When I first started creating this website and my blog, I wrote most of the software on top of Apache myself. I want this to be a learning experience as a webmaster. A very good learning experience it is, I did not have to put in too much time as the software is rather simple. But overtime, things need touch up here and there. Some reorganization is needed to make it friendly to search engine. The layout and presentation also need some fine-tuning. The trouble of building my own website is these maintenance task tend to be deferred and piled up.

Finally I have taken some time to build an index for my past articles. I have also adopted some microformat to help the search engine to find my articles. I feel good after completed long neglected work today.

2007.09.06 [tech] - comments (0)

Web discussion forums is often a place to find useful information. Search engine make this valuable archives only one query away from us. However, I am surprised at how many forum software still present their information badly. For example, I found this from some slashdot derived forum:

Posted by xyz on Saturday February 07, @02:02PM

It gives fairly detail posting information down to the minute. The big question is, what year is it? Why is it omitted. This is especially important for discussions on tech issues, which evolves quickly. A recommendation made 3 years ago is often irrelevant to the problem on hand today. Indeed one important Internet literacy skill is to be able to discern the relevancy of information and disregard obsoleted information.

This is not an isolated instance. I write this because I have seen this same problem crop up in multiple and varied places. The only explanation I can give is that the forum owner only focus on the present, such that the year is understood and preferably omitted. But the archive will continue to be a valuable resources 3 years, 5 years or 10 years on. In such case it is ridiculous to omit the year from the entry.

Think into the future, not just the present.

2006.11.25 [tech] - comments (0)

• Mike Arrington of TechCrunch started the second day by giving his take on startup winners and losers. He defines winners as who accomplished a liquidation event like IPO or acquisition. But look carefully at his list, which includes companies like myspace and skype, all of them got acquired. There was no IPO!

  Further he think this list is hot for new startups:

  • Flex apollo
  • platform/backend
  • developer tools
  • bring desktop app online
  • office efficiency
  • cloud storage
  • identity
  • market destruction
  • ENTERPRISE

  Right after this he downplay his own words by saying best entrepreneurs would avoid this type of advice :)

• Ted Rheingold from Dogster took us for a tour to what he calls passion centric communities, from his own dogster and catster to many many more. What make them attractive to their loyal communities? Entertainment, information, sociality and services are what they should have. As Ted say, you can't fake sincerity. I can certainly feel the passion he radiates.

  Here is one detail he think would make some of these sites better - use something other than white as the background color.

• Cal Henderson talked about things they have learned from building flickr. His advices focus on operation issues. Plan for downtime and maintenance. Setup a process to handle it. Communicate with the users about the problems and so on.

  Also interesting is that flickr does not have a separate QA team. The developers do everything from coding, testing to fixing bugs. He also emphasize the importance of tools. A good example is an one touch deployment tool. They also employs many other tools for monitoring.

My final observation is that the sessions are dominated by people from small companies, most of them no more than five years old. Also a sizable portion of the audiences are very young people. This is the age of new ideas. I am certain that next year we will see even more innovations.

More on day 1.

2006.09.15 [tech, futurewebapps-sf06] - comments (0)

I was excited to be at this year's the Future of Web Apps summit. This two days conference has gathered many great speakers to look at some of the web's most successful sites and applications and practical advices on creating your own web app.

Here are some highlights from the first day.

• Dick Hardt from Sxip take about identity 2.0. He's points about identity is based on past history, which in turn predict future behavior is well taken. He later demonstrated logging on to websites using different protocol such as OpenID, i-names and Infocard. Since all discussions remained at high level, it is not obvious to the uninitiated to understand what is the magic behind these new technology.

• Kevin Rose of digg recounted an interesting anecdote that the first version of digg was built with a budget of $2000. He has contracted a programmer through elance who later became a shareholder.

• Tom Coates from Yahoo has examined social software, what it is and what makes it works. Two points are especially noteworthy.

  He noted there are two models of social software. The consensus model, represented by wikipedia, presents one complete version of information created by many contributors. On the other hand the polyphony model, represented by flickr, presents many different versions all at once (e.g. many different photos for any give tag). He believes consensus is difficult to achieve and the success of wikipedia is more of an exception. People would be better off focusing on the polyphony model.

  Secondly he cautioned about rewarding user, either using virtual points or worst of all, actual money. He believe an imperfectly designed reward system would easily skew user behavior or even unravel the entire community.

• Tantek Çelik from Technorati gave a good introduction on microformat, which allow us to embed structured data in HTML. It so simple that I have updated my about page right away to include a hCard and a link to add my contact to visitor's address book.

• Steve Olechowski of Feedburner gave us an interesting tidbit. He said adding contextual ad on RSS feed does not perform at all. He explained that ads network like Google AdSense or YPN are geared toward offering product to people searching for something. For feed readers they are not really looking for anything specific but just interested in what you have to say. Since feed subscribers are people who come back daily, he suggest a brand awareness campaign would work well in this media.

• Carl Sjogreen of Google told the story behind Google Calendar. He talked about how hard they work to design the UI that is really easy to use. Things like quick add box (natural language input box) and in-place editing has made it a much more fluid user experience.

More on day 2.

2006.09.13 [tech, futurewebapps-sf06] - comments (0)

Coming from a culture where we associate news with accidents, violent crimes and disasters, I have came out and start to question our perverse obsession on negative news. Not all news agencies are created equal though. The Christian Science Monitor, for example, put a lot of focuses on education. Here is one good piece of story about an experiment in India where they put free computer stand in slums and found kids made good use of them. Surprising they found that with no supervision or instruction

the children "download and play audio and video, send and receive e-mail, chat, and so on," he says. They quickly move on to learn some English from English-language websites, read Indian newspapers, and even "look for jobs for their fathers,"

And it does not end there.

Many have changed their goals from say, rickshaw driver to engineer, and most now want to go to college.

Seeing this kind of heart warming notes really light up my day.

Thinking more about this, it is not all that surprising that kids can teach themselves about computer without any instructions. Afterall I learned much about computer and programming by myself before even set foot in the college, and all these without really spending much in gadgetry.

2006.06.01 [education, tech] - comments (0)

I am a little behind in blogging. Last week I have went a few separate but all very interesting events. Last Friday was the Long Now Foundation's seminar The Long Time Tail by Chris Anderson with Will Hearst. Chris has introduced an influential concept of long tail. This time he has extended the concept in the time dimension. The sales of an item would decline over time. Traditionally an item would become unavailable or go out of print because it becomes uneconomical to continues to sell it. But in the world of unlimited virtual inventory with a searchable, discoverable front end, Chris discovered that the demand, though diminished, would extend well into the future. This is the same as the original long tail framework. And it applies very well to the time dimension.

The business implication? A content archive has great and untapped economic value. The implication to the consumers? We will enjoy a much wider selection of quality works. And we will be less subject to marketing hype that tries to push hits to make money in a flash of time.

Chris' presentation was full of studies of different markets. One study of Netflix users says the customers are a lot more satisfy with old movies than new movies. There is a logical explanation for this. Good movies stand the test of time. I for one are very glad to find the old collections available on the market again. For some time old Hong Kong movies have been release in VCD format, a lower quality, lower cost predecessor to DVD. Despite the lack of a searchable online front end and generally available only in a uncategorized, flea market like retail outlets, I still manage to amass a good collection great movies over time. I have fond memory of many old movies I have seen one or two times when they were released years ago. It is great to be able to enjoy them again many years later. The old releases also reportedly turning a good profit. Yet another confirmation that there is a demand of the long tail.

2006.05.19 [tech, economics] - comments (0)

Tim O'Reilly posted some analysis and visualizations of book sales data on his website. It analyzed book sales trend in several dimensions. Java is shown on a steady downward trend. On the other hand C# is clearly gaining share. The idea is the state of book market would be a technology trend indicator. Of course this is more likely to be a trailing indicator. Idea has to be mature enough before it can turned into a book. And it would takes months for it to be published and adopted in the mass market. Nevertheless it gives a good picture what people are really buying into.

2006.04.20 [tech] - comments (0)

Ray Ozzie made an extraordinary demo of Live Clipboard in ETech. He illustrated a very simple concept to enable users to pass data between web applications and the PC, by using the clipboard! All this involves only defining the live clipboard data format and simple enhancement to the applications.

He reminded us the clipboard was a great GUI innovation that allows transfer of data between multiple applications. This is also a glimpse of the power of the semantic web.

2006.03.09 [tech] - comments (0)

Practical Internet Groupware
by Jon Udell

In his book "Practical Internet Groupware", former BYTE magazine editor Jon Udell layout an architecture that links human minds into collaborative relationships. Base on his actual experience in building BYTE's intranet as well as the magazine's public online services, he gave his insight on the powerful use of Internet.

Among the many IT books I have read, this book stand out as sublime, even avant-garde. Got a question? Search the Internet, send a follow up email to folks you have never met. That's something many of us have probably done without much thinking. Yet Jon would step back and reflect on the dynamic that had happened. An ad-hoc workgroup was formed between him and several person on one particular task. The collaboration was unbounded by time, geography or corporate affiliation. He strived to grasp the subtle interactions and to facilitate this flow of information on the Internet.

People are lazy and do not like to learn or adapt to complex rules impose by computer systems. On the other hand simple rules and clever UI tweak can often make interactions spontaneous and effective. Use an appropriate subject for a message is one good example. The author discussed one of the oldest groupware on the Internet, the Usenet newsgroups. He termed it conferencing and explained why it is a better channel for some kind of interactions compare to email. Many of us who get caught in lengthy email debate would be delighted to know there are more effective way to conduct this kind of discussion. Indeed a seamless integration of web, email, newsgroup and a searchable document database are the components that make a formidable groupware application.

Unlike most IT books, he did not focus on any single platform, computer language or a technology. Whether it is a tool from Microsoft or its competitors, a freeware or a commercial product, he would use it if he see fits. Throughout the book are short, unglamorous, but nevertheless working code samples. Given I read this 6 years after its 1999 publishing date, many of the code or specific technologies are already considered obsoleted. Yet the insight that stem from these early system are just as relevant today. Think just what is the core component of web 2.0 technology? User participation!

Perhaps nothing reveal more about this book than its front cover. The 'practical' in the 'Practical Internet Groupware' means everything is derived from actual experience and real code rather than a theoretical discussion. Yet it is in small print while the 'Internet Groupware' is emphasize in the banner. That's because the code and the actual systems are just starting points that spawn the exploration of threads that link people into collaborative relationship. This is an immensely powerful Internet application we have yet to master.

2006.02.26 [book, tech] - comments (0)

I hit upon an entry of my family name Tung 董 in Wikipedia yesterday. What I saw doesn't delight me. It looks like a mischievous teenage has put his twisted bio into the entry. I went in to clean it out and added Tung Chee Hwa as the sole representative of the Tung clan for now. Although it has only 3 lines, this is my first original content contributed to Wikipedia. And a little something I have done for my family name. Furthermore I am delight to find the prankish entry has only been up for 10 days before I shot it down.

Wikipedia is frequently looked as a rival to traditional institutions like Britannica. People like to pick out bad entries from Wikipedia and complain how professionalism is being overran by amateurishness. I of course have many counter arguments. But today I have realized something more. If encyclopedia is to be a most comprehensive reference of knowledge of all kind, then it is a long tail business! I don't expect anyone would care enough to put an entry of my family name into an encyclopedia (I have been generous to call Tung "a common Chinese family name"). There is going to be many many knowledge important only to a small group of people and few others. In the age of information explosion Britannica would have a hard time to hire enough experts to write about each and everything important and remain economically viable. Just 4 years since inception Wikipedia already claim more articles than Britannica. It is really no accident.

