Huibert Aalbers Portfolio

Chapeau!

May 7, 2007

Over the last few years, in most countries, statistics have shown that, in general the trend is pointing towards lower participation in public elections. In some cases, a particular election can break the trend, specially when important matters are at stake, people live in fear or when the country is highly divided on a particular subject. However, analysts usually assumed that citizens had progressively lost interest in politics. As a result, in some countries it has become usual to see participation scores well below the 50% mark. This obviously raises a lot of questions about the legitimacy of those elected under these circumstances.

Raising participation numbers in local, state and national elections should be the primary goal of every true democrat. Too often do we see a candidate win because it’s supporters are more committed than those who back the opposition. Some may consider that fair but I disagree. We should only be satisfied with election results when they are accompanied by high participation rates across the board. That is why I was so glad to see an amazing 85% participation rate in France’s Sunday presidential election. This result really deserves a standing ovation and we should all work in our countries to make sure that we obtain similar results in future elections.

Even though France cannot claim to be the oldest modern time republic (that honor goes to the Dutch), it is well known that many of the ideals that inspired both the US Constitution and the French Revolution, came from French thinkers. That leads many to consider France to be the cradle of democracy. Seeing that 218 years after the French Revolution, democracy is still alive and well in that country is something to cheer about.

More on productivity

May 1, 2007

According to this BBC article, the authorities in India’s premier engineering institute, the Indian Institute of Technology (IIT) in Bombay (Mumbai), have cut off internet access to students in hostels at night. They apparently feel that 24-hour internet access is hampering students’ academic performance and overall personality development.

I agree. I am most productive when I do not use the Internet. Opening a browser window is just like opening a chocolate box (if you are into chocolate), there is such a wealth of interesting and entertaining information available that it is hard to just look at what you really need. Fast Internet connections only make this problem worse. A slow connection helps people realize that they are wasting their time. With fast Internet connections, you do not have to wait and therefore time flies. Before you realize it, you have spent too much time surfing the web.

Shutting down Internet access is not a good solution to improve sound Internet usage. Reducing the bandwidth is. Modem speeds are generally decent enough to allow students that require critical information at 3AM to complete a term paper, to access Google and Wikipedia. However, slow access will make sure they do not waste their time on YouTube or other Web 2.0 sites.

Once again, I’m headed to WWDC

April 29, 2007

After Apple announced two weeks ago that Leopard would be delayed until October, I was disappointed, like almost everyone else. I expected to be able to play with the new APIs in June, as promised. Even though I had much fun last year at WWDC, I had decided not to attend this year. I thought that it would be much cheaper to read the online documentation published by Apple, which I expected to become available in the same timeframe.

The fact that Apple’s WWDC will now happen well before the actual OS launch suddenly makes the event much more interesting to me. The Leopard innovations track seems specially interesting for developers who want a small head start.

Last year, WWDC was really only about discovering the future of the Macintosh platform. The Leopard preview distributed to developers was much too far from completed to be any useful. This year’s WWDC, with a feature complete OS, should be all about how to use the new APIs in existing or new applications. To me, that is a very compelling value proposition.

So, just when it looked like I would miss this year’s WWDC, I changed my mind. Once again, I’m headed to San Francisco. See you there.

Good question

April 24, 2007

I am a big podcasting fan. I usually spend about two hours a day listening to mainly tech personalities interviews. This is a nice way to understand industry trends and learning about cool new projects. Here is a list of some of the podcasts I subscribe to:

IT Conversations (In particular Technometria)
Late night Cocoa
TWiT (This Week in Tech with Leo Laporte)
The Scoble show and other podcasts from the podtech network

The more I listen to interviews, the more I get upset when I hear that most interviewees follow the exact same pattern when asked a question. They usually will start answering by saying something like “Excellent question” or “That is a great question”. That feels so fake when you hear it over and over again. It is obvious that there is some form of consensus in the US (I haven’t seen signs of this spreading to other countries yet) that this is the correct way to behave during an interview. I understand that by acquiring this habit you some additional time to think about your answer and you may even please the interviewer. However, while this technique may have been effective at the beginning, when few where using the technique, now that everyone has learned it, it just feels unnatural and even plain silly, specially when the question isn’t that brilliant.

Inconvenient solutions

April 20, 2007

There is no doubt that the movie “An inconvenient truth” was able to change the perception of where our world is heading if we fail to act. The movie, combined with high gas prices and the war and Iraq was able to shake consciences and get our society moving. I am amazed by the incredible amount of business initiatives related to new renewable energy, plug-in vehicles, recycling and conservation in general and I am sure that everyone gets excited at the prospects these new technologies create.

I think however, that we are being misled. Most people seem to believe that we can solve the problem by changing slightly our behavior and applying new technologies. The truth is that we will need to go further. Initiatives like ethanol adoption may reduce foreign oil dependency in the US but consequences on the environment are far from clear. The Brazilian experience with bio-fuels is far from conclusive with many experts blaming in part deforestation in that country on the growth of the sugar cane plantations which are needed to produce that fuel. Other fuels perceived as alternatives to oil like diesel or hydrogen have their own issues. With really clean energy sources such as wind or solar growing too slowly, our only real option is to reduce energy consumption. Besides energy, the world is facing other problems such as lack of drinking water or food. How is that possible? Never in history has so much food been available and the same applies to water. The problem is simple, there world population is now six billion and counting. Since we cannot easily increase food, water or energy supply, the solution has to come from reducing the last variable, human population.

Do not worry, I am not proposing at all to start World War III or executing minority groups. What I would like to see is an open debate around the subject. What people should understand is that if for one generation (25 years) each woman only had one child, all the catastrophic prophecies mentioned by Al Gore would probably not happen. I am really puzzled by the fact that a solution as obvious as this one (although extremely hard to implement) is not even mentioned in most discussions. However, if we do not discuss the matter, it is very unlikely that parents will even consider the impact on the environment of having a large family.

Over the last few decades, in most countries (even the emerging ones), birth rates have been decreasing steadily. One reason has been the fact that parents have been bombarded with (not always obvious) messages on TV, movies and radio telling them not to have more children than they could afford. As a result, parents in those countries feel that they have the obligation to provide their children with good education, long vacations, and abundant clothing. As a result, couples have reduced the number of children they decide to have. That message has been good for children, parents and society in general. Why not slightly update that message now and include the fact that small families are good for the planet too and repeat it all over the world until everyone gets it?

We all know why. Such a message goes against the holy scriptures of most large monotheistic religions. Of course, at the time, the goal of their leaders was to outgrow their competitors to convert non-believers, by force if necessary, which made large armies a necessity. Today we face a planetary crisis. We need to discuss the options with an open mind. If this means upsetting some religious bigots, I personally don’t care much. More important issues are at stake.

The value of Product Certification

April 20, 2007

During this week Technical Leadership Exchange, attendants could take up to three product certification tests for free. On day 1 I decided to take the “Developing WebServices with Rational Developer 6.0” exam since it was a topic I was quite familiar with. I found the exam to be pretty easy, specially considering the fact that many questions were related and that some questions provided the answer to others.

The same day I passed that exam, one of the members of my team failed the “DB2 version 9 fundamentals exams”. That was totally unexpected, since he is an experienced Information Management IT Specialist. I decided to take the same test the next day to find out how hard the exam really was. Just for the record, most of my database experience has been working with Informix. I am a DB2 certified professional but I took the exam four years ago with version 7.1 and I have never installed or used DB2 version 9. To make thing worse, I had to complete the exam in 45 minutes (instead of 90) because I was running late for dinner. So, how did I do? Not extremely well, but I was able to pass the exam.

So, how is it possible that I could pass the exam while a more experience engineer failed. One explanation is the test’s heavy focus on SQL. Despite being called DB2 fundamentals, few questions are really DB2 specific. Anyone with a solid SQL knowledge can obtain a good score even without having used that particular database engine. That is a serious problem, we cannot expect employers to know the contents of a particular exam. We should expect the title of the certification to clearly describe the skills being tested. However, that does still not explain why an experienced professional can fail such a test. The answer is quite simple, English fluency. Most product certification tests tend to be quite confusing with many answer options looking very similar. That is fine if English is your primary language, but not for those who are not fluent in English.

I feel that this is unfair. Although I do not care much about product certification when evaluating candidates, many managers do. Candidates who do not speak English well are being double taxed, first when being evaluated on their foreign languages skills and later when asked about their technical skills. Now that I am thinking about it, I wonder if the fact that a country like India is known for the technical skills of it’s people this is due to the fact all students speak English and therefore have easier access to technical information. Having lived in many countries around the world I can tell you that, in my experience, stupidity is evenly distributed and so is intelligence.

So, what is the solution? For starters, product certification tests should be available in multiple languages. This shouldn’t represent a large effort for companies such as Cisco, IBM or Sun. I am convinced that if this happened, the number of certified engineers in regions like Latin America, Europe and Asia would rise dramatically, unveiling new business opportunities in countries that currently are not seen as investment worthy.

However, that is obviously not enough. Universities in non-English speaking countries should acknowledge how key the proficiency in English is to their student’s success. Unfortunately, in many countries national pride is preventing government from taking serious actions. This is too bad as it is taking away opportunities from people that truly deserve them.

IBM’s Technical Leadership Exchange (TLE)

April 16, 2007

I am currently spending a couple of days in California (Anaheim), attending IBM’s Technical Leadership Exchange. This event is organized once a year to train technical talent and also to serve as a forum where participants can share ideas about the future of technology. It is an interesting event because unlike other events organized by different brands within IBM, here it is possible to meet people from all the organizations that form the company.

So far I have been enjoying myself. Besides being able to spend some time with my colleagues at Disneyland, I have been able to attend some very interesting sessions. Topics range from software to hardware to Web 2.0. Product certification is free and many are lining up to get certified. This may not seem very fun, but the truth is that I see it today mostly as a game. It is a challenge after all, just like playing Sudoku or Crosswords. Now that I am a manager I do not expect people to judge me on my technical certifications, but I still find it amusing to pass an exam once in a while.

The truth is that I feel very fortunate to participate in this kind of events. A couple of days ago I was interviewing a candidate for a job. He told me that his current manager did not give him any learning opportunities in the last three years. This may sound ridiculous to many, but engineers really enjoy learning new technologies and get quickly depressed when they feel that they are falling behind. IBM offers many learning opportunities to its employees. This year, most of my direct reports will spend four weeks on training. That is good for the business because customers want to deal with knowledgeable vendors but it is also excellent for the technical team’s morale.

We all saw the signs

April 13, 2007

Over the last few months several rumors indicated that Mac OS X would be delayed. Often, the alledged reasons were not really credible. The last rumor, that Leopard would be delayed because of lack of proper Vista support in Bootcamp, launched by Taiwan’s Digitimes , was simply laughtable, and as such was quickly dismissed by the Mac community.

However, there was a reason these rumors emerged. There were signs all over that Leopard would be delayed, and we all saw them. How many wondered why Apple hadn’t started touting the new features of Leopard just two months before the official launch date? Weren’t we surprised to see that recent builds still had long lists of known bugs? Come on, we all knew deep inside that this was coming, but we had faith, even after the AppleTV was repeatedly delayed.

What strikes me is the reason argued by Apple to delay the release of their next generation OS. Blaming the iPhone seems simply silly. If it is true, it sends the message to the Mac community that the iPhone is more important than Leopard, something that is not so smart to do, since many Mac users already feel some kind of iPod envy, not appreciating that the music device is taking most of the spotlight. Most Apple fans want Apple to be successful in order to increase their computer market share and fight Microsoft head to head. Whether this is realistic or not is another question, the fact is that that is exactly what they expect. Removing the word computer from the company name and choosing the iPhone over the Macintosh does not bode well with the core Apple supporters either.

The truth is that a four month delay is not something to be truly concerned about. In fact, this delay solves many problems for Apple. Launching two major products (Leopard and the iPhone) the same month is something very difficult to execute properly. By only showcasing Leopard in June at the WWDC they can focus their marketing machine on the iPhone launch while starting to create hype around the new OS to ensure a successful product launch, just in time for Christmas. Quality of the code is obviously a concern also, as mentioned in the press release. Mac users are used to the highest standards and a buggy OS would probably have translated into a major PR disaster.

However, what I think most die-hard Mac users fear is a slow death of the Macintosh platform. We all know that personal computers (both Macs and PCs) are an endangered species. The Internet is slowly but surely killing them. Since the introduction of the idea of the network computer, the need for desktop OSes has been declining steadily and most of us spend most of our time working online. However, people (specially Mac users) seem to like their personal computers. They do not want to lose them and are asking for reasons to remain passionate about a product that every (smart) analyst sees going the way of the dodo. People are asking for 3-D interfaces with gorgeous graphics and in general all those things that cannot be achieved by a browser. That means that until web browsers become able to offer a true rich user experience, there will be a need for desktop computers. I hope that Apple realizes this and does not start to milk the platform while focusing too soon on alternative platforms.

Fortunately, Apple clearly mentioned that the technical resources assigned to the iPhone were “borrowed”, not transferred which leads us to believe that they should be soon back working for the Macintosh division. However, the clear disappointment felt throughout the Mac community should be received by Apple as a clear warning sign that many loyal customers are feeling nervous and that they need to see a very strong commitment from Apple to the Macintosh platform. We will all be watching closely Steve at WWDC.

Eight-core Mac Pro, are we ready?

April 4, 2007

The extremely low-key release of the new eight-core Mac Pro earlier today illustrates perfectly some of the problems the computer industry in general and Apple in particular are facing.

When Apple decided to move away from the PowerPC it was simply because IBM could not deliver a competitive processor to challenge Intel. At the time it was clear that they had hit a wall both in terms of performance as well as heat dissipation and that they were looking for alternative ways to solve the problem, with innovative chip designs such as the Cell processor found inside Sony’s PS3. Apple wasn’t interested in new designs, they wanted a quick fix to solve the performance gap and Intel provided it.

However, Intel couldn’t defy the laws of physics either and now, less than two years after Apple switched chips we see clearly that, in the future, GHz will matter less and less, specially after chip manufacturing moves to 45nm this year and further improvements become harder to achieve. Instead, what will matter will be how many cores a chip includes and how efficiently they can communicate in order to collaborate to solve complex problems, something that IBM had recognized long ago.

Why do we need eight-core workstations? In the server arena, the answer is obvious, all the applications are multithreaded and can easily use all the cores you throw at them. The more processors you have, the better a database or an application server performs. On a workstation, the situation is totally different. Most applications can only use one application. Sure, Having two cores on my Mac Book Pro is nice because watching a movie does not negatively impact the performance of my machine if I am using another application. But that is what I call user-multithreading. Since users can perform multiple tasks at the same time, having multi-core chips helps avoid performance issues. That is nice, but what we really want is to have applications such as Pages or Excel use those extra cores so that a four-core computer delivers four times the performance of a single core box. That is not the case today. Almost none of the applications a normal person uses daily are multi-threaded.

The main reason most desktop applications are not multithreaded is because writing that kind of applications is hard, very hard. That is why everyone has been dragging their feet when Intel and others have urged them to rewrite their applications to take advantage of multiple core processors. What applications need to be upgraded first? Those that require the most performance, like games, video editing systems and number crunching applications come immediately to mind.

While this move from single-threaded to multi-threaded applications is affecting the whole industry, Apple faces some unique challenges. Tests run by AnandTech demonstrated that Linux created processes and threads, raised signals, and performed other interprocess communication far more rapidly than Mac OS X. This does not really have an impact today when most personal computers have only single or dual core processors but could quickly escalate into a major issue if eight or sixteen core machines become common.

The good news is that Apple is well aware of this and seems to be working on the issue. I expect Leopard to offer a partial answer to many of these problems. The developer community has to work also on their part of the problem by delivering scalable and efficient applications that fully use the hardware they run on.

That is why an Apple spokesperson said that “The 8-core Mac Pro gives pro software developers a platform to prepare new versions of their applications for the future, when 8-core technology is more prevalent on the desktop” according to MacWorld. We are clearly not ready to harness that much power today, but we will get there, eventually.

Idea for a new Mac application

April 3, 2007

I still haven’t decided whether I will buy an Apple TV or not. The only thing that is holding me back is the fact that in order to stream video to that device, the files have to be added to an iTunes library. Since my main machine is a laptop, my HD is relatively small and I cannot afford to keep all my videos on it. Instead I have to rely on an external disk that is seldom connected to my MacBook Pro. Therefore, I do not feel that iTunes is the best tool to manage my video files. Obviously, another problem is the fact that many of these files are encoded in formats not compatible with the AppleTV, although they can be played fine using quicktime with additional codecs.

I know that I am not alone, many share this very same problem. Right now everyone seems to be either re-encoding all their video library or trying to solve the problem by hacking the Apple TV. One solution requires to open the device, extract the HD, mount it on a Mac and add additional quicktime codecs. Even though there are still some remaining problems with this solution, I am confident that, in time, it will work. However, no matter how clever the solution, it will still require users to open the AppleTV and void the warranty.

I think there is better way. Why not write an application that pretends to be iTunes (by advertising as such using Bonjour) in order to communicate with the AppleTV. This would allow to stream any video file to the AppleTV, since iTunes and the Apple TV most probably exchange video frames, no matter what format the file is encoded in on the remote Mac.

I know, this is not a trivial project, otherwise I would try to do it myself over a week-end. However, I am confident that it can be done. Airfoil did something similar for AirTunes by allowing any sound produced on the Mac to be streamed to an Airport Express, not just iTunes songs. The nice thing about such an application is that it would make it easy to play any file stored in a designated folder without requiring to hack the AppleTV. So, what do you think, is anyone up to the challenge? I pledge US$100 to the project. Is anyone else interested?