An important omission by critics of Airbus A380

Airbus' newest aircraft, the A380 (ignoring the A350 XWB), is often pitched against Boeing's latest aircraft, the 787 Dreamliner. Not directly, since these aircraft aren't comparable, but from a strategy perspective. It has been frequently said by analysts and others that Boeing has made a bet that flying passengers will migrate from hubs towards direct flights between points. In contrast, these analysts say, Airbus' development of the A380 assumes that passengers will migrate towards the model of flying between large hubs, and later taking connecting flights to the smaller destinations.

Who is right? We can't say today, and we'll have the answer only in a few decades. What we do know for sure is that the critics of the A380 aircraft [and these are frequently the same people who favor Boeing's strategy and predict its victory] have overlooked an important point - significant traffic will continue to fly between the most developed, most urban cities of the world.

There is no reason to believe that significant numbers of people will stop flying between these cities. For example, there's little reason to believe that the following routes will get less crowded in the future:

1. Paris - New York
2. London - New York
3. Mumbai - Delhi
4. Johannesburg - Cape Town
5. Hong Kong - Shanghai
6. Singapore - Kuala Lumpur
7. Paris - Montreal
8. London - Washington

Which aircraft can best serve these busy routes? Which aircraft can reduce congestion by lifting double or triple the number of revenue passengers in a single flight, all while dramatically reducing seat mile costs? It's the Airbus A380. Because I expect a high volume of air travel to continue between the most important cities of the world, I believe that the respective strategies, if any, by Airbus and Boeing might not amount to a zero-sum game. It's Boeing's alleged strategy that might turn out to be not so correct - Airbus' bet on hub-to-hub travel looks just fine to me. More so, since I haven't even figured the rise of new cities in the above thoughts. The A380's future looks safe.

The best list of tips and tricks for scoring high on the GMAT

This list is not finished yet. I'll add more points to it as they come to my mind.

A month before the GMAT:

1. Fall ill a month before the GMAT. I know it sounds weird, and is something that isn't exactly under one's control, but when I was down with common cold, fever and a sore throat about a month before my GMAT date, I was actually happy that I fell ill. August and September are the months in which it's easy to fall ill in the North India region. By falling ill a month before the GMAT, my body's resistance to common fevers got refreshed, virtually ensuring that I wouldn't fall ill again till my GMAT date. Of course, you can't afford to be ill close to the date your GMAT is scheduled.

On the day of the GMAT:

1. Carry some food stuff with you that's light, nutritious, and stimulating. Consume this during the break to both energize and stimulate yourself. An agile body and an agile mind can sail faster and more smoothly through difficult questions.
2. Before going to your test center, read some news stories to lubricate your brain's information processing systems (this is similar to warming your body before doing a hard workout at the gym).
3. Utilize the tutorial time (before the actual timed test starts, the GMAT software presents a brief tutorial) to make sure that your keyboard and mouse are working properly. Smoothly working keyboard and mouse save valuable seconds and also reduce anxiety.

Thoughts on Boeing's win of the KC-X aerial refueling tanker contest

I’ve been avidly following the Airbus vs. Boeing (vs. Antonov) saga for the KC-X tanker aircraft contest for many months now. I’ve gobbled every important development in this interesting competition, and frankly, I’m not too surprised by Boeing’s win. Although I did have a strong feeling that it’s obvious that America would award a prized military contract to only an American company, my feeling was definitely not so strong as to make me predict Boeing’s win on my newly-launched Predictions blog.