2005.11.11 [tech] - comments (0)

As a software developer I am a strong advocate of open source software. They are used extensively both at my work and for my private projects. In retrospect, open source platform, often referred to as LAMP, has long past the stage being just an useful extension to proprietary software. It has become my dominant development platform. If I were to build a server application today, it would make very little sense for me to consider Microsoft servers. Why choose a framework that leave you with a single tools vendor. LAMP has proven to be technical viable, cost nothing to experiment and distribute, and more importantly I trust them because of its openness. Nowadays I need little justification to pick LAMP over Microsoft.

What a big leap from just a few years ago when it looks like Microsoft is going to take over the world.

2005.11.01 [software, tech] - comments (0)

Reuters reports low-cost computers and open source software making it cost less to invest in a startup. I always think open source is an under appreciated revolution. Not only is it redefining laws of economics and inspires people to collaborate, now it also form the basis of a low-cost computing platform and sprawl a new wave of innovations.

2005.04.03 [tech, business] - comments (0)

I got to try out the newly released Yahoo API. Yahoo's move to open up its services to developers is a good news. But specifically I want to applaud them for designing a very developer friendly interface. Registering an application ID a breeze (and it doesn't seems to be mandatory for a quick experiment). They have the sensibility to use REST services over SOAP. It takes me 2 minutes using the web browser and text editor to experiment with their services and come out with a good understanding of what it can do. All these can be done without specialized tools required by SOAP. Google should have go this way in the first place.

2005.03.02 [tech] - comments (0)

A very interesting article The Long Tail from Wired took a look at a phenomenon how obscure products found their audiences thought online stores. Even though these low volume sales form the tail end of the sales chart, their aggregated size is so significant that it might spur a new form of economy as oppose to the mass-market economy today.

2005.01.19 [tech, economics] - comments (0)

I have looked back to a previous post "Bill Gates' Inbox" and was very disappointed to find the link to the very funny mocked mailbox of Bill Gates' was broken. Did the author took it down after getting some letters from Bill Gates' lawyer? I turned to the wonderful Wayback Machine, which crawls the whole World Wide Web and archives historical content for public reference. Easy enough the picture is still available here. Indeed I can find the historical record of my own website. Thanks to the Internet Archive for bringing this invaluable service!

It turns out the link is broken not because of Bill Gates lacking a sense of humor, but for a simple reason that the author has moved the URI. Alright, I will update the entry with the new link then.

2004.08.19 [web, tech] - comments (0)

I have the opportunity to attain Tim O'Reilly visionary talk before. This time I come up on his powerpoint presentation The Internet Paradigm Shift. It summarized many important ideas of the future of computing, like the commoditization of software stem from open source software development. He also have some insight in suggesting hidden business model and that owning user data would be the new form of customer lock-in. I think the new era of internet applications will not be far away. In a few years we will be so use to it that PC computing would become a relic.

2003.12.24 [tech] - comments (0)

In this Scientific American article, Armando Fox and David Patterson discussed an unconventional approach to build reliable computing systems. Instead of focusing on improving software and hardware reliability, they consider failures inevitable. Their recovery-oriented computing (ROC) approach focusing on bringing the service back quicker. Some method such as micro-rebooting is considered.

Their study also revealed that operator errors cause most of the system downtime. Perhaps the most important boost to reliability is to improve system usability.

2003.05.31 [tech, computer] - comments (0)

I have been using O'Reilly Network's Safari Bookshelf for over a month. The Safari Bookshelf is an electronic reference library for programmers and IT professionals. It provides online access to over a thousand titles from O'Reilly as well as several other major publishers, like Addison Wesley. [more...]

2003.05.08 [tech, book] - comments (0)

My server is now using its idle time to help search for a cure for SARS. The infectious diseases has already causes hundreds of deaths. [more...]

2003.05.03 [tech, health] - comments (0)

I have not used Usenet News for a few years. With the low signal-to-noise ratio and the unattractive text message based interface, I expect it would soon go obsolete and replace by the more attractive web based forums.[more...]

2003.04.25 [tech] - comments (0)