Yahoo Was Never A Search Engine”) that she really said that Yahoo was never a search engine. And then he repeatedly tries to prove that Yahoo was a search engine, largely ignoring Batz’s words search company. Typical blogger disingenuity.

So who is the greater revisionist here, Bartz or the blogger?

Finding Fault is pleased to give you a more enlightened perspective. We will present our analysis in multiple parts. In this part, we will give you the simple big picture.

The Simple Big Picture

Google states its mission as:

[T]o organize the world’s information and make it universally accessible and useful.²

Google achieves its mission, and makes money while doing so, as follows.

1. Crawl the Internet, index the content, make it available in searches.
2. Host content on Google’s own servers, index it, make it available in searches.
3. Buy fame, not advertising.
4. Sell relevant advertising to make money.

Consider these Google services:

Blogger, Books,³ Code, Docs, Groups,⁴ Health, Knol, Life Photo Archive, Mail, Newspaper Archive, Patents, Picasa Web, Scholar, Sites, Voice, Youtube.

What do all of these services have in common?

Here’s the answer we wanted you to give us: Content. Specifically, content hosted by Google. Content that Google can index better than anybody else can, because Google doesn’t have to crawl it remotely.⁵

Some of the services we listed contain include private content: Mail, Voice, Docs, Picasa Web, Sites, Health.

Private content is important:

• Private content can be included in search results.⁶
• Private content can be used for selecting relevant ads.
• Private content cannot be indexed by competitors.

Now let’s talk about advertising versus fame.

When you buy billboard space and TV ads and newspaper ads and radio ads and web page ads—that is advertising.

Fame is different.

Do you remember the publicity that Google got from Gmail, when it was announced on April Fool’s Day with a one-gigabyte mailbox size? And then more publicity when people realized it was true? And then even more when people begged one another for Gmail invites and bought them on Ebay?

That is fame.

Google didn’t put ads into Gmail until some years later, but the monetization had begun.

You might have seen bloggers claiming that Google doesn’t know how to make money from Youtube.⁷ Or that Google Docs has failed to replace Microsoft software.⁸ Or that Gmail, Docs, and other products somehow distract Google from search.⁹

All this is probably true, in a literal sense, if you miss the big picture. Finding Fault is pleased to set you straight.

The bloggers consistently miss the synergy between Google’s many products.¹⁰ These products obviously do compete with one another, because resources expended on one product can’t be expended on others. But there’s a bigger picture.

What if Youtube is just a mechanism for Google to get an opportunity to collect and index and search petabytes¹¹ of content? Then that mechanism is working extremely effectively. Google has access to this content in a way that no competitor is ever likely to match. Everybody else crawls websites that link to Youtube content, or crawls Youtube remotely; Google indexes Youtube directly. Everybody else indexes text surrounding Youtube links; Google indexes the audio within the video.¹²

Yahoo announced audio searches several years ago,¹³ but Google has the content.

Youtube as a mechanism for buying fame is working awfully well. Do you know of any other product that presidential candidates publicly use and argue about and, by doing so, implicitly endorse? Not to mention famous athletes and big corporations. Money can’t buy this type of fame.

Is Finding Fault just guessing, or can we verify this?

Finding Fault is guessing—which is our most favorite activity next to finding fault—but that doesn’t make us wrong.

Google made high-resolution videos (first “high-quality” and then “high-density”) available on Youtube, which must cost Google much more bandwidth.¹⁴ Google has spent programming resources to provides a Youtube application for the Android platform (hence for G1 mobile phones).¹⁵ To make this possible, Google now dispenses videos in mp4¹⁶ format.¹⁷ The upload limit for Youtube vides used to be 100 megabytes. It’s now up to two gigabytes¹⁸—taking up up to twenty times more disk space, but providing more content for Google to index. Google is clearly treating Youtube like a valuable property. If Youtube was as big a burden as bloggers would have you believe, Google would not be expending such significant resources, especially in a down economy.

