All Things Techie With Huge, Unstructured, Intuitive Leaps

Toshiba Satellite Lap Product Review

I just purchased a new Toshiba laptop because the old HP dual core just died. Actually it is in the process of dying. It is having mini-strokes. When I boot it, it doesn't boot. It doesn't POST. It just sits there and the HDD disk light flashes orange. This usually flashes orange when the accelerometer detects that the laptop is moving and parks the disk heads. After a while, the thing boots up, runs for awhile and then dies again. It is a crapshoot as to whether it boots and how long it takes.

So onto the Toshiba. It has a dual Core i3 genuine Intel processor and the hard disk is half a terrabyte. It should scream, but it comes with Windows 7 so it just marches. Nothing like an old XP for speed. (Did I mention that Microsoft sucks?).

Well, I am generally pleased with the laptop except that for human factors engineering, it is a FAIL. The keyboard is all wrong. When I am typing without looking and want to hit the shift, I hit the control. The shift on the left is way too small, and way too high. The return key is small below but enlarges into the row above, where the little pinkie on the right hand can't reach it. It is a pain to type on.

Other ergonomic things bug me. It takes me forever to find the little button to eject a DVD or CD. The delete button isn't handy. The track pad is over too far left. It is not centered like on any other normal computer.

The other thing that is slightly ticking me off, is that it regularly makes weird noises like a chirping for a few seconds then quits for a couple of minutes. I have to debug this one.

So while the computer seems okay, it is a bit of a pain to operate it. I have to learn to live with it, because Future Shop, where I bought the thing, had the wrong price on it, and they honored the tag that said it was $120 less that it should have been.

Update: The keyboard is huge pain, but I love the way that it runs cool compared to the HP.
And it really boots up fast and shuts down fast unlike the HP. I used the like the HP, but this
one is growing on me. The keyboard is still HUGELY frustrating though. But there are no more cooked skin marks on my thighs from an overly hot laptop.

Update 2: The speakers on the Toshiba sound tinny compared to the HP laptop. Still having difficulty with the keyboard, and now the sound sucks. But I really like how cool it runs. The HP actually left burn marks on my lap when I took the computer outside to program on the balcony overlooking the water.

Mechanized Attack Detection Algorithms

I was on a CNN sub site and they were talking about security and threat of attack. One of the types of attacks that they were profiling, was a mechanized terrorist attack by robot or remote control machines.

This got me to thinking. I am a technical architect for software, and its about time someone starting putting down some thoughts about algorithms for detection of mechanized attacks. Here are my thoughts on the subject.

1) The "Does This Thing Belong" algorithm. This algorithm is for machine monitoring of real time CCTV images. A camera aimed at a street would have a shapes learning module where each shape that went by the lense would be recorded. If a shape of an anomalous thing went by, an alarm would sound. A good analog of this is an electronic filter. You have high pass, low pass and bandwidth filters. In our algorithm, if you had a city street with no commercial traffic, then you would have a low pass object filter. Only small cars would be unalarmed. If a truck (possibly filled with explosives) went by, then the alarm would sound. The opposite of that would be in an industrial area, if a small vehicle was in an area where large trucks were the norm, an alarm would sound. The bandpass filter would detect things like robots, and remote controlled small vehicles sent against an urban target. Anything extremely large or extremely small would get noticed.

2) The purpose of movement filter would be next. Say you have a robot moving down the street, self navigating. One would expect some jitter in the movement. There would be small delays and corrections in self-navigation. However, if a CCTV camera was near a potential target, and it detected movement with no jitter or no evidence of a random walk navigation, it could mean that the direct route to the target was programmed in, and an alarm would sound.