I have a conspiracy theory about Boeing's win, explained by the following three points:

1. It was already decided by the US that Boeing would be awarded this contract. In fact, it was obvious – it’s unthinkable that the Americans would risk depending on a European company for a strategic piece of their defense setup for many decades to come. How the political climate of the world is going to change in the next few decades, nobody can predict. America doesn't know what type of relation it’ll have with European countries in, say, 2035. So, the Americans effectively used EADS North America to squeeze the best deal out of Boeing. They knew they’re gonna award it to Boeing, but they just wanted Boeing to give them the cheapest deal possible. In the end, it’s a win-win-win outcome for America – its Air Force gets the cheapest possible deal, America gets an American company to build the product, and the Americans also get to humiliate the Europeans.
2. The Europeans knew all too well that Boeing would be awarded this decades-long contract. They participated will full (apparent) seriousness, lost like a hero (and gained sympathy as a result), so as to set a strong pretext for European governments buying only European products in their own subsequent procurements (presumably, by now they're sick of the US talking about free markets and consistently acting in a protectionist manner).
3. The US Air Force and Boeing were probably sleeping together since the very start of this contract. After EADS North America submitted its final bid, that bid was secretly disclosed to Boeing and Boeing covertly submitted a carefully revised bid, ensuring its victory (and also ensuring that any appeals or complaints by EADS to GAO, etc., are easily won by the updated Boeing offer).

In summary, it was all too clear right from the start that Boeing would be awarded this contract. All those discussions about cost, performance, technology, etc., ignored the most fundamental and most important fact – that Boeing is an American company (and also a jewel in America’s crown) offering the only truly American tanker, and that only an American-made tanker can deliver the maximum economic benefit to the American economy (no matter how much EADS North America trumpeted that its tanker is also truly American, the fact remains that it’s a European product that benefits Europe more than it benefits America, and that the money earned by EADS North America would’ve been sucked back by Europe). Why should an ailing America give billions of dollars to Europe when it can very well give that money to American companies (and thus American people), without compromising in any way? Why should America, a nation that uses every evil tactic to help Boeing sell stuff in the world, buy stuff for itself from a European company?

Equally importantly (perhaps even more), had the US bought a European product worth so many billions, the political fallout in the US would’ve been almost catastrophic. This contract wasn't merely a military contract. It was also an economic and political contract. Boeing's win can be summarized in three words - foreign risk, economic benefit, and sheer politics.

Through this win, Boeing has hurt Airbus in several ways:

• Humiliation.
• EADS’ desire to establish a factory in the US for its A330-based freighters is effectively killed for now.
• Boeing’s dying 767 product gets a fresh life, bringing tremendous commercial benefits (for example, the profits from this contract effectively act as a huge subsidy for Boeing, and will be used against Airbus).
• Post this win, a rejuvenated Boeing tanker product will compete more fiercely against the Airbus A330 MRTT for worldwide tanker procurements. 

Cleverly, and for obvious reasons, post this win, Boeing’s statements seem to have dropped the words America and American entirely (anyone who is aware of Boeing’s and EADS’ fierce campaigns for this contest is all too aware of Boeing’s overuse of the American-made tanker card in this supposedly fair contest).

A limitation of the "Like" button used by Facebook and others

Would you Like the following photo, if it appeared on Facebook or some other website? Some people would, but many probably wouldn't, because it feels a bit distasteful to Like this sad photo.

Would you bookmark this photo, if it appeared on some website? I hypothesize that a larger number of people would be comfortable bookmarking it than Liking it.

I believe that the process of hitting the Like button includes two implicit sub-actions - an expression of approval/liking for the item, and a desire to bookmark it, with the former action more dominant than the latter. It appears to me that it's for this reason that some people would refrain from hitting the Like button on a sad/sensitive/unhappy item, because although they were awed/touched by this item, they, perhaps subconsciously, do not want to indicate their approval for the situation depicted in the item, even if they are comfortable with the idea of bookmarking it.

In contrast, although the process of hitting a bookmark button also includes the same two actions, the sub-action of expressing approval/liking for the item is more subdued here, and thus merely bookmarking such a sad photo doesn't appear as an endorsement of the item - neither to oneself, nor to others.

Of course, as search engine experts at Google will probably concur, the Like button and my explanation of its mechanism (in contrast to the mechanism of a bookmark button) is extremely valuable from ranking perspective.