If we recognize Google Docs as a mechanism for Google to collect and index and search more content, then it’s working exactly as it should. Because most Google Docs content is private, no Google competitor can crawl this content and index and search it. And what matters is not Google Docs’s¹⁹ market share relative to anybody else, but the total amount of Google Docs content. The more there is, the more there is for Google to index. Plus, any time you share a document with anybody else, you double the opportunity for Google to search the content—for you, and for the person you shared it with. Triple if three people, and so on.

And don’t forget that Google uses optical character recognition to let you search text within images to a great extent. Its Catalog service (now discontinued) consistently did this, and its Patents service has been quietly doing it for some years. Recently, this feature became official.²⁰ The consequence of this is not only better searches, but searches within Google’s growing collection of older printed content that is only now being scanned into images.

Almost every Google service enhances Google’s brand recognition.²¹

Let’s analyze one Google service that you might naively think is a sure money-loser: Google Code.

If you are an expert programmer who uses Google Code to host his²² project, you will very likely also use Google for all your searches. Because:

• All your Google Code stuff will show up in those Google searches more reliably than anywhere else. Who among us isn’t a little vain? It’s like looking into a mirror as you walk by one.
• Google makes it easy to cross over from each of its services to the others.
• You are using something useful from Google, and you will likely be loyal and stay in the Google world.

But there’s more.

If you are enough of a programmer to be uploading to Google Code, the ten or fifteen people among your friends and family who are struggling with computer crashes, hangs, and viruses, will likely come to your for advice.

Quite likely, your advice will mention Google now and then.

And when you fix these people’s machines, quite likely you will remove those “other” annoying search engine toolbars.

So one Google Code user translates into ten Google Search users.

The bloggers who claim that Google is losing hundreds of millions of dollars on Youtube have no idea how much companies pay for advertising that doesn’t get them a tenth as much publicity. Here are a couple of advertising numbers for the fiscal year 2008.²³

• Microsoft; $1.2 billion.
• Google: $25 million.

All the content Google now hosts is slowly becoming part of its search results.²⁴

Let’s look forward a little. What else could Google add to its enormous collection of information? Finding Fault is pleased to make some wild guesses.

• Old movies. Audio indexing and search will come first. Indexing and searching of video images will eventually follow.²⁵
• Ancient sheet music. Not easy to encode via OCR, but should happen eventually.²⁶
• Other ancient handwritten manuscripts. Again an OCR challenge.
• Old source code, retrieved from punched cards, paper tapes, and other media.
• Old binaries, yes, old binaries. There is a wealth of ideas locked up in old binary-only software. Somebody should collect it and save it so it can be reverse-engineered in the future, long after the original authors are gone and forgotten. Google is the likely candidate.²⁷
• The Internet Archive and its Wayback Machine.²⁸ An acquisition or a cooperative search agreement.

And, by the way, those of you obsessing about antitrust,²⁹ this is where the problem, if it exists at all, might lie in the long run. Not a search monopoly, but a content monopoly.

Concluding, then:

Google doesn’t make money from Youtube,³⁰ or Google Docs,³¹ or Google Code.³²

Google makes money from content: collecting it, indexing it, organizing it, searching it.

And the key to the simple big picture, Dear Reader, is not the indexing, not the organizing, and not the searching, though they are critically important too.

The key is: collecting the content and owning the copy. Not necessarily owning the copyright, but owning the physical copy.

In a future part in this series, we will give you the simpler even bigger picture.

In Red Hat’s 2009 SEC form 10-K,¹ the “risk factors” section lists the company’s estimated risk factors under almost 80 different headings.²

We think such long listings of risks provide questionable value to investors. Investors are looking to figure out what is likely to happen, not the worst that could possibly happen. But once something bad does happen, investors look for somebody to blame.³ Anticipating that, company management tries to make sure that it can later point to its SEC filings and show that it did, in fact, warn these investors of the risks.

Any investor who actually bases his⁴ investment decisions on these risk factor listings in SEC filings will likely end up putting his money under a mattress. Smart investors know how the game is played, and they know what these risk factors are intended to do—protect the company management from liability.

But give a blogger⁵ an opportunity to get a few more hits, and common sense goes out the window. Instead of recognizing these risk factors for what they are—essentially CYA for company management—bloggers get all hyperactive instead.

In Red Hat’s filing, we see this risk:

We are subject to risks of currency fluctuations and related hedging operations.

