史蒂文 :: Release 2.0

I haven't seen this one, but this is from a visit by Ali G to Harvard in 2004. Oddly i found it while looking into some Semantic Web stuff - love to see a personalization engine give me those kinds of results!

"To fink dat so many great people has been educated ere like Lyndon Banes
Johnson, or as he is better known - JFK, George Clinton was also ere I fink ,
and de one before him, and also...William Tell - is he one of your lot,
probably, and dat bloke wiv de hat, but most importantly dat really fit honey
from Star Wars - if u iz out dere, me'd love to - me iz stayin at de Best
Western Hotel - me's got a really nice room, altho since dis morning dem has put
a parental lock on de tv. "

 read 0 comments | 

I pretty much write everything in C#. For years I was annoyed that some really cool ideas were coming out of the Java community and the effort required to really learn C++ and then be able to learn the cool things you could then invent to do with it often put me off.

I did write Java way back in 1996 - in fact i was in a Computational Physics class that day it was released. Back then it was awful and of course as come a long way. However, I never kept up and although I liked the langauge my first job was in a Microsoft shop and so that's the direction my career took.

I primarily used Xml and procedural code was often written to suppport that, but when C# arrived, it was the first time i really thought of any programming language as a primamry tool for my work. Like Java, it's really easy to do some vey cool things.

However, the last while has seen me read more about Patterns that i can ever remember and now I see patterns in everything. Of course i've been using them for a while without really understanding the plumbing, but now I am starting to really understand them in depth (my day job has had me implementing them in a real application).

The GoF book has been real interesting (and was even pre-Java). However what has appealed most is that communities of programmers can think of problems in an abstract way and so the implementation becomes less important. I now read a books in the C# and Java worlds, knowing it's often a case of simple conversion.

Recently I've been thinking about Xml patterns and more. I've not had a great time to think in detail, but I wonder whether these kinds of abstract patterns can be applied to Xml, Xslt, XQuery and so on and to what extent. In other words, is their easier ways of viewing the real world and realizing a set of patterns in many domains and implementations that could be combined to make problems much easier to solve.

I'd like to hear from anyone who has been doing work - or knows of work going on in these areas.
 read 0 comments | 

Interesting discussion on Xslt, XPath and XQuery over at TheServerSide.

I posted my own thoughts.

As much as I have seen people struggle with Xslt, it has ofetn been templates and their lack of knowledge around XPath. The complain that they never know when something is going to happen, but that is usually because everything is written as "//element" and with very little control over what the template is actually supposed to match. I have often seen a single tempate matching the root and then a shed load of HTML in the middle. That's not because there is a problem with Xslt - that's because they don't get how to properly use templates and XPath.

I do and i'll keep fighting the cause (special note to my friends over at ACS!).


 read 0 comments |