3) The aerial threat probabilities calculator. If I were looking for aerial threats from airborne machines ranging from a toy helicopter with a hand grenade attached to a drone, I would create a huge software 360 degree virtual pillar around the sensors and camera, and then do a threat probability assessment of each sector of that virtual pillar. The highest probability sectors are from where the sun is the brightest (one of the tenets of air attack is to have the sun behind you to blind the observers). Then I would consider line of sight cover. Other factors that would either mitigate or enhance the risk would be no-go zones for an attack (a tall building is in the way). The threat probability from each sector would be calculated and monitored, and the system resources would be concentrated where the biggest threats lie.

4) Potential launch site evaluations. If one were to launch a mechanized attack, is there somewhere (either an entrance or egress) from where the attack would come. These areas would be heavily monitored.

These are just some of the algorithms to consider. The idea of a terrorist or other threat by robot or some other mechanized means is sort of remote now, but in ten years, I bet that this kind of software will be ubiquitous, and built into cameras. And you read about it here first.

If Browsers Were Guns

(click for larger image)

One of my business partners sent me this graphic. This tongue in cheek graphic is so true. We have been struggling to make our mobile web app work as intended with that stupid piece of crap called Internet Explorer 9, by that dead-man-walking company called Microsoft.

The Perils of Technical Trading in Stocks

I was first introduced to technical trading as a software analyst. I was hired to write some proprietary algorithms for a stock and commodities trading group. When I was first introduced to the principles of technical trading, I was blown away. I had thought that fundamentals were the only way to trade (in fundamental trading you look at the companies balance sheet and how well it is doing and whether its markets are expanding and contracting. In technical trading, you just look at the performance of the price alone to gauge what will happen next.)

Here is an example of a very simple algorithm in technical trading. The red line is an exponential weighted moving average of the stock price plotted against the actual price day by day. The rule is when the red line is above the stock price, the price will fall, and when the red line is below the stock price, the price will rise. And it can work rather well. Look at the 5 day example below:
(click for larger image)

It works even better on the long run. Here is a one month run. It is all neat and tidy. It looks like any fool can make money, and that this is a huge money spinner. It truly is the Dummies Guide to Getting Rich. So what's the catch?
(click for larger image)

The catch is that technical trading doesn't work all the time. Fundamentals can trump technical trading. Emotions of the marketplace can trump technical trading. If there is panic because of a terrorist attack, or if there are bad job creation numbers, stocks fall. That is why I don't trade stocks based on my proprietary systems. But I do like to watch what happens.

Do you know why they call computerized stock traders "day traders"? Because it is too dangerous in technical trading to hold a stock overnight. This is a prime example illustrated below. I am using my favorite whipping boy stock Facebook. All of the above charts are the official stock prices during stock market trading hours. But there is pre-market trading going on. Examine the stock chart below. If you will notice, the blue stock price is above the red line meaning that the stock should go up and up. But the stock is Facebook, and today Facebook will announce their earnings. They aren't expected to be that hot.

(click for larger image)

So if you went to bed last night, dreaming of counting your money after Facebook stock rises, you woke up this morning to a nasty surprise. The stock is down over 6% before the market opens. If you were holding a long position overnight, it's not champagne and caviar tonight, but dog food spread thin on crackers.

Technical trading is not for the weak of heart, and as the Gambler says " You got to know when to hold them, when to fold them, when to walk away, and when to run"!

Simple Yet Effective UIX For Dummies

Let's get real for an instant. A website generalist is someone who knows less and less about more and more, until he knows nothing about everything. An SEO/UIX website guru gets to know more and more about less and less until he knows everything about nothing. They both end up in the same place.

When building a web site, all sorts of UIX experts tell me that I have to wireframe websites, and then use video to watch people using them to optimize the (sarcasm on) user experience (/sarcasm off). UIX is king.

If you are a regular reader of this blog, you know that I follow the Google sect of online design -- a super powerful website with minimalist design. When you abstract to the layer of why have a website, you come up with two answers: to sell and to inform.