Does this mean that Red Hat management is shivering from a fear of currency fluctuations? Spending sleepless nights keeping a gun by its side in case ruthless related hedging operations try to break in and burglarize the Red Hat headquarters?

No, it just means Red Hat has cautious management that wants to cut off shareholder lawsuits and SEC action at the border so far as possible.

It’s like those sun protection covers that warn you not to drive with the cover on the windshield. A person sufficiently lacking in sense to drive with the cover on isn’t going to be bothered by the warning, but it may protect the manufacturer from liability later on.

Microsoft’s stock has not been growing much lately. In anticipation of future shareholder lawsuits, Microsoft management is doing the smart thing—thinking of every possible accusation that could be made against it in a future shareholder lawsuit, and making sure it covers itself by listing every possible risk in its SEC filings now, under 20 different headings.⁶ This is not fear, it’s caution and self-defense.

But throwing Linux in front of a blogger can sometimes be like throwing a ball of wool in front of a cat. Cat instincts take over and, for all practical purposes, at least until the cat gets bored, the wool is a fleeing rat.

It all began (this time around) when Blogger 1 happened to note⁷ that “Microsoft’s latest 10-K … adds Canonical, distributor of Ubuntu Linux, to the list of acknowledged competitors for Microsoft’s Client division, which makes Windows for PCs. Also notable is the addition of Linux distributor Red Hat to the list of Client division rivals. Previously, Red Hat was mentioned only as a competitor for the Microsoft Business and Server & Tools divisions.”

Give that man a cigar—straightforward and factual reporting like this is unfortunately rare in the blogosphere.

Blogger 2 took the simple and straightforward quote above and made the following headline out of this: “Microsoft says it’s fighting on two fronts against netbook Linux, browsers”. And then proceeded to claim that Blogger 1 “[N]oted that Microsoft—the company which European regulators continue to allege abuses its dominant position in multiple markets for even greater gains—paints a picture of itself facing significantly greater threats from stronger competitors in two key market areas: the “UX” brands of operating systems, including Linux and Unix; and netbooks, whose rise in popularity has opened up a toehold for Windows’ rivals.”⁸

Blogger 1 does nothing of the sort, so far as we can tell. How do you go from “list of acknowledged competitors” to “fighting on two fronts” and “significantly greater threats”, unless you have an overactive imagination?

We know, of course, that Microsoft has been often described as trying to keep margins up on netbooks. But we have read that elsewhere. We can’t find that anywhere in Blogger 1′s article or in Microsoft’s SEC filing.

What we see Blogger 2 doing is essentially giving his own opinion, and making it seem authoritative by citing another blogger and Microsoft’s SEC filing.

Microsoft’s filing mentions “strong competition” from Apple, Canonical, and Red Hat. And also from Google, Mozilla, and Opera Software Company.

And Microsoft also concludes that section by saying: “Our operating system products compete effectively by delivering innovative software, giving customers choice and flexibility, a familiar, easy-to-use interface, compatibility with a broad range of hardware and software applications, and the largest support network for any operating system.”

Which are you going to believe—the “strong competition” part, or the “compete effectively” part?

We recommend: Neither.

Blogger 2 also claims: “[T]he battle lines [Microsoft] draws for the SEC in its theoretical market map is that of a fortress defending itself from strong, simultaneous attacks on two sides.” Microsoft does nothing of the sort. This colorful description comes from Blogger 2, not from Microsoft. Remember, Microsoft listed 20 different types of risks, which included not only competition from Linux, but also: “We depend on our key personnel that we employ,” “Our quarterly operating results may not be a reliable indicator of our future operating results,” and “We may be subject to greater tax liabilities.” This is just a standard disclaimer.⁹ whose intended audience is a future jury or judge in a future lawsuit.

Blogger 3 chimes in with the headline: “Microsoft admits that it feels threatened by Linux”.¹⁰ Where did “admits” and “feels” come from?

