Tuesday, December 02, 2008

Semantic space

A semantic space explorer. Unfinished, but a great metaphor for exploring all kinds of databanks and databases.

Monday, November 03, 2008

Pricing of information

An interesting piece of research from one of my professors from INSEAD, Markus Christen: it seems that it is more profitable for information providers to offer lower quality information to their clients. This will force their clients to use multiple providers to arrive at a 'truth', and will also allow them to raise the prices of the (unreliable) information they offer. This works if there is a low correlation between the information offered by the providers, and if there is a relatively low number of providers.
Reminds me of IDG, Gartner, et al. Which incidentally are among the examples mentioned in the paper.

Friday, September 26, 2008


Not sure what to think about this, other than having my vanity reach new levels as I look at my face in a Blade Runner-esque virtual environment... thanks to ExitReality. The navigation metaphor is ok-ish, I had a similar idea for exploring databases; but I'd guess that most of the information-rich web sites aren't amenable to this kind of visualization.

Super cool toy, though.

Monday, August 18, 2008

Friday, August 15, 2008


Nice, Wordpress. To redirect a WP-hosted blog, one must pay $10/year. To use a custom CSS (I hate the font used by the current one I have there), it's another $15. I understand they have to make a living. They should understand I can use Blogger instead.

Applets once were embeddable

Ok most of my recent posts have actually been embedded scripts or badges from various external sites; long live mashups! Today however, I had to deal with a problem I thought had been fixed since 1996: embedding a Java applet in a web page. Such a simple task was made unpleasant by the need to have it done in XHTML and supported across various browsers, including IE, Firefox, and Safari.

Well, the OBJECT tag works slightly different in each of the above. And it is extremely poorly documented. If you want your JAR to reside in other directory than the one where the HTML file is, tough luck.

This is plain stupid. Why is it so hard to standardize such a basic premise of the internet? Ok Java applets aren't used that much anymore, but why did they have to go and make something that was once simple, complicated?

Friday, June 13, 2008

Landing page project for my personal web site

Courtesy of Gliffy.

Work in progress:


Editable chart:

Monday, June 09, 2008

Alexa Site Compare widget

...cannot be pasted here apparently. So it can be found here.

Sunday, June 01, 2008

Twitter's db maint

...is down for database maintenance. DBCC I guess? A bit funny to see yet another reason why the service is down. I can think of several ways to monetize it, but certainly not with the amount of downtime they are having.

Saturday, May 24, 2008

Widgets ecosystem

Allright, widgets are the new rage. Here are your options:

Then there are the 'social sites' plug-in widgets (eg for Facebook: Washington Post's collection of widgets). Also, offerings such as this one that provide the backdrop for widgets to work. Or, for a really useful collection of widgets, look here.

Finally, there are aggregator sites such as these: WidgetBox and SexyWidget.

A cool company that authors widgets.

Saturday, March 22, 2008

Blogger API

Using the Blogger API: of course, from JavaScript you are restricted by the same domain origin security requirement. Google has a nice API that allows JS to work... still trying to figure out a way to use this feature. The irony is that the demo I link to reads this very postings.

More programmable web

Lately the 'programmable web' has become more and more of a reality. Amazon, Google, and Adobe among others have made big contributions. Here are some:
  • Amazon S3
    • REST/SOAP based storage
  • Amazon SQS
    • SOAP based queuing
  • Amazon SimpleDB (still in beta)
    • similar to GoogleBase?
    • attribute-based semistructured repository
  • Amazon EC2
    • on demand virtual servers accessible via web services
  • GoogleBase
    • shared repository that can be queried using a SQL-type of language
  • GoogleGears Database
    • browser extension allowing the use of a local SQLite database (SQLite is just taking off: Adobe AIR also uses it)
  • GoogleGears WorkerThread
    • browser extension allowing multithreading
    • the threads are created locally; they are real OS threads spawned within the browser process
    • the threads communicate via text messages
On one hand, Adobe and Google bring web functionality to the desktop (stepping out of the browser sandbox: the AIR environment; GoogleGears); on the other hand, Amazon and Yahoo offer services that extend typical desktop or client/server functionality to the web.

I do not think the day is too far away when you will be able to compile a program online (if Office is going the software as service way, why not Visual Studio?) and deploy it to a virtual environment on the cloud, and run it there. Possibly, the only thing that is missing is a set of libraries to abstract the communication protocols (SOAP, XML-RPC, ATOM, etc... still too many).

Friday, March 21, 2008

Google Gears demo

Had some fun with Google Gears. To run this you need Google Gears installed, and it will not work with Safari.

Google Gears demo

Ruby and SQLite3 in Windows

If you need to run Ruby with SQLite in Windows, since you cannot set up the environment in the script like you can in Unix (#!), make sure you have the sqlite3.rb and sqlite3.dll in the same directory where the Ruby script is. Figuring this out caused me much grief.

Thursday, March 13, 2008

Postgres and Ruby

On Mac OS X the Postgres binaries are saved in /usr/local/bin. Here is a handy Postgres launcher that can be saved in the user’s directory (as postgres_start.sh for example):

su -l postgres -c "/usr/local/bin/pg_ctl start -D /Users/postgres/datadir"

Where datadir is the directory (owned by the postgres user) where Postgres has the data files.

To create a custom tablespace in Postgres, make an empty directory first and use that as the location for the tablespace in pgAdmin3.

Sample Ruby code to access Postgres; notice that the first line is needed to set up the environment; without it the require fails.

#! /usr/bin/env ruby
# original file src/test/examples/testlibpq.c
# Modified by Razvan
# Calls PL/SQL function in Postgres

require 'postgres'

def main
norecs = 0
pghost = "localhost"
pgport = 5432
pgoptions = nil
pgtty = nil
dbname = "razvan"

conn = PGconn.connect(pghost,pgport,pgoptions,pgtty,dbname)

res = conn.exec("BEGIN")
res = conn.exec("SELECT * FROM insertrt('another row')")

if (res.status != PGresult::TUPLES_OK)
raise PGerror,"RB-Error executing command.\n"

res.result.each do |tupl|
tupl.each do |fld|
norecs = norecs + 1

res = conn.exec("END")
printf("\nRB-Records: %i\n", norecs)

rescue PGError
if (conn.status == PGconn::CONNECTION_BAD)
printf(STDERR, "RB-Connection lost.")
printf(STDERR, "RB-Error:" )
printf(STDERR, conn.error)
end #rescue
end #end def main

main #invoke code

This calls the following Postgres plpgsql function:

CREATE OR REPLACE FUNCTION insertrt(data character varying)
id bigint;
id := 0;
id := id + 1;
INSERT INTO "RTable" ("Data", "Id")
VALUES(data, id);
RAISE NOTICE 'New id is %', id;
ALTER FUNCTION insertrt(character varying) OWNER TO postgres;
GRANT EXECUTE ON FUNCTION insertrt(character varying) TO postgres;

To execute this function do a SELECT * FROM insertrt( ‘parameter’ ). Interesting in the function, SELECT INTO variable. Also notice “ ‘s used to enclose field names and table names. The output of RAISE NOTICE is displayed by the Ruby console.

Since gems does not work with OS X Tiger’s Ruby, the Postgres adapter has to be built manually.