The primary raison d'etre for a business is to sell. So never mind all of the booshwa design crap for a minute. Malcolm Gladwell tells us that people thin slice -- make a decision within the first thirty seconds. I was told that when I wrote for a major outdoor magazine. You have to capture them in the first thirty seconds. Your value proposition will do that.

So if I am a business, and the sole purpose of my website is to sell and inform, then the landing page of my website has to boldly proclaim my value proposition. Let me repeat that: the landing page of my website has to boldly proclaim my value proposition.

When my value proposition is accepted by the surfer, I have to account for two alternate possibilities. That is where UIX comes in. The first possibility is that I have made the sale. You have to get the customer to get to the buy zone quickly. The second possibility is that I have sold the value proposition, and now the consumer wants details. Simple. (There is a third possibility in the fact that my value proposition bombed, but then the surfer is off elsewhere and out of the equation).

So, I have to unambiguously give the consumer the immediate buy option or arrange the information to answer the putative questions in his/her quest for more information. I have to arrange it such that the most frequently asked question is answered first.

Psychologists tell us that we like a lot of choice when it comes to menu items, but more than 4 or 5 choices actually hinders the buy process.

So the navigation and usability of the website hinges on providing more information to support the value proposition. In a general sense you can't go wrong with W-5 - Who, What, Why, Where and When. When is always NOW, so HOW can be substituted. And of course, you need a call to action. If you don't ask, you won't get.

The above graphic illustrates these principles completely. You don't need to pay big bucks to a UIX expert to produce wireframes and usability reports. Most consulting work is applied common sense. When usability experts fail, is that they are not subject matter experts in whatever you are selling. They would tend to put the specs and fine print in the back, but if you are selling electrical couplings and such, those data sheets are what is required up front. Or they would spend $10,000 filming someone using wireframes to discover what you already know.

If you put considerable amount of forethought into the value proposition and how the user gets to the supporting information, then you have already mastered the Dummies Guide to Extreme UIX.

Facebook's Socialcam app shows friends what you watch - FOX Carolina 21

Facebook's Socialcam app shows friends what you watch - FOX Carolina 21 (Click to read) If you ever needed another reason to quit Facebook, here it is.

DISCLAIMER: I AM NOT A REGULAR OF FOX NEWS or EVEN A FAN !!!!!!!!!! However this article is interesting.

MySQL Cleaning Out Database Duplicates

I had a MySQL database that was automatically populated with raw data and no checking. So as a result, I had duplicate rows of data and each data row had a unique key id number (auto increment). The idea was to get rid of the duplicate rows. I was lucky that one of the fields in the row was an equipment serial number, and I used that fact to easily eliminate the duplicates.

Lets say that the table was called equipment_table. I would copy the table to make equipment_table_clean with the following:

CREATE TABLE equipment_table_clean LIKE equipment_table;

We know that the the serial number is the thing that defines duplicates so:

ALTER TABLE equipment_table_clean ADD UNIQUE (serial_number);

Then feed the data into the new table with the following command:

INSERT IGNORE INTO equipment_table_clean (serial_number, name, col1, col2) SELECT serial_number, name, col1, col2 from equipment_table;

Works like magic and saves a lot of work. Then drop the old table and rename the new table (after you have checked the data).

How To Crack Passwords

I was reading yesterday how 450,000 passwords on Yahoo were cracked. First LinkedIn and now Yahoo. Obviously a lot of people don't take a lot of precautions when choosing their passwords.

I was particularly intrigued by the numbers. Of the 450,000 passwords, close to a half a million, there were 2,925 instances of "12345" being used for the password. That's 3,000 users who had a guessable password. If I ever became a hacker, and wanted to get into someone's account, I would try 12345 as a password.

But there are other choices. As we saw with the LinkedIn crack, a viable percentage used 'linkedin' as the password. In other words, they use the site that they are accessing as the password. Even more simple, in the Yahoo case, 780 people used the word 'password' as their password.

Lax security measures like this make account hacking very easy.