Blogger 4 has a similar headline: “Microsoft admits to SEC: We fear Linux, Ubuntu, and Red Hat”.¹¹

Blogger 5 concludes, “[I]t seems that Microsoft is preoccupied with open source as a threat, rather than as an opportunity. That’s unfortunate.”¹² But if that is true, then isn’t Microsoft also preoccupied with security and privacy breaches? (“Security and privacy breaches may expose us to liability and cause us to lose customers.”) With contract problems? (“If we fail to comply with our customer contracts or government contracting regulations, our business could suffer.”) With its own quarterly results? (“Our quarterly operating results may not be a reliable indicator of our future operating results”.)

Could Microsoft also be preoccupied with Global Warming, tsunamis, meteorite strikes, explosions, wars, and assassinations? (“A disruption or failure of our systems or operations in the event of a major earthquake, weather event, cyber-attack, terrorist attack, or other catastrophic event could cause delays in completing sales, providing services, or performing other mission-critical functions.”)¹³

Blogger 6 claims, “Microsoft is feeling the heat of Linux competition.”¹⁴

Either Microsoft management is shuddering in fear of being attacked from twenty directions (if you take the 20 risk factors seriously) or they are simply anticipating arguments that could be made in shareholder lawsuits and inserting all the appropriate warnings into the SEC filings. Our money is on the latter.

We recommend taking it all with skepticism, because it’s all a game.

Legislators play the game, by making laws that make people feel like they are being taken care of.

The SEC plays the game, by making rules—thousands of them—and designing forms to be filled out.

Corporations play the game, by listing every possible risk that could ever occur.

And finally, bloggers play the game, by desperately grasping at every possible straw and trying to make breaking news out of it.

The proof takes the form of a chart²

Google hits for allegedly bad PR (bluish) versus allegedly good PR (greenish)

showing that a web-based service called Wordnik,³ which supposedly does allegedly bad PR, is getting only limited web traffic, while a different web-based service called Topsy,⁴ which supposedly does allegedly good PR, is getting a high amount of traffic.

Hidden in this argument is an assumption that the Writer never analyzes: That all other factors are essentially similar enough between the two web sites (Wordnik and Topsy) that we can safely conclude that the PR from his blog’s blessing was the major, if not the only, reason why Topsy is getting so much more attention from users.

In other words, he is asserting, in effect, that Wordnik and Topsy are essentially equivalent services differing only in how they do their PR. And from his argument we can conclude that, had Wordnik been the one doing allegedly good PR and Topsy the one doing allegedly bad PR, the results we see would have been essentially reversed.

But has he provided any facts to support his assertion?

None that we can find.

Is he arguing that Wordnik and Topsy perform the same searches?

Not that we can find.

Is he arguing that Wordnik and Topsy perform their respective searches equally well?

Not that we can find.

Is he arguing that Wordnik and Topsy have essentially equivalent funding, so they could invest essentially equivalent resources into their respective services?

Not that we can find.

To us, Topsy looks like a search engine focusing on Twitter, while Wordnik looks more like a dictionary with bonus features. It’s hard to mistake one for the other, if you try a few searches. And neither would replace the other, if you wanted the functionality that each offers.

In our tests, we found that Topsy, which searches Twitter tweets and the web pages they point to, gave us a large number of hits for most searches of words and phrases. But we found that Wordnik gave us useful results for single words but gave us only a few search results from Twitter, and almost nothing else, for phrases. Even though one normally looks up single words, not phrases, in a conventional dictionary, we think Wordnik’s inability to cope with phrases is a bad thing. We also think, but are not sure, that Topsy searches both tweets but the pages they point to, while Wordnik searches only the tweets.

If Wordnik is broken for phrases as we think it is, and limits its Twitter searches to only the tweets themselves, then perhaps this explains why Topsy might be making its users more happier.⁵ Of course, we haven’t done detailed enough research to be sure of the above. But at least we took a look at both services that noticed that they do different things and one of them does certain things quite poorly. Something that the blogger in question didn’t even mention in his comparison.

Not every correlation implies a causation.

In the world of science, when you look for causation, you try to vary only one thing and keep everything else constant.

In the world of blogging, apparently, you just pick the result you want and add some pretty, if blurry, pictures.

We would like to see bloggers come up with better research and better reasoning.