There's an opportunity for airlines to launch a new type of return ticket, something I call Nearby Return Ticket (NRT)

Traditionally, a return ticket from place A to place B allows a passenger to travel from A to B and back from B to A. It's no news that a return ticket between A and B is usually significantly cheaper than two individual tickets (A to B; B to A). However, the traditional concept of a return ticket assumes that a passenger disembarking at B will board, a few days later, from B. What if B is in a different country, and the passenger travels to a location C in this country using the road, and wants to fly back to A from C? The traditional concept of a return ticket doesn't apply here, and sometimes a passenger is left with no inexpensive choice.

The best way to explain this is with an example. An Indian wants to experience South Africa, and wants to take a flight from Delhi to Johannesburg. He wants to roam Jo'burg for 2 days, before making a quick hop by the road to Durban and Kruger (another 3 days). Upon returning, he then wants to fly from Jo'burg to Cape Town on a local airline, and experience Cape Town for 3 days, before taking a flight from Cape Town back to Delhi.

The current model of return ticket doesn't apply in the above scenario, a scenario which I believe is a quite common passenger requirement, and such a traveler will have to shell out a total of INR ~63,000 for the DEL-JNB and CPT-DEL flights on Emirates. However, both DEL-JNB-DEL and DEL-CPT-DEL flights on Emirates are significantly cheaper (INR ~44,000 each) than the DEL-JNB plus CPT-DEL combo on Emirates. Why? Because DEL-JNB with CPT-DEL doesn't qualify as a return ticket.

This is unfortunate, because from an operational perspective, DEL-JNB-DEL and DEL-CPT-DEL journeys don't give any extra benefit to Emirates, compared to the DEL-JNB plus CPT-DEL combo. That is, for Emirates a DEL-JNB-Missing-CPT-DEL journey by a passenger is not significantly different from a regular return ticket journey by the same passenger.

And so, I believe that there's an opportunity for airlines to launch a new type of return ticket called Nearby Return Ticket (NRT). NRT will allow a passenger to get the benefit of significantly lower overall price of a regular return ticket, as well as the flexibility of arriving at and departing from different destinations that are nearby. For example, Emirates will allow a passenger choosing the DEL-JNB-Missing-CPT-DEL type journey to fly for INR ~44,000 using NRT (or INR ~49,000, if Emirates wants to make a few extra bucks for providing the NRT service), instead of INR ~63,000.

After all, when it doesn't make any/much difference to an airline from an operational perspective, there's no reason why an airline shouldn't be willing to give a passenger the fare of a regular return ticket on NRT too.

LocalMaps or MicroMaps - an idea for maps at the "micro" level

In Aug'09, I went to the IGI Airport (DEL) to pick a friend who had flown to Delhi, and who had a few hours before she took a connecting train to her final destination. After parking my car at the parking, I wondered where the Arrival terminal might be!

I digress for a moment.

What does one do when one is in Gauteng and wants to drive to Kruger National Park? Simple, one feeds Kruger National Park into one's GPS device and the device guides one to one's destination. This is an example of mapping at the macro level.

I think we also need maps at the micro level - maps that work inside large airports, large entertainment parks, large shopping malls, etc. I name such a feature LocalMaps (or MicroMaps).

Back to IGI. Had there been such a feature, my phone would've sensed that it's inside an area that supports LocalMaps, and would've automatically asked me to either choose (from a list discovered by my phone in that zone) or search for a person, place or product (or service) I'm interested in (in that zone). As soon as I would've started typing Arri, the phone would've suggested Arrival Terminal, and choosing this option would've either displayed (or spoken) the direction in which I should start moving (towers in this zone would've helped my device to pinpoint its location within the zone, and to suggest directions).