A Jungian View of an Internet Connected Computer

During my long sojourn in the tropics, there were and are times when I am amazingly bored. You can't hop into a car and drive anywhere worthwhile on an 18 mile island inhabited by a quarter of a million people. There are only 7 or 8 restaurants worth going to. After you have had your swim on the beach, perhaps shot a rock lobster with your spear, and had your requisite mojitos for the day -- then what. That's when I decided to go to a meeting of Jung Society. Carl Jung was a contemporary of Sigmund Freud, and those two ruminated on psychiatry, dreams, sex, mental illness and the unconscious in a way liberated from biology, real brain chemistry and actual scientific fact.

Let me tell you, the Jungian theories are the biggest bunch of booshwa that parallel other strange beliefs like UFO abductions, crazy religious dogmas and the honkingest pseudo-science promulgated anywhere. The thing that really made me laugh out loud, is that grown men were explaining the story of how Jung discovered the collective unconscious. It turns out that Jung the psychiatrist (or fisseekee-atrist as Ricky Ricardo used to say) was treating a man that was cuckoo for coco puffs. He gave that man a crayon and a piece of paper and the man drew a rudimentary face. Jung was struck with the inspiration that the man tapped into the collective unconscious and drew a face mask of a Polynesian king on some far away Pacific isle from years ago.

Astonishing. Those were the days!!! When anyone saw something that humankind never understood properly, you made up theories and stories, and if your stories and theories were accepted by a number of people, then it was considered fact. This was how the flat earth, spontaneous generation of insects and other expositors of human knowledge came to be. Carl Jung did the same thing. The sad part is, now that we know the brain is neural nets, synapses, massively parallel biological systems and such, affected by dopamines and serotonins, and yet grown men still believe in Jungian explanations. That would be the equivalent of cosmologists today believing that the earth is supported in space on the backs of stacks of turtles.

So lets suppose that you had a time machine, and you whisked Carl Jung into a darkened room with an internet connected computer in the present era. What would Jung say about the experience? I will now channel Carl Jung.

I have met an intelligent alien being today who's thought process scrolled across it's face. It seemed to have an internal light in its brain, and rather than sound going in from square ears on the side, sound came out of them. There was a phallic kidney or testicle attached to its square box body, and one could control expressions on the lit face by manipulating this organ.
The machine thought in language as well as pictures. It communicated with other beings through letters that were sent through the ether and displayed on its face.

This alien being is highly sexual, as it displayed many many moving pictures of naked human women indulging in sexual acts with well endowed men. I tend to believe that this sexual expression is as a result of its control organ continuously being handled and stimulated.

The alien being definitely has a collective unconscious. It can through a dream mechanism called a Googel, draw information from its universe at a whim.

The alien being is introverted in the sense that it requires a prime mover for any actions, but it is quite extroverted in displaying all sorts of ideas and pictograms on its lighted face.

The alien has a limited complex in terms of emotional array, but that probably is due to its low power status as it is constantly being poked by humans.

The archetypes of this alien are thusly:

1) The self is a self effacing almost inanimate object except for the lighted face
2) The Shadow archetype is that of a know-it-all that is ego-less
3) The Anima, the female component of this alien is almost absent as it does not complain, has no negativity, works diligently without complaint and does not speak while doing tasks
4) The Animus archetype is that of the strong silent male without a hint wrath or rancor
5) The persona is that of a usually predictable person capable of displaying deviant sexual thoughts and raucous music

All in all, this alien being is much too advanced for life on this earth. Because of its total lack of need for psychoanalysis, I recommend destruction of all of these aliens except for one in my closet where I can watch it sexual vignettes and create more theories.

Picture of Offshore Development Amenities

(click for larger glorious image)

I need your opinion. I am going to be hiring code monkeys for offshore software development. Pictured above are the amenities. Do you think that I will be able to attract candidates from the East Coast?