And oh, so long as we are finding fault, we would like to see better naming too. If you tell somebody about Wordnik on the telephone, the other person will more likely type Wordnic (analogous to Internic) or Wordnick.⁶ On the Internet, if they can’t spell your name, you don’t exist.

AP sells five words for $12.50

But before we rush to judgment about AP, let’s just make sure that trying to exert a copyright on five words really is marginally dumb.

Can you get copyright protection on just a few words? Apparently, yes. What matters is not the length but the degree of creativity.¹ The phrases “meter drop” and “rolling stock” in a specific context got copyright protection just a few years ago, at least so long as the parties had not yet settled out of court.² A priori, one has no reason to believe that AP isn’t capable of coming up with short creative aphorisms that might deserve copyright protection.

But it’s hard for news-gathering organizations to come up with such phrases during the normal course of their work. There aren’t that many different ways of talking about elections, assassinations, wars, and airplane crashes. It’s unlikely, we at Finding Fault think, that most short sentences that the AP could come up with would deserve any noticeable copyright protection on their own. To have a blanket policy of asking for $12.50 for five words is a bit silly, it seems to us.

Let’s not forget the “hot news” doctrine, though. According to this much-maligned principle of law, there is value not just in the creativity of words used to report news, but also in their timeliness. Although you can read plenty of discussion about this,³ we get the impression that “hot news” applies only in extreme cases where somebody systematically copies a direct competitor’s headlines or sports scores etc. We doubt that “hot news” would apply in a typical situation where you seek to sporadically quote five or six words from AP.

So we don’t consider it improper to call AP’s five-words-for-$12.50-policy “dumb”.

AP’s website permissions form lets anybody buy a copyright license for anything. It doesn’t have to be any content owned by AP. Just enter any random words, and the website will let you buy a license to use those words.⁴ You can pay AP to give you a license to use public domain words or even your own words, for example.

It would be make no sense if you did so, of course. Nobody in his right mind would actually knowingly pay AP for a license to use public domain words. And besides, it’s questionable that the license would have any validity, because (a) AP can’t sell you what it doesn’t own and (b) you would simply be exploiting an instance of bad web design. To pay for a license of dubious value given to you by a broken website would be a dumb thing.

Just because somebody else lets you do something dumb is no reason to actually do it.

Or so we would have thought.

But there’s no accounting for what a blogger will do to get more hits.

Not only are bloggers going to AP’s website and paying good money to buy the rights to use public domain material,⁵ but they are actually being cited with approval by other bloggers.⁶

And why are they doing this? To prove it can be done, apparently.

Here’s the meta-dialog as we hear it in our imagination:

Website: Licenses for sale! Four random words free, fifth random word for only $12.00. Get yours now!

Blogger: Now that’s pretty dumb. I can buy five random words, even public domain, for $12.00? Yup, that’s dumb. I will take it. Here’s my credit card number.

Website: Um, we’re not the ones acting dumb here sir, you’re the one making the purchase.

Blogger: Got it! Got this license! Now that’s dumb.

Tell you what Dear Blogger,⁷ there are manholes out there that trap people now and then.⁸ Next one that you can find with a loose cover, why don’t you jump in, just to prove that it can be done? Try to survive so you can blog about it afterwards.

Gears is a software facility provided by Google that allows browser-based applications to be productively and efficiently used offline.²

Some months ago, an updated announcement told us that the device now had 1 gigabyte of RAM and still a 4-gigabyte solid state hard drive that would hold the cache.³

More recently, pictures have been published⁴ showing the entire front page of the New York Times apparently viewable on the device.

In essence, the proposed device essentially does everything with a web browser and includes several gigabytes of local storage and runs Gears.

The Gears facility and local storage will let the user continue to edit and read documents and review his saved email. And, presumably compose email for sending, which would remain queued until a network connection was available.

Local storage, offline access to cached documents, offline access to cached email—yes?

No, apparently not, according to this Doomsday headline:⁵

Why The CrunchPad Is Toast.

And here, according to that Writer, are some of the reasons why the CrunchPad is allegedly toast:

The device has no local storage.

The device has no local apps, and only runs Web sites and Web apps.

This, again, tethers you to an Internet connection for even the simplest function, like skimming an old email, reading an e-book, or looking at a to-do list. This also means that app performance will also depend on your Internet speed.

