Featured Post
These days, I mostly post my tech musings on Linkedin. https://www.linkedin.com/in/seanmcgrath/
Wednesday, March 12, 2003
Python gets an O'Reilly Nutshell Guide
Python gets an O'Reilly Nutshell Guide. Every shelf should have one.
First and last anti-war-in-Iraq blog entry
The best, most beautiful, most poignant anti-war song ever written. (Read it but listen to it if you can - at least twice.).
The Island (Paul Brady)
The Island (Paul Brady)
Lets move XML-DEV
Google coming to Ireland
It seems Google are coming to Ireland. I guess that demotes us to the second biggest users of Python in this country:-)
Tuesday, March 11, 2003
Book != ISBN
Five books?. Well, only three really with two "packs". I've only ever seen the packs for sale in one book shop.
Recursions - I can't resist them
I note today that the invented term googlechopped is still far from being googlechopped.
On the radio I heard that some behind-the-scenes ruminations on proposed changes to Ireland's FOI (Freedom of Information) act, are now public thanks to an invocation of - yup you guessed in, the FOI act.
I really need to get out more :-;
On the radio I heard that some behind-the-scenes ruminations on proposed changes to Ireland's FOI (Freedom of Information) act, are now public thanks to an invocation of - yup you guessed in, the FOI act.
I really need to get out more :-;
The Web Services Scandal
Paul Kulchenko points to an EAI Journal article well worth a read. As is this which resonates strongly with my own beliefs:
Think about systems in terms of data flowing through processing nodes. Decompose the problem based on how data changes with respect to time. For too long now we have been subordinating the data to the high priest of algorithms. Putting all our efforts into cleverer and cleverer ways to decompose algorithms. A momumental waste of time and money.
Its the DATA stupid! :-)
Oh, and lest we forget - there is no such thing as "the structure" or "the semantics". These are local to a process in a time and place. At each flow point, you need transformational capabilities. Zen, flow and emergence in information models.
...That's probably why documents should be the primary focus and the way you send bytes and bits is of secondary importance. Build your vocabularies, define your semantics, structure your documents, then talk about SOAP, REST or pigeon mail to deliver them.
Think about systems in terms of data flowing through processing nodes. Decompose the problem based on how data changes with respect to time. For too long now we have been subordinating the data to the high priest of algorithms. Putting all our efforts into cleverer and cleverer ways to decompose algorithms. A momumental waste of time and money.
Its the DATA stupid! :-)
Oh, and lest we forget - there is no such thing as "the structure" or "the semantics". These are local to a process in a time and place. At each flow point, you need transformational capabilities. Zen, flow and emergence in information models.
Subscribe to:
Posts (Atom)