Welcome to the Digital Playground!
I hope you enjoy my constantly evolving website. It's my blog, personal web-coding test area, and file server. Please come by and get to know me! As for all my worldwide readers, a big hello goes out to you from the great state of California in the USA.
webmaster@ordorica.org - a.k.a. Hector Ordorica
Clear Channel
Clear Channel stream extractor
I've had a perl script located at http://ordorica.org/tune.pl that can get a Clear Channel stream, like KFI 640AM , and send you the .asx file for streaming. This way I can avoid Clear Channel websites which want to show ads during the stream, and also use a different audio player that supports windows media format.
They recently made an update to their website so the web based player now authenticates with javascript instead of just using HTML. So today I worked on updating my tune.pl code to support their new contrived authentication and "station id."
To use the script, just provide the channel's website. For example:
http://ordorica.org/tune.pl?station=kfi640
http://ordorica.org/tune.pl?station=kiisfm
The script returns an asx stream, which works with VLC, foobar with mslive plugin, and others.
You can get the script here:
http://ordorica.org/tune.txt
#!/usr/bin/perl
# Genasx Grabber
# Grabs the asx file from the ClearChannel radio stream.
use LWP::Simple qw(!head);
use CGI qw(:standard);
#the call paramater in the command line;
#tune.pl station=kfi640 in a terminal,
#or tune.pl?station= on a website
my($station) = param('station');
#print a blank page if the station param isn't given
print "Content-type: text/html\n\n" if(! $station);
#so far all ClearChannel streams seem to be .com
$content = get("http://www.$station.com/cc-common/ond....< More on this story >
Posted by Hector on Sunday, 01-Mar-2009 21:48 PM
More Bug Fixes
More Bug Fixes
After trying to deal with xhtml validation errors:
http://validator.w3.org/check?uri=referer
I started rewriting the Etomite news module again to automatically finish closing all html tags. I also adjusted it to show more of the story by default.
Then I ran YSlow and optimized my site to get a 89 (B) grade. I can't get an A without moving to CDN servers. This was worth it! My site is much more responsive, smaller, and renders without errors.
One of the easy fixes YSlow suggested that made the most difference was to set my Etags:
http://developer.yahoo.net/blog/archives/2007/07/high_performanc_11.html
By adjusting my .htaccess file on my Dreamhost server. I added the line:
FileETag MTime Size
And now the browser doesn't have to recheck every image on page reloads. It's actually a Dreamhost bug that causes it.
I also added:
<FilesMatch "\.(ico|pdf|flv|jpg|jpeg|png|gif|js|css|swf)$">
Header set Expires "Thu, 15 Apr 2010 20:00:00 GMT"
</FilesMatch>
To my .htaccess to set a longer expiration date on my files.
Overall I got my http requests down from ~5 on a reload to 1 (I have a random i....
< More on this story >
Posted by Hector on Sunday, 22-Feb-2009 04:51 AM
SunSpider 2/22/09
SunSpider on Chromium 1.0.165.0 build 10150
Wow, Chrome realy improved its javascript performance with a new version of their javascript engine v8 that now supports faster string processing. The results speak for it self. The Chromium builds are a bit buggy, but are damn fast if you are willing to give it a try.
Also, Windows 7 x64 works without the --in-process-plugins tweak with the later builds.
RESULTS (means and 95% confidence intervals)
--------------------------------------------
Total: 671.2ms +/- 10.3%
--------------------------------------------
3d: 87.2ms +/- 17.4%
cube: 23.6ms +/- 20.6%
morph: 37.0ms +/- 21.6%
raytrace: 26.6ms +/- 28.5%
access: 54.8ms +/- 18.2%
binary-trees: 3.8ms +/- 27.4%
fannkuch: 17.6ms +/- 25.3%
nbody: 21.6ms +/- 34.2%
nsieve: 11.8ms +/- 13.7%
bitops: 42.2ms +/- 23.2%
3bit-bits-in-byte: 3.2ms +/- 32.5%
bits-in-byte: 6.8ms +/- 20.0%
bitwise-and: 12.0ms +/- 19.4%
nsieve-bits: 20.2ms +/- 31.8%
controlflow: 2.6ms +/- 26.2%
recursive: 2.6ms +/- 26.2%
crypto: 36.4ms +/- 12.5%
aes: 12.2ms +/- 13.3%
md5: 13.2ms +/- 16.8%
sha1: 11.0ms +/- 13.8%
date: 97.4ms +/- 7.9%
format-tofte: 48.8ms +/- 10.7%
form.... < More on this story >
Posted by Hector on Sunday, 22-Feb-2009 04:09 AM
Fixed Bug
Fixed bug in News Parser
I've finally fixed a bug in the News Parser module for Etomite. I had to strip out some <span> and <fieldset> tags from the news stories to display them on the main page.
It's interesting the bug never showed up in Internet Explorer since it appears to insert its own </span> tags to fix the formatting. Anyways, I'll update the module a bit more and submit it back to the Etomite snippet library.
Posted by Hector on Wednesday, 28-Jan-2009 11:52 AM
Chromium Updater
Chromium Updater version 0.1
Today I present a simple batch file that I made to download the open source Chromium browser. For those who don't now, Chromium is the open source branch of Google Chrome's browser. It doesn't have a EULA (End User License Agreement) to agree too, unlike Chrome.
Warning, this is bleeding edge. You could download a build that just doesn't work, or a build that has tons of rendering errors. Report bugs to the dev team if you find any!