How do we reconcile these claims with the published descriptions of the CrunchPad that we cited above, and of the well-known ability of Google’s Gears to let you run your Gears-enabled applications efficiently without networking connectivity?

We speculate that the Writer had nothing useful to say, decided to come up with a headline that would get him some hits, did a hasty web search and found a page where Michael Arrington stated that the CrunchPad “has no hard drive”,⁶ figured that “no hard drive” must mean “no local storage”, and that that was enough for a Doomsday article.

If this isn’t an example mindless blogging, we at Finding Fault don’t know what is.

A tablet with no local storage? What nonsense. Even low-cost cellphones have local storage these days, and Apple’s iPod line of gadgets have multi-gigabytes of it.⁷ Any device that lets you access email and do web browsing has got to have local storage. And it would make no sense to not use Gears, since it’s free, and it works, and it’s already used to allow browser-based applications to run offline.⁸

A blogger with a publishing background asked a silly question¹ a couple of years ago. And he answered it with a yes.

The question was: Can blogs do journalism?

This question is a silly one not just because a blog just sits there doing nothing, while it’s bloggers that do the work. In this case, the author clarified that he wasn’t referring just to the content management system behind a blog, but also to the people writing for the blog.

The question is a silly one also because answers to questions of the type “can x do y” rarely give us much insight. With some exceptions not relevant here, we generally care more about what somebody usually does, not about what somebody can ever do.

Can Google ever give you useful search results? Does Google usually give you useful search results?

Can The Economist ever publish an informative article? Does The Economist usually publish informative articles?

These are very different types of questions.

So Finding Fault asks: Do blogs—or bloggers—usually do journalism?

If by journalism you mean the type of writing that uses facts and reasoning to yield a useful and reliable conclusion, then we think the answer unfortunately is no.

In the absence of good statistics, let’s just consider some anecdotal evidence.

We will present a brief example first, and then a longer one.

A Writer with whom we are about to find fault goes on and on about how he could not do his work over an iPhone.²

It’s well-known that the iPhone does not allow multitasking. The Writer complains about lack of multitasking, as if it’s news to anyone.³

It’s equally well-known that the iPhone has no keyboard⁴. The Writer complains about the lack of one, as if it’s news to anyone.

The iPhone apparently doesn’t have a very good client for Google Mail. The Writer complains about it. The iPhone provides notifications for various events such as calendar events and instant messages, and the Writer complains about these.

And he asks: “We don’t work like this on our computers—why does Apple think we want to work like this on our phones?” Say what? What makes anybody think that Apple thinks that? Apple makes laptop computers too, for those who need one.

What makes the whole thing even more mindless is that just a few months ago, the same Writer reviewed a device⁵ that has all the features that he is complaining about not having. That device, a tiny netbook that would fit on his palm and in his outer jacket pocket, augmented with tethering to a suitable mobile phone or even his iPhone—which any committed iPhone user can accomplish with only a little effort—would have let him do his work as he wanted to. And yet, in that previous review, the Writer emphatically rejected that device without, so far as we can tell, even trying it out. All this, to us, clearly falls into the mindless blogging category. Complain about something. The complain about something else because it’s not enough like the previous thing that you complained about.

Finding Fault would have been more impressed if the Writer had evaluated the tiny netbook more carefully and then, while complaining about the iPhone, referred back to the tiny netbook and explained why that type of device would not have met his needs either.

Proceeding now to our longer example of mindless blogging, let’s get some background knowledge first.

Lohr and Helft in The New York Times⁶ discuss Google’s antitrust risk:

Google’s power is a cause of worry in many industries—media, advertising, telecommunications and software. Yet being large, successful and ambitious is not an antitrust violation. “You’ve got to be big, and you have to be bad,” observed Andrew I. Gavil, a law professor at Howard University. “You have to be both.”

In the Microsoft case, the software giant’s monopoly in personal computer operating systems was not an antitrust problem. It was its corporate actions, including using contracts and bullying tactics to stifle competition, that broke the law, the federal courts ruled. Such strong-arm practices, legal experts say, have not been part of the Google story.

It’s not the monopoly, it’s how you achieve the monopoly.