Location: Bahamas

Looking For the e-Commerce WalMart Killer

Dear soul-destroying small town commerce Walmart. You have been a blight on civilization with us now for 50 years. I just read it in my online news source today. I am gleefully serving you notice that this is your obituary.

Oh you were innovative in your day as you made America the servants of the Chinese yourself the biggest corporation in the world.

Your competition killing strategies innovations were many. You introduced every day low prices by having things made by suicidal children in communist in China which rapes the environment to build cheap junk.

You gave us a massive selection of clothing and goods which tore or broke days after using them and an embarrassment of pirated intellectual property riches.

You gave us the urban blight of big box stores.

You gave us slavery again with thousands of minimum wage jobs that prey and consume the weakest of our society.

You created unique supplier partnerships by buying out the entire inventory of suppliers, and then dictating a wholesale price with minimum profits.

You instituted data-driven management practices which squeezed the last drop of blood from every pound of flesh penny out of every cost.

You had an almost invisible corporate culture which created the largest carbon footprint in the world for retailing.

But do you know what? There is someone planning an e-Commerce startup to take you on and slay you like you smote the thousands of small hardware and mom-and-pop stores across North America.

This startup will be based on sustainability. It will offer better choice but incorporate just-in-time inventory management. Gone will be the big box stores (Note to self: Call stockbroker and short real estate companies who develop and manage retail space).

This startup will go a long way to curb over-consumption because the market will not be flooded with cheap toxic goods.

This startup will be data-driven as well, but in a very different way. It will offer the online shopper just the array of articles that fit the wants of needs of that person's profile.

This startup will have the corporate culture of a knowledge industry. Smart.

This startup will not worry about currying supplier favor and relationships across the globe, because supplies can be dropped and replaced with the touch of a few keystrokes.

This startup will erase the low paying retail sector jobs and replace them with higher paying jobs in fulfillment centers.

This startup will provide selection without inventory.

This startup will have everday low prices because it will combine the middleman with retailer.

And hopefully, this startup will change the paradigm of retail shopping for good, and it will be incubated on the East Coast, and prove that visionaries like Frank McKenna were ultimately right. After all if WalMart had its start in Bentonville Arkansas, there is no reason why the successor cannot start in Riverview, Hillsborough, Sussex or a plethora of places where innovation lives in the minds of the enlightened.

The Problem With Social Media Monetization

Now that I have a moment on a weekend, I've been meaning to pen this article before I get down to playing with code. In many previous articles on this blog, I have outlined the failure of Facebook to monetize its almost-billion followers. I point out reasons for the failure of Facebook user monetization and how the stock hasn't been anywhere near its opening day high. As I see more and more startups trying to crack this nut, I had a full insight yesterday as to the intense problems in social media monetization.

I had a visitor from the UK who is a project manager for a global company that is in the top ten of the Global 500. His sons (both of them) work for a small extreme sports company in the UK. One is an instructor, and the other son is in charge of social media marketing. The center is located in a UK city with a population of around 100,000 people. The center offers indoor extreme sports with simulators and expensive equipment.

When the social media marketing guru took over, the social media effort was essentially null. Within a very short time, through the marketing efforts, the center attracted over 10,000 Twitter followers and 600 Facebook followers. Youtube videos have less than 500 views each. The center has less clients on an annual basis than its number of Facebook followers. The effect on business has not been measurable, and the marketing guru is seeking greener pastures.

What this proves to me, is that social media marketing in itself, is a bit of an uncrackable nut. This view was expressed by the social marketing guru's father who said that it is almost impossible to glean KPI's or Key Performance Indicators except by inference when looking at the bottom line. Many bricks and mortar companies have been closing their Facebook stores because of poor performance, and hence poor ROI (Return On Investment). LINK STORES CLOSING ON FACEBOOK

Many of the "Likes" on Facebook are from people who have no intention of visiting the extreme sports center, but want to convey the image of someone who does. On Twitter, many followers hit the follow button to get a reciprocal follow. Youtube is the ultimate arbitrator of numbers because it shows you the hardcore numbers of people who would sit through a video. However, this doesn't necessarily translate to a visit to the center either.

