30 April 2006

The more I got these good signs, the further I feel away from home

And this one is very interesting.

That is not from Da Vinci Code, this the ISO/IEC 15434 2D Data Matrix barcode. So what's the big deal? Here's one story. I got a package from the Philippines that comes with this kind of a barcode, and this package is not your courier's package(Fedex, DHL or whatever), it's a biscuit package from a taken-for-granted-but-well-known biscuit company in the Philippines, imagine a cheap Philippine biscuit company can easily adapt to that. I am researching about Data matrix for a different purpose. Until I found this box with a Data matrix on it! Wow! Cool! Awesome! Data matrix is more reliable and efficient than barcode and has broader range of applications, take for example snapping a datamatrix in a newspaper that is printed beside a ringtone advertisement with a handphone camera, it will then decode the Data matrix and sends a GPRS or SMS/WAP push to the wireless operator with instructions to download the ringtone and bill the subscriber, since Data matrix can be scanned at 20% contrast ratio it means that it's almost foolproof(like 99%). Unfortunately, some or maybe most systems in Singapore will have a hard time catching up on Data matrix due to the fact that most of the systems are tightly-coupled, codes are copy-and-pasted from one module to another just to make everything work on the first cut without really minding to look far, far ahead in the future. International POS providers and packaging specialists who has ready systems to be deployed will start doing the killing.

I am just spilling some few beans of opportunity. I'll collect my payback and go home soon.

23 April 2006

My fearless forecast: Linux demand in Singapore will grow

I was at Sim Lim Square the other day to bring one of our servers for servicing. This is related to my previous post about the RAID controller. I talked to one of the technical support engineer who was there busy fixing another customer's PC so I went in to their work area and told them about our RAID card issues, but due to time constraints (because it's almost closing time when we get there) I feel that our server will not be fixed in the same day which is not acceptable to us. While waiting at the work area I noticed that the engineer was installing Fedora Core 5 on an Asus blade server, Fedora Core 5 on Asus blade server! For wherever Linux World I came from this may sound preposterous. Fedora Core 5 are regarded as for sixth-graders not blade servers!

It's understandable that this is still Microsoft Country(because MS thinks for them!) but at least there's now a bit of a traction in terms of Linux acceptability. So I asked the engineer on how many Linux installations does he do in a week, he can't give an estimate or maybe he doesn't want to but he hinted that there's a lot and I asked who owns the blade server he's working on, he told me it's from a medium-sized company. In the end, my server wasn't fixed so I decided to replace the RAID controller with a more Linux-friendly one, the 3ware 9550SX(you see, great products really get to plug here). No magic required, it worked.

There's now an opportunity(in fact, many) and my palms are itching. The business of Linux Technical Support will soon be sprawling in this island like a "locksmith service" some of the most daring one's will even post stickers at your door with "LINUX INSTALLATION/SERVICE HP:9XXX-XXXX" alongside those locksmith stickers. Anyway, my number is ...

RAID Controller for CentOS 4.3

I don't have too much exposure on different RAID controllers for Linux. Of course, I am a software developer not a systems administrator or not even a hardware specialist. And I have wasted so many hours trying to make one of our RAID'ed servers to work and only to discover that the card is lousily supported. So here's my recommended hardware:

3ware 9550SX, it's S$700 and it's worth it! I have searched tons of forums trying to get our crappy RAID controller to work but what's echoing in the community is the 3ware 9550SX. See, those wasted hours are meant for development work.

18 April 2006

Ladies and Gentlemen, The IBM Websphere Application Server Community

Though a bit late at this time, it's nice to revisit how this thing has progressed in the past few months. But the days of experimentations are now over, it is now the dawning of the "full-court-press" GTD(Getting Things Done) days. To some it will be a renaissance period, to others it will be the dark days. There are so many exciting things coming out, new frameworks, new ways of doing old crap etc. etc. but the buzzword "Getting Things Done" is the most important part in today's Java-Whatever-Development. "WAS CE" as it is widely known, is still too young to be judged as to whether it will guide you naturally to the GTD path or "Need-to-know-more-and-waste-my-time" path. But I will nominate WAS CE to be the application server with the highest "GTD percentage" in the near future.

12 April 2006

Why RCP?

Eclipse Rich Client Platform has been gaining traction in some sectors of the Java development community. But it's really hard to justify the return of the client/server development nowadays, to convince technical managers who are deeply in to web development to gradually consider moving to rich client platform. Some of the people will always recourse to deployment costs that comes with the rich client as the strongest excuse which in some part a valid argument. If you're one of those "unplugged" within your development team and wants to jump in this rejuvenated platform here are the questions you need to answer:

  • Why do we need to switch?

  • What are the constraints?

  • What are the driving forces for the change?

  • What critical problems will it solve? For good?

  • How will I demonstrate?

Demonstration is always a big turning key in any argument. There's got to be one. This is how you will start your RCP ecosystem. By getting your hands dirty.

05 April 2006


I've been away from watching what's going on with Java Web development these days until I got my eyes on Waffle. Yes, yet another web framework. Ok, no XML tweaking in the name of annotations, feels like Rails, of course, it's Rails-like! But frankly, I am not yet motivated to use or adapt it. To me there's nothing really significant about annotations. If there's any framework that needs no XML tweaking, dang, that's the web.xml of the Java Servlet Framework every so often, it is the heart of every mistake Java web-based developers make. Do we need another web framework? No, I think what we need is for some smart-nuts to overhaul the Servlet specification.