On now to our longer example of mindless blogging.

A Writer complains⁷ in an article entitled “Google Antitrust Case Misses the Point” as follows: “The issue with whether Google is or is not a monopoly isn’t merely its search dominance: It’s Google’s ability to control online advertising and, increasingly, what we read or don’t read.”

The Writer misinterprets the target of his own link. His “Google is or is not a monopoly” clause fake-links⁸ to an article⁹ entitled “Reports: DOJ Turns up the Heat on Google’s Book Deal” about a settlement between Google and the plaintiffs suing Google. The settlement will effectively give Google a pricing monopoly over copyrighted works whose owners cannot be located. This settlement involves multiple parties making an agreement that creates a monopoly—a likely antitrust violation. Hence the U.S. Department of Justice has been investigating the situation.

What does the Writer do with this? Well, he writes: “The issue with whether Google is or is not a monopoly isn’t merely its search dominance:…”, thus implying that the article to which he linked is about Google’s search dominance. But the article to which he links really refers to an agreement, not so much about causing search dominance, as about giving Google a type of monopoly over copyright.¹⁰ Google already has search dominance and the settlement doesn’t increase Google’s search share.¹¹ After the settlement, Google will be able to sell access to copies of some copyrighted works with little or no risk in a way that nobody else will. This confusion between search and access to copyrighted works is the Writer’s first mistake.

The Writer’s second mistake is to confuse between two types of antitrust situations. There is the type that involves multiple parties making an agreement that might create a monopoly. Then there is the type that involves a single party creating a monopoly. Wikipedia has an informative article¹² in which we find:

A distinction between single-firm and multi-firm conduct is fundamental to the structure of U.S. antitrust law, which, as noted antitrust scholar Phillip Areeda has pointed out, “contains a ‘basic distinction between concerted and independent action.’”

Those who trust neither this Wikipedia entry nor the references it cites could look up the Antitrust page¹³ provided by Cornell University’s Legal Information Institute and then follow numerous links to get the same information. In referring to an agreement between multiple parties, and then complaining about the monopoly of a single party, the Writer shows no awareness that antitrust rules might apply differently to the two types of situations. If multiple parties act together, this can violate antitrust law even if nobody does anything “bad” and no harm to competition occurs.¹⁴ But a single party that attains a monopoly doesn’t violate antitrust law unless that it does something bad.

Accusing an article about the DOJ’s actions, or accusing the DOJ itself if that’s what the Writer intended, of missing the point because the DOJ is enforcing the law as it stands, makes no sense to us. We think the Writer is the one missing the point.

The third and final error the Writer makes is a subtle one. He links to an article about the application of antitrust law and, in this case, about a settlement that brings the lawsuit to an end, and about the attention the US Department of Justice is paying to this because the settlement involves multiple parties making an agreement that may lead to a monopoly. Then, without ever making it clear, he switches to an argument based purely on morality and not at all on law. For there is no antitrust violation when a single party attains a monopoly—like Cisco¹⁵ had for many years—without using wrongful means. Going back to the New York Times article, and adding badly-needed emphasis: “You’ve got to be big, and you have to be bad”. Any argument that complains solely about Google being big, as the Writer’s does, without ever mentioning badness, must be a purely moral one, as it has no basis in antitrust law.

If you are going to make an argument based purely on a moral code—your moral code—that has no connection with the applicable law, then Finding Fault suggests that you not begin by linking to an example of the application of the law. It muddies up the situation.

Mostly, the bloggers do one or two searches, and get a screenshot or two to drive traffic to their blog, as if nobody else could repeat the same search and see for themselves. Then they ignore Yahoo and declare Google or Bing the winner of their benchmark.

Finding Fault is hereby pleased to find fault and point out three major flaws that we have observed in numerous blog postings.

The First Small Flaw

The bloggers are judging Bing’s results by how Google-like they are. They do a Google search and a Bing search, then they look to see how similar they are. If Bing gives you essentially the same hits as Google does, then Bing must be good.²

Really?

Did we judge Google by how similar to Altavista it was?

That was a rhetorical question, but we will forgive you for answering no.