Chromium_Updater.exe [1495KB] [Windows XP, Vista, Server, Windows 7]
It's a simple script that uses wget (I've included the dependencies) to downloaded the latest build in Chromium's repository. Here:
http://build.chromium.org/buildbot/snapshots/chromium-rel-xp/
When a new build is available, wget downloads the build. I then append the changelog for that version to changelog-history.xml; in the case anybody is interested.
Finally, I extract and replace the chromium build in the directory (using 7zip , also open source).
Personally, I run it as a cron job. Use the task scheduler and schedule chromiu....
< More on this story >
Posted by Hector on Wednesday, 28-Jan-2009 11:39 AM
Opera 10 alpha
Opera 10 alpha on SunSpider for comparison.
My opinion:
I've been using Chrome since it has the nicest interface with Vista Aero. It's really become stable compared to it's initial release, and it even updates nicely in the background. It doesn't nag with multiple allow prompts like Firefox 3 to allow SSL exceptions. I use it's application mode for Gmail and Google Docs. It loads faster than a real application, and is accessible anywhere you have internet access. The new tab feature is the best, and no firefox plugins offer the same level of integration with user history and a search pane. I now bookmark everything of interest and just really on Chrome's ability to quickly find what I'm looking for based on a keyword.
I still use firefox though because of noscript which protects you from javascript and xss attacks. And it's almost as fast as Chrome's v8 engine. Plus you can somewhat replicate Chrome's look with the Chromifox2 theme. The Chrome theme really makes the tabs stand out.
Opera is the slowest in terms of javascript. But it's rendering is excellent, and fast on most sites. It did get stuck on gmail for me (it never finished loading).
Once Chrome has a "only load javascript from first party domain.com" then I'll switch permanently. But Chrome otherwise is fast, and loads instantly. I don't have to worry about sites crashing and taking out the whole browser.
RESULTS (means and 95% confidence intervals)
--------------------------------------------
Total: 2815.8ms +/- 5.9%
--------------------------------------------
3d: .... < More on this story >
Posted by Hector on Monday, 05-Jan-2009 00:33 AM
Sunspider
1/5/09 - Happy New Year
And a new set of benchmarks.
Chrome:
============================================
RESULTS (means and 95% confidence intervals)
--------------------------------------------
Total: 1068.2ms +/- 7.0%
--------------------------------------------
3d: 126.2ms +/- 9.7%
cube: 31.2ms +/- 13.0%
morph: 60.4ms +/- 27.2%
raytrace: 34.6ms +/- 11.0%
access: 76.4ms +/- 5.8%
binary-trees: 4.6ms +/- 14.8%
fannkuch: 24.0ms +/- 20.4%
nbody: 30.8ms +/- 14.1%
nsieve: 17.0ms +/- 5.2%
bitops: 57.0ms +/- 6.4%
3bit-bits-in-byte: 4.4ms +/- 15.5%
bits-in-byte: 9.6ms +/- 7.1%
bitwise-and: 19.2ms +/- 2.9%
nsieve-bits: 23.8ms +/- 13.5%
controlflow: 3.2ms +/- 17.4%
recursive: 3.2ms +/- 17.4%
crypto: 49.4ms +/- 8.1%
aes: 19.2ms +/- 40.5%
md5: 15.4ms +/- 22.5%
sha1: 14.8ms +/- 13.8%
date: 121.0ms +/- 18.2%
format-tofte: 63.0ms +/- 19.7%
format-xparb: 58.0ms +/- 16.5%
math: 98.6ms +/- 6.9%
cordic: 53.2ms +/- 1.0%
partial-sums: 33.4ms +/- 13.3%
spectral-norm: 12.0ms +/- 19.4%
regexp: 226.8ms +/- 13.5%
dna: 226.8ms +/- 13.5%
string: 309.6ms +/- 17.0%
bas.... < More on this story >
Posted by Hector on Sunday, 04-Jan-2009 21:52 PM
Chrome Firefox
Google Chrome 1.0.154.39 vs. Mozilla Firefox 3.1b2 nightly 12/19/08
I just did a quick Sunspider benchmark using these latest builds. A firefox nightly, and Chrome dev branch.
Here are the results:
Firefox:
============================================
RESULTS (means and 95% confidence intervals)
--------------------------------------------
Total: 1330.6ms +/- 6.3%
--------------------------------------------
3d: 179.4ms +/- 11.2%
cube: 61.4ms +/- 1.1%
morph: 31.2ms +/- 15.8%
raytrace: 86.8ms +/- 20.2%
access: 147.0ms +/- 11.9%
binary-trees: 44.0ms +/- 20.1%
fannkuch: 59.2ms +/- 19.0%
nbody: 30.6ms +/- 24.6%
nsieve: 13.2ms +/- 7.9%
bitops: 38.6ms +/- 9.3%
3bit-bits-in-byte: 1.8ms +/- 30.9%
bits-in-byte: 10.0ms +/- 0.0%
bitwise-and: 2.6ms +/- 26.2%
nsieve-bits: 24.2ms +/- 17.6%
controlflow: 48.0ms +/- 4.8%
recursive: 48.0ms +/- 4.8%
crypto: 63.2ms +/- 3.2%
aes: 38.8ms +/- 4.8%
md5: 17.8ms +/- 3.1%
sha1: 6.6ms +/- 16.8%
date: 223.4ms +/- 15.1%
format-tofte: 125.2ms +/- 27.1%
format-xparb: 98.2ms +/- 21.1%
math: 63.2ms +/- 11.4%
cordic: 32.0ms +/- 26.8%
partial-sums: 20.6ms +/- 5.4%
spectral-nor.... < More on this story >
Posted by Hector on Thursday, 18-Dec-2008 21:42 PM
Older News: