Software

Squeezebox Boom

A while ago, I wrote about how I would love a Squeezebox that is integrated with speakers.  This would make it a perfect solution for our kids rooms.

Squeezebox/Logitech just announced the Squeezbox Boom, which is exactly what I want.  With this, we don't have to deal with playing cds that skip because the kids have scratched them.  I am definitely ordering one of these.

What Did You Say?

Like a few million other people I recently bought an iPhone 3G. But unlike a few million other people I bought TWO of them -- one for my young and lovely wife. That puts me in the rare position to actually speak from experience about a current networking issue: what's the deal with these iPhones? Is it the phone or the network that is causing problems? And the answer is: both.

Our experience with the iPhone 3G is not good, though not terrible. It is a fantastic device, if flawed. The main flaw is the phone, rather than the iPod Touch bits that comprise the rest of the unit. Voice service is not good, calls are dropped, and the iPhone won't work places where my old Nokia N80 easily did. But my iPhone, oddly enough, works better than does my wife's. This variation is how I know at least part of the problem is with the phone, not just the network.

But the network sucks, too. We switched from Verizon (I know you can't use an N80 on Verizon, smarty-pants -- I have more than one cell provider) which claims the best network but had the nagging problem of delivering the odd voicemail 7-10 days late. Verizon claims never to have heard of this problem but ask a few Verizon users and they'll confirm. Now that we are full-time on AT&T we might blame the lousy service on the phones except I also got a Samsung AT&T phone for Mimi, my mother-in-law, who is on our family plan. Her service sucks, too, so it is not just the iPhone.

I'm sure AT&T has oversold their network. You can tell because the worst service of all is from one iPhone to the other. If the call doesn't spontaneously disconnect half the time you often still can't understand what the other person is saying. Service is somewhat better going to landlines or other mobile providers.

I'm sure AT&T will fix this eventually but I don't like being treated this way. No wonder they are so hot to keep that iPhone exclusive, since half the iPhone users I know would jump to T-Mobile if they easily could.

Last week's column about the population of CCIEs and global development raised quite a ruckus -- a word I include to confuse the non-native English speakers who saw last week's column as discriminating against them. We can argue a bit about the numbers and their meaning, but I think it is fairly obvious that: 1) Cisco dominates the Internet core router business, so this is a real issue no matter what your language, and; 2) CCIEs are NOT just network techs. It is an extremely difficult certification to get and typically costs in the neighborhood of $30-40,000 by the time you are finished, sometimes a lot more.

While it may be patently obvious that China and Korea will be more important 30 years from now than India and Japan, that wasn't my point. Anyone can express that opinion. What I was trying to do was to show a reason WHY that might be the case as evidenced by this CCIE data. Why shouldn't India be just as successful as China? Their populations are comparable and they both have good educational systems with large numbers of graduates. They both value science and technology. India even uses English as one of its official languages. Both have booming economies with plenty of room for growth. Well this CCIE analysis gives one empirical reason why it should be so. While the Indians are developing their IP expertise, the Chinese are developing their IP networks, simple as that.

Another reason to talk about this subject is because there is far too little actual thinking on the Internet these days. The blogosphere is full of opinions but not very much solid discussion of why things are the way they are. Agree with me or not -- I don't care -- but I'll always make you think. The whole point of this column is getting people to think and discuss.

Now to the problem from last week of how YouTube and similar video services can better appeal to advertisers. I foolishly thought last week's tease might coax a few bucks out of the bushes at a time when I could use them, but no. So I'll reveal my solution anyway, even without a reward.

YouTube would love to make lots of money from ads that would play before its 100 million videos per day, but they have had difficulty appealing to traditional advertisers, not because of the quality of the viewers but because of the quality of the videos, themselves. There is a huge variety of content on YouTube and while advertisers are willing to stretch a bit in what they'll sponsor, they are afraid of making a mistake and sponsoring the wrong videos, like those that contain nudity or other objectionable content. Short of watching all the videos, how do we best avoid this problem? That's what has Google scratching its GoogleHead.

It all comes down to the quality of the metadata -- the data describing each video. I think the answer is obvious and is composed of three parts: 1) notification, 2) structure, and 3) standardization.

I'm probably listing these backward, but I want to get notification out of the way first. Whatever system YouTube chooses to manage its metadata won't be perfect. There will be errors -- the Internet equivalent of Janet Jackson's "wardrobe malfunction." So there should be a facility for viewers to notify YouTube if they feel that there is a significant mismatch between the likely target of the content and the target of the accompanying ad. No condom ads before Dora the Explorer clips, for example. No bacon commercials with vegan cooking shows. Having a way to report such errors will diminish them in both number and importance.

Next comes structured metadata. It floors me that YouTube doesn't enforce this already. If you ever took a journalism class you learned that the first paragraph of any news story (called the "lead" or "lede" -- same pronunciation) is supposed to answer the questions who? what? why? where? when? and how? YouTube could use a form for each video submission that used these categories, possibly minus the highly suggestive "why." To submit your video you have to fill in all five blanks. Leave any blank unfilled and your video bounces. These five metadata categories really ought to cover the gamut of describing most any video. If they don't, or if they are improperly used, then we are back to notification and getting users to help fix our mistakes.

Finally there is standardization. Even within the five structured metadata categories there can be great variation in the meaning of the chosen terms. That's why we need to standardize those terms. Rather than reinvent the wheel, I would rely on the best current system of standardized metadata, which is Wikipedia. Behind each structured entry would be those Wikipedia terms that would seem to be appropriate, along with links to their definitions, just to be sure.

With this system it would take a minute or two longer to submit each video, but what the videos were about would be much clearer to viewers and advertisers alike. Paris Hilton is a "who." The Paris Hilton is a "what" or perhaps a "where," though the system would probably force them to be separated. Either way you end up with a hotel if you want a hotel and a sex video if you want a sex video.

Now it's time to recharge my iPhone (again).

links for August 27, 2008

• DeltaCopy - Rsync for Windows
• Linux.com :: Making secure remote backups with Rsync
• cwrsync - Rsync for Windows | ITeF!x

Windows Media Center

I am thinking about installing Microsoft Windows Vista Media Center on our Mac mini, espcially since our Xbox 360 is doing everything that the Mac mini used to do.  I am planning on keeping the Mac mini in the closet, near our Windows Home Server, and then I would view this content on the Xbox 360 through Windows Media Extender.

There are several things that I would be able to do with this setup:

• Install TVTonic to have access to "Internet TV" and video podcasts  (The 2008 Olympics is available this way)
• Play ripped DVDs that are hosted on the Windows Home Server (1, 2)

When I get this set up, this would allow me to remove our 300 disk DVD changer, as I can select any of the movies that we have through a nicer UI. 

Also, potentially this would allow us to get rid of our cable TV.  This would save us a lot of money. I think that the vidoes available over TVTonic will be fine, when the kids want to watch something.  If there are current TV shows that we want to watch, we can buy them through the Xbox Live Marketplace.  Buying each episode individually would still be cheaper that paying for our monthly cable bill (especially since we wouldn't have to rent the extra cable boxes)

Leading Men

Leading indicators are measurements that change over time and suggest future trends for important second-order results like population growth and economic development. Economists in particular are often looking for indicators that have been known historically to lead the overall economy. If unemployment goes down, for example, it is a good bet that shortly thereafter income will rise and the economy will improve. It's for this very reason, then, that economists and Wall Street fund managers are always looking for newer and better leading indicators. But such indicators needn't be limited to the economy: they can apply to technology and technical culture, too, which has its own feedback loop to economic development. My friend George Morton, who figured this all out, says that by knowing the right numbers to look at we can have a good idea what countries will be leading in technology -- and presumably in economic development and power -- in the years ahead. The measure George likes is the number of Cisco Certified Internetwork Experts or CCIEs.

The CCIE is Cisco's top certification category and VERY hard to earn. Being a CCIE doesn't mean you have Len Bozack on speed dial, but it might as well. Cisco products dominate the Internet and CCIEs are Cisco gurus, so if you are serious about the Internet as a nation you'll have CCIEs hanging about, or that's the theory. Conversely, if you just talk a good game as a country with technological aspirations, maybe you won't have many CCIEs at all -- maybe none. It's one way to determine who the posers are.

Cisco publishes the total number of CCIEs and their geographical distribution four times per year and George used the Internet Archive to track down the last nine years of data to look for trends. All of this is behind one of this week's links.

Where I took a step further was to divide the number of CCIEs into each country's population, then do the same for each country's Gross Domestic Product and correct for widely varying populations and states of economic development. For a baseline, then, the U.S. has at present 5,863 CCIEs, which is 1.947 CCIEs per 100,000 population and $2.2 billion of GDP per CCIE.