We judged Google by noticing that: Oftentimes we weren’t sure what we were looking for, but Google seemed to know.

That was why Google impressed. It made Google Googly. It was the quantum leap.

Yahoo eventually claimed to have caught up with Google, and now, apparently, so has Bing, some bloggers report.

But we see no new quantum leap. And without this new quantum leap, we expect no search victory over Google. The blogger benchmarks are close to worthless.

The Second Bigger Flaw

The second Googly property of Google was that it seemed to give us the truth, so far as anybody could define truth. The only people who seemed to complain about Google lying were the spammers and the search engine optimizers, which only made the rest of us even more sure that Googliness was Truthliness.

And so far, Google in its searches has not let us down. If we are looking for something bad about Google, we generally do a Google search, with the implicit belief that Google won’t lie to us even about itself.³

If you don’t realize how unique a quality this is, you haven’t been paying attention.

Yahoo, too, has adopted this philosophy.⁴

Finding Fault has seen no reason to believe that Bing at its core has this essential quality of Truthliness. Although it does look like it does. But we suspect Bing is faking sincerity.

Why do we suspect this?

Because, ultimately, Bing is just another of Microsoft’s faces. We won’t bore you by listing the many ways in which Microsoft and its officers and employees have acquired a reputation for insincerity.

This is why we at Finding Fault don’t trust Bing, and this is why we think the bloggers who evaluate Bing on the basis of a few search results now, today, are missing an important point.

They should be looking for reasons why Microsoft and its officers and employees won’t in the long run give us the sanitized truth according to Microsoft.⁵ If we see anybody find such a reason, we will let you know.

The Third Fatal Flaw

The biggest flaw in the blogger benchmarks of Bing and Google, the fatal one, is that they make the wrong assumptions, look at things in isolation, and don’t do real research.

They are looking at search as a service in isolation and as a service provided to customers. This is wrong, wrong, wrong. The service that Google ultimately provides is not search to its users, but rather, advertising to its customers. This is why Google engineers obsess over the number of kbytes sent to the browser, and over the grey-ness of text and the blue-ness of links, and over page load times greater than a small fraction of a second. Because, ultimately, the search service must attract viewers for the advertising that Google sells, and cause them to not use it just once, but to keep coming back, and keep clicking.

Bing’s background images and clever JavaScript will, for sure, attract attention, as they have already. The search engine Ask has been using clever JavaScript for some time to provide visitors with an ehnanced search experience. Unfortunately, while this pleases us we are in the mood, most of the time these enhancements just come in the way. And this is where Google has consistently won with its sparse design. The blog reviews we have seen of Bing seem to miss this point entirely.

And all this is why you cannot meaningfully do one or two tests, either yourself or with a random sample of users. Suppose you ask your random sample to test both Google and Bing and tell you which gives them better results. And suppose, just for the sake of argument, all of the users say they get better results with Bing. Would that mean anything?

Unfortunately, no. They are using Bing because you told them to.⁶

If they are doing something you told them to, then it doesn’t matter how good or bad it is—you don’t know what they would you do if you didn’t tell them what to do. If 100% of your sample of users prefer Bing when told to use it, but only 12% of them would use it if you didn’t tell them to use it, then your test results mean nothing.

In fact, just the users’ knowing that you are studying them may skew the results and make them unreliable.⁷

What you ought to do, Dear Blogger, is take a large enough random sample of users, let them do searches over a period of weeks, make them use Google and Bing and Yahoo and that other new one that sounds like Quill, and also Ask, and Mahalo, and a few others. This way they don’t know which search engine you are really evaluating, so they have no incentive to try to please you.

Collect a lot of data. Analyze it and publish results, so your subjects don’t feel they wasted their time.

Then throw it all away, because it means nothing.⁸

Now, after the experiment is over, you must surreptitiously monitor how these subjects do searches. This is the real, covert experiment. Which search engines do they now prefer to use when they think nobody is watching? And how has this changed since before the experiment?

And most importantly: Which ads do they click on?

The test subjects are now trying to please only themselves, not you, and their behavior now will tell you far more than their behavior during the overt experiment.

Those of you who want to take time off from mindless blogging and do some real experiments, we would love to see you publish your results.⁹