I believe LocalMaps can save people time when they're inside large, well-defined areas. Looking for a specific Axe item at a huge shopping mall? Type Axe into the search and you'll be told not only about the availability (and price, etc.) of the item, but also the direction in which you should move to get to the item. Order the phone to check if a friend is inside the same mall, and you'll be told both the answer and the direction, if applicable, to get to him/her. Inside a Pick n Pay store but unable to locate a particular product? Worry not, for you can query the location of the product.

The world needs a "SearchMyStuff" service

I've created content on numerous Web properties:

1. Emails in Gmail, Hotmail and Yahoo Mail
2. Chats in Gmail
3. Photos on Flickr, Orkut and  Picasa Web
4. Tweets on Twitter
5. Comments on Facebook
6. Bookmarks and Videos on YouTube
7. Posts on GMAT Club
8. Questions on LinkedIn Answers and Yahoo Answers
9. Blogs on Blogger
10. Scraps on Orkut
11. Contacts in Google Contacts and Yahoo Contacts
12. Documents in Google Docs and Zoho
13. Files on SkyDrive
14. Events, etc., on Google Calendar
15. Etc.

I know no service that allows me to conduct a search on all the content created by me, lurking on various Web properties owned by different companies. And I think that the world needs such a service. A service that allows me, for example, to conduct a search for the term Chrome, and retrieve all the content related to Chrome that I've created online. Creating such a service will be a tall order, and will involve APIs, business-interests, privacy, security, etc., but that doesn't mean that such a service won't be worth the effort.

Building Google-like search engine using Google Search Appliance

The idea below is based on certain assumptions made by me. It seems that this idea is workable, unless this "misuse" was foreseen and features/measures were put in place to thwart such a use. I first wrote this idea in Jun'09, but never posted it on the Web.

Online news media is abuzz about the release of an updated version of Google Search Appliance (GSA) - version 6.0. Among the many updates to this version, one feature has intrigued me the most - the ability to index billions of documents (using clustering). (1)

To understand a possible implication of this, it's important realize that Google's search algorithms are one of its most important pieces of IP. What makes Google Google is its secret sauce - the algorithms it uses to rank Web content. Everyone can crawl the Web, but it's the relevance-determining algorithms that give Google much of its competitive edge over rival search engines such as Yahoo Search, Ask.com, and Bing. (2)

It's also known that GSA uses Google's ranking algorithms to rank indexed content. (3)

We also know that the ability to crawl and ingest the Web is not a major source of competitive advantage for search engines. Even a simple program such as HTTrack can do a relatively decent job of downloading a website by jumping from URL to URL. The process that HTTrack uses to crawl a website is similar to how contemporary search engines crawl the Web. It should be easily possible to configure (or customize) HTTrack to crawl the Web, rather than just a website. (4)

Click image to enlarge

What all of this leads me to believe is that it should be possible to cluster multiple GSAs to create a pseudo-Google - a search engine that uses Google's secret algorithms to rank the Web, but is powered by a cluster of GSAs. If this is indeed possible, it'll make it super-easy for clever entrepreneurs to launch new search engines that provide high-quality results.

References:

1. Google Search Appliance Now Can Index Billions of Documents, PC World, June 2, 2009
2. The Big Cheese - Powerful Version Of Google Search Appliance Can Grow Exponentially, TechCrunch, June 2, 2009
3. Google Enterprise Search - Search Appliance
4. Google Search Appliance on Wikipedia

By not serving environment-optimized binaries, Mozilla is denying Firefox users an optimum experience (and hurting Firefox)

To summarize, I've observed that Mozilla's practice of distributing official binaries that aren't optimized to take advantage of new features and other improvements in modern CPUs (and possibly operating systems) results in a relatively inferior user-experience for Firefox users, ultimately hurting Firefox (and thus Mozilla).