It is logical to assume that nations with adjusted numbers that exceed those of the U.S. for CCIEs per 100K are Internet up-and-comers and ought to fare well in the decades ahead. Beyond the population statistic, countries that have significantly less GDP per CCIE are those that would seem to have made networking a national priority. Countries that are significantly behind the U.S. on one measure or another are just that -- behind the U.S. -- which is not good.

Here, then, are some of the numbers I calculated:

Canada, not surprisingly, is similar to the U.S. with 2.2 CCIEs per 100K and $2.45 billion per CCIE, as is the UK with 1.5 CCIEs per 100K and $2.12 billion in GDP per CCIE. Ireland is very similar to the UK with 1.48 CCIEs per 100K and $2.95 billion in GDP per CCIE. The really interesting European numbers to me come from Germany, with 0.74 CCIEs per 100K and $4.3 billion per CCIE, and France, with 0.36 CCIEs per 100K and $8.11 billion per CCIE. Both of these countries appear to be underinvesting in network technology, with France especially lagging.

Latin America is dramatically behind Europe and North America, though I found it interesting that Argentina, with 0.17 CCIEs per 100K and $8.94 billion in GDP, is 50-100 percent ahead of both Mexico and Brazil.

It is especially interesting to compare India with China and Japan with South Korea. India has 0.036 CCIEs per 100K to China's 0.22 per 100K -- a 7X differential -- while India has $10 billion in GDP per CCIE to China's $3.3 billion. There is no doubt that there is plenty of network expertise in India, but these numbers show that expertise isn't making it out of the technology centers to the rest of the country. China, on the other hand, is developing its IT infrastructure much more broadly. This also doesn't take into account the simply huge numbers coming out of Hong Kong, where there are 3.3 CCIEs per 100K and $1.13 billion in GDP per CCIE -- numbers that might properly be added to the rest of China in some accounts.

Japan has 1.23 CCIEs per 100K to South Korea's 1.9, but the significant difference between these two countries is the $4 billion per CCIE in GDP for Japan compared to $1.28 billion in South Korea, which is clearly investing massively in network infrastructure.

Looking 30 years into the future I think it is clear that the regional leaders will be China and Korea, NOT India and Japan.

Israel has numbers very similar to Korea with 1.43 CCIEs per 100K and $1.4 billion in GDP per CCIE, which is more than double on both scales that of the other Middle Eastern leader, Saudi Arabia, with 0.42 CCIEs per 100K and $3.2 billion in GDP per CCIE.

And don't count out corporate haven Bermuda with its five total CCIEs and its population of 66,000. That's 7.5 CCIEs per100K and $900 million in GDP per CCIE.

"When looking at the countries a key element for IT is the English language," said George. "Yes I know Harvard still publishes diplomas in Latin. Where English is a first language, or an important second language is key to the number of CCIEs. Sorry C, PERL, JAVA, Cisco IOS, Basic, &t. are all English context. The Moore's CCIE Law also brings into question the ability of countries to attract IT capital with open or closed network infrastructure.
The last point -- over 50 percent of all CCIEs in the US work for Cisco. In Mexico, and India I bet it is over 50%. Both are very large call centers for Cisco Support. Morton's CCIE Law = Moore's Law accelerated by the technological support available to exploit Moore's Law. The degree of acceleration is measured by the available number of CCIEs. It's Newton's Second Law: F=ma, (or Force = Moore's Law * CCIEs)."

What we're talking about here is the future of world power in this century, but I can understand if your eyes have glazed over a bit. Maybe something more interesting would be the big question reverberating through the hallways at Google: why the heck can't they make any money from YouTube?

This is, on the face of it, the same question that eBay must be asking about Skype. Both were fairly large investments yet neither is contributing significant revenue to the parent company. Poor Skype can't feed us ads before or after every phone call, especially since the only way to have contextual data to make those ads more valuable would be to listen in on the calls -- a no-no. But for YouTube the problem is different because it ought to be fairly easy to figure out what a video is about then sell ads against that metadata.

Easier said than done, my friends. The problem with YouTube and advertisers is that the nature of the videos, themselves, is too varied and the metadata too easily wrong. A hotel chain that might well want to advertise before a video of the Paris Hilton, for example, might be extremely reluctant to advertise before a video OF Paris Hilton, yet from the perspective of metadata both are extremely similar. This is a HUGE problem for Google and all the other streamers of user-generated videos, which leads some people to believe that amateur night will eventually end and the Internet will return to being like most television -- a place for predominantly professional video.

What's funny about this is I think I have the problem figured out. The answer seems obvious to me, but I'm not sure Google would even listen if I told them my answer.