So what gives? It is my opinion that way too much value is given to social media interaction. It is also my opinion that when one visits social media, one is not in a frame of mind to shop. Shopping and socially interacting take place in different areas of the brain. Shopping fires the hunter-gatherer neural nets and socializing is the semantic opposite of that. In addition, it is part of the human psyche to have a delta between what people say they do, and what they actually do. That is the key difference between social media and other internet interactions. Social media is based on what we say, and may or may not have a basis in reality.

So how have companies actually cracked the nub of the problem of social media monetization? It is a tantalizing problem because the rewards of solving it are huge. The way that companies have attacked this issue, is to actually track what we do, not what we say. They do this not with social media, but with other websites.

For example, when you visit the New York Times website, the results of a tracking cookie is sent to 103 separate companies. They know more about you than the government does. Do you want an example of how this happens? Go to and if you have a Safari browser, you can download a plugin to see where your data is going. If you don't, you can watch a simulation.

When you actually make it a point to visit the New York Times or the Huffington Post or any other website, you are making a statement about yourself. Through data mining, they can tell your demographic, your ethnicity, your age, your socio-economic status and anything else they need to know to market to you.

Facebook can't work that way. Facebook itself, admits that 5% of its accounts are fake, and I am willing to bet that the percentage is much higher. And if the account isn't fake, then the personal information such as birthday is fake.

So essentially what I am saying is that Facebook will never see the realization of the full potential of monetizing its information. Companies who mine social media for marketing data will at best have so-so results.

As a good example of this, yesterday I drove a Chrysler 300 automobile -- brand new. It wowed me. There is no key to start it. The key is a fob with a near-field RFID chip. The bells and whistles were mindblowing. The engine was 5.7 liters and when I tramped it, the car almost launched airborne. It had a video camera with a proximity detector while in reverse. In reverse the sunscreen on the back window lowered and came up again in forward gear. The car recognizes various occupants in their seats by their weight, and settings for air can be adjusted on an individual basis. There were lights on the bottom of the doors to light your footsteps.

It was one of the most amazing vehicles that I have ever driven. If I had tweeted that I was in love with the car (which I was) that lead gleaned from social media would be useless. My wife thinks that the car is a leviathan, an example of gross consumerism with a large carbon footprint, and our family would have a nil chance of owning that vehicle.

However, if I had gone to a search engine and looked at various sites selling that car, and I Googled the price, that would be a valid indication that I would be seriously interested in acquiring the car. It would be a matter of checking what I do, not what I say.

So is there a way of cracking this nut? There could be, but I don't think that it come from social media. I ran into a startup last week called . They are redefining analytics using the premise that we as social beings, cluster around and belong to various tribes. It is an intriguing concept and one that could have legs because it has a stronger premise than the one that currently exists where "I should be able to make money off people because I have what they told me is their date of birth and what they say they like".

Since this startup doesn't have the pull to do the tracking cookie bit with Fortune 500 websites, they have to be smarter about it. I really have no idea of how they assemble the tribal information, but there are a few things that they should do. They should use search engines, data mining, mathematical models and Bayesian Inference to tease out the tribes. This process is taking facts and integrating them into knowledge with models that have a very high correlation with real life.

It was the data miners who figured out that a teen at Target stores was pregnant by the probability of buying vitamins and skin cream. It was the data miners who made a mathematical model that showed that combining beer and diapers on a Saturday in a supermarket gives a huge boost to sales. It all comes down to the data miners.

For to succeed, they have to add business intelligence and data mining to their technology portfolio, and then they will have a better chance of making money from their analytics than that of Facebook. This is my own humble opinion.

Disclaimer: I am not associated in any way with any of the companies mentioned above.