More specifically, the binaries distributed by Mozilla use the Greatest Common Divisor (GCD)/Highest Common Factor (HCF) approach (this approach is mistakenly labeled Least Common Multiple (LCM) by many, many people) for compilation, resulting in binaries that don't make use of relatively newer instruction set extensions such as SSE, SSE2, etc. Additionally, the compilation settings used by Mozilla result in binaries that are optimized for size, at the expense of performance. The result is a non-insignificant deterioration in Firefox's performance (I've observed non-insignificant performance difference between non-optimized and optimized versions of Firefox on popular benchmark tests such as V8). This difference is visible to, and hurts those Firefox users whose systems allow Firefox to provide only slightly less than just-acceptable performance (such users would probably get slightly more than just-acceptable performance with optimized builds).

Mozilla's practice also denies better experience to those users who have faster connections (so they don't mind downloading a larger installer) and newer CPUs (so their systems are capable of performing better).

One of the computers I own and maintain is exactly such a system. Chrome runs smoothly, whereas the official version of Firefox 3.6.8 struggles, as if it's at a loss of breath. Switching to an optimized version of Firefox has improved performance noticeably, and has prevented me from switching full-time to Chrome on this system. What my own example shows is that there's a bunch of users whose machines don't provide them with a satisfactory Firefox experience because Mozilla doesn't optimize Firefox for performance. Is this number of such users large? Is it so large that a switch by these users from Firefox to the undeniably-snappier Chrome will materially hurt Firefox/Mozilla?

How can Mozilla solve this problem? The following options come to my mind:

1. Fat installer, normal install: An installer will contain multiple versions of binaries, intended to cover the most-common types of systems in existence. During installation, the installer will perform system capability assessment, and those binaries will be installed which are best suited to the environment. This solution will most likely result in a decrease in the number of downloads, due to increased installer size.
2. Normal installer, normal install: The installer will carry only the GCD/HCF code (thus keeping it small-enough for most people to download), but will perform system capability assessment during installation. If it determines that more optimized binaries can be run on this system, it'll either quietly pull those binaries from the Web, or will ask the user for his consent before pulling the more optimized files.
3. Separate installer, normal install: Since only power users are expected to be bothered about using optimized binaries, Mozilla should make available official but optimized versions of Firefox on Mozilla.com, albeit hidden from general users (so they don't accidentally download an installer which their system might not support). The installer, even for an official but optimized version, will perform system capability assessment to ensure that installation is being conducted on a supported system. This approach has the disadvantage that it serves only power users. It fails to provide the benefits of optimized Firefox to the masses.
4. Regular installer, deferred optimization: Install the regular Firefox, and on first launch, it'll itself perform system capability assessment and ask your permission to optimize itself by downloading and installing optimized versions of files that are best-suited to the current system. Alternatively, it'll quietly download and install these files, and the effects will be visible after a restart.
5. Thin installer, normal install: This is the approach that I favor the most. A tiny installer that doesn't carry any binaries will perform system capability assessment and submit the results to Mozilla.com, which'll supply the installer with the files best-suited for the system. This type of installer - Chrome also uses a thin installer, although I'm unaware if it pulls system-specific binaries - appears to be suitable for both amateur and power users.

Related content: Swiftweasel; Swiftfox

We remember "random" sequences of alphabets much better than random sequences of numerical digits. Why?

Ask me to remember the following three words, and I guess I should be able to (quickly):

• Nipitto
• Quasquam
• Zyrectica

Ask me to remember the following three numbers, and in all probability I won't be able to:

• 5686990
• 21342137
• 984631237

Why the difference? There are 26 alphabets in the English language, compared to only 10 numerical digits. And yet we're able to remember thousands of words - most of which are basically random sequences of alphabets (it's we who've given meaning to these random sequences) - but we have trouble remembering more than a few phone numbers. Why?

On a completely fundamental level, alphabets and digits are nothing but visual representations of certain values and sounds, respectively. And words are formed by combinations of letters, while numbers are formed by combinations of digits. Why then is an average human able to remember thousands of words in his memory - sometimes even from multiple languages - but he can't remember too many numbers?

I suspect that sounds are a reason. Since all words in the English language have a unique pronunciation (in the form of a sound), it's possible that we're better at remembering sounds. Numbers - such as 32748746 - don't have any single sounds associated with them, and this might be a reason why numbers are much tougher to remember. The concept of sounds also seems to be applicable to the examples I've given at the start of this blog post - all of the three words can be converted into sounds, which're much easier to remember than pure sequences of alphabets. It's possible that we remember sounds, and later use a sound-to-text engine present in our brain to convert the sounds to sequences of alphabets. It's also possible that each sound can be stored as a single data-chunk, thus requiring less memory compared to a sequence of numbers.

Case in point - remembering the following words is as tough as remembering numbers, because these words can't easily be converted into easy sounds (absence of vowels):

• Zvtrxttr
• Nytrpqstj
• Bcddfktr

These aren't words anymore - they've become as tough to remember as numbers!

Related content: Human brain could be storing & retrieving information as 'related blocks'

Related content: Human brain's information-retrieval system is imperfect (apparently)

How I keep my new computer safe from malware (spyware, viruses, worms, etc.), network attacks, phishing websites, spam, etc.

I use the following policies, tools and techniques to safeguard my new computer from all sorts of privacy and/or security breaches:

1. DNS level: Use of OpenDNS resolution service ensures that malicious addresses are blocked at the DNS level. Even the basic/free version of OpenDNS provides a decent level of customization, including allowing blocking of specific website categories.
2. CPU level: The CPU in my new computer supports DEP at hardware level.
3. BIOS level: The computer's BIOS includes a "Virus Warning" feature, which warns whenever an attempt is made to write to the MBR of the hard disk. Additionally, enabling passwords in the BIOS helps prevent unauthorized changes to BIOS settings, and unauthorized entry into the system.
4. OS level: Windows 7 provides me with these helpful features - Action Center, ASLR, Automatic Updates, DEP and UAC. Additionally, the computer is normally run with standard user privileges, rather than administrator privileges, and the user accounts are password-protected.
5. Firewall level: Use of a good firewall such as the one by COMODO keeps the machine secure from unauthorized access through network. Additionally, COMODO's firewall includes useful features such as Image Execution Control, Sandbox, etc., which protect against unknown/untrusted executable code. I would've liked it if COMODO's firewall included its own feature to block malicious websites.
6. Anti-malware level: An anti-malware application such as ESET NOD32 Antivirus (or Norton AntiVirus) used in real-time protection mode provides effective protection against all sorts of bots, rootkits, spyware, Trojan horses, viruses, worms, etc., and also mandatorily scans all Web traffic. Behavioral-analysis/heuristics features enable such an application to detect new/unknown malware, while the unique Download Insight feature of Norton AntiVirus provides vital crowdsourced information about individual files. Frequent updates ensure that definitions and modules remain up-to-date. NOD32 Antivirus also blocks webpages known to include "potentially dangerous content".
7. Additional anti-malware: Windows Defender and Windows Malicious Software Removal Tool - both provided by Microsoft - provide additional protection against common malware. Finally, a monthly scan with the free and nice Windows Live OneCare Safety Scanner both checks and optimizes the system.
8. Web browser level: Use of Chrome provides these benefits - it's secure by design, it includes its own anti-phishing feature, it runs Flash inside a sandbox, and it updates automatically and forcibly. Additionally, I use HTTPS (whenever supported); I've even made bookmarks with HTTPS prefixed, so I'm directly taken to secure pages (whenever supported).
9. Search engine level: Google's search engine includes its own warnings against malicious results, and thus, use of Google provides an additional layer of protection.
10. Frequent, regular updates for everything: I've configured the system so that everything - anti-malware applications, browser, drivers, operating system, etc. - updates frequently and regularly. I manually update the applications which report that updates are available.
11. Remove everything unnecessary: Removing everything that doesn't contribute is another policy that increases security. Disabling really unneeded services (without breaking essential functionality), disabling unneeded ports, uninstalling non-essential components of the OS (Internet Explorer, etc.), etc., are examples of this policy.
12. Functional-but-secure settings throughout: Applications, devices and the OS include many configurations options which allow a trade-off between functionality and security. Some of these include AutoPlay/AutoRun, router settings, USB-related settings, etc. Striking a good balance throughout is essential to increasing security without breaking functionality.
13. Stay abreast of whatever's new: Stay aware about the latest developments and discoveries, and adopt whatever's good. Google's Chrome is a good example. It's newer and safer than Firefox. Someone who uses Firefox and doesn't stay updated about new developments would've missed Chrome entirely, thus making him a little less safe (than he could be).
14. Avoid paying too much price: Additional protection is possible by using additional tools such as Malwarebytes' Anti-Malware, McAfee SiteAdvisor (or Norton Safe Web), SUPERAntiSpyware, Web of Trust, etc., but I don't use these as I don't see as much benefit from their use as is the price to be paid, especially since the above measures have already kept me safe for years.

I use the above methods (except the 14th point) to safeguard my new machine. One good aspect of the above policies, tools and techniques is that all of these need to be deployed and setup just once. Everything works automatically thereupon, except for occasional manual checks/updates. The few hours it takes initially to set things optimally have paid off nicely for me - I've been breach-free and malware-free for years now!

Google isn't giving away Android for free. Steve Ballmer, was wrong

I can't recollect the article or the video in which I read/saw Steve Ballmer make a claim which meant that because Google is giving the Android operating system for free, Android can't be of as much quality as our paid Windows Mobile operating system can be. Because we charge for Windows Mobile, we can put a lot more investment into the product, compared to Google, which can put only so much investment into its free OS.

I had thought this when I had read those words of Steve Ballmer, and I thought this again when I read a few statements made recently by Eric Schmidt.

Google is not giving away Android for free!

Google has strong reasons - both tactical and strategic - to invest in Android and ensure that it achieves a high adoption on smartphones and other devices from multiple sellers. What Ballmer said appears to imply that Google has neither an incentive to fund Android's development, nor cash (because Android's free, the money has to be pulled from elsewhere inside Google, but doesn't come from the Android product itself).

Incorrect. Google has both.

1. Cash: Android does generate net revenue for Google. So much, that it's more than enough to fund its development (as quoted recently said by Eric Schmidt. Unfortunately I don't have a link to that news story either). Ballmer's assertion that Android is a free product rests on the incorrect assumption that only the upfront taken revenue (from end-users or hardware-sellers) makes a product a paid product. His claim apparently overlooks the fact that Android-based devices generate advertising revenue for Google, and this revenue is large-enough to more than cover Android's development. Further, Google doesn't have to split advertising revenue generated on Android, unlike the split it does with Apple (for the revenue generated on the iPhone, iPod touch, etc.). Android might come free to the hardware-makers and hence to the end-users, but from Google's point-of-view, it is a paid product - it's the advertisers who pay for it! How do Google's wonderful services such as Web Search, Gmail, etc., come for free to the end-customers? Somebody must be paying Google, after all. It's services such as AdSense and AdWords which allow Google to make both Android and an array of wonderful online services free for both end-customers and hardware-makers. And who funds AdSense/AdWords? It's the advertisers! It's this indirect revenue-stream that Ballmer apparently overlooked.
2. Incentives: This aspect can be understood better if one believes that market share and revenue are distinct goals. Generating revenue aside, the Android operating system ensures increased exposure to, and adoption of Google's products and services (by being both "defaults" and tightly-integrated). This makes both tactical and strategic sense, and appears to be a sufficient-enough incentive (net revenue incentive aside) to encourage Google to fund Android's development and adoption.

In summary, it's in Google's interest to make sure that Android gets heavy adoption - which means Android must out-innovate rival operating systems, which means the Android team works as furiously as the Windows Mobile team - a contradiction to what Ballmer implied.

A favorable side-effect of increased adoption of Android is that this helps Google to contain and hurt both heavyweight and upstart competitors, including Microsoft.

Steve Ballmer, was wrong.

Why is the combined ticket price of three legs of a particular flight lower than the combined ticket price of only the first two legs?

Perhaps I don't understand the economics of civil aviation that well, but what I just saw has puzzled me. A lot.

Ticket price for a commercial flight from DEL to JFK (via HEL) on Finnair is lower than the ticket price for a journey from DEL to YYZ (via HEL and JFK, with the journey from DEL to JFK operated by Finnair, and the journey from JFK to YYZ operated by American Airlines).

(Click to enlarge)

(Click to enlarge)

This weird pricing looks even more alarming because an American Airlines flight from JFK to YYZ is priced at INR ~8,500, or a good over 27% of the price of the DEL-HEL-JFK flight operated by Finnair.

(Click to enlarge)

This effectively translates into: A+B>A+B+C, where A, B and C are positive real numbers. How's it possible? Who's losing money here, if anyone?

iTunes on iPhone/iPod touch is significantly different from the Macintosh/Windows version of iTunes on a fundamental basis

This is what I observed during my past few weeks I with an iPod touch.

 iTunes for Windows (source)

The Mac/Windows version of iTunes is a place where you do at least the following things:

1. See your complete media collection (assuming you use no other media management software), and choose what music/movies/videos you want to play
2. Playback any of the content you can see in iTunes
3. Discover new content, using both catalogs and search queries
   
4. Acquire content, by either downloading free content or purchasing paid content
5. Sync content (including applications, et al.) with your portable Apple device
6. Etc.
   

In a nutshell, iTunes on the "desktop" operating systems is a one-stop place, which allows you to do everything that you can possibly want to with content (keeping Web browsers, YouTube, etc., out of the picture for the purpose of this post)

iPod touch (source: Apple)

In contrast, on iOS devices such as iPod touch, there are three distinct applications to do what all could be done in iTunes alone on a Mac/Windows machine.

1. See: Use 'Music' or 'Videos' applications. You can't see your media collection in iTunes!
2. Playback: Again, you can't playback locally stored content inside this "iTunes"!
3. Discover: Use 'iTunes'
4. Acquire: Use 'iTunes'
5. Sync: iOS
6. Etc.

And after this analysis I'm able to understand why I kept opening the iTunes application on the iPod touch every time I wanted to consume the stored music/videos.

Google should launch a contextual advertising product for PDF files

I've wondered for years why we don't have an AdSense for PDF product from Google (or others). Why can't I embed code from this hypothetical AdSense for PDF product into a PDF ebook I've written, and distribute it for free on the Internet. This hypothetical PDF shall be secured in such a manner that it shall open only when there's a Web connection available, so that ads can be pulled from Google's servers in real-time.

I believe that an AdSense product for PDF files can bring a revolution in the commercial use of the PDF format. It shall allow both experts and semi-experts to publish individual articles as well as full-blown ebooks in the form of well-protected PDF documents laden with advertising code, without worrying about unauthorized copying of their work. In fact, under this model authors will encourage free-distribution of their PDFs to more and more people.

Such a product shall allow website owners to publish select documents in PDF rather than HTML. Use of PDF shall provide the following benefits to the publisher:

1. Setting permissions using industrial-strength features of PDF
2. Not losing out on the per-document ad revenue available in regular Web documents
3. Providing the usual benefits of PDF files - such as a consistent layout, paper-like readability, etc. - to the users

Because the PDF format is such a tightly-knit format, compared to the relatively loose HTML format, I have a high degree of confidence that a well-designed contextual advertising product for PDFs will usher a revolution in the way PDF files are used. However, I also have a strong belief that a prerequisite for such a product's success is that the use of PDF files be made as seamless as the use of HTML documents currently is. Google's integration of PDF support in Chrome is a good step in that direction.

For Google, such a product can mean a new source of cash - whilst making minimum investment - to feed its desire for continued growth.