Archive for March, 2008

Yesterday I attended a presentation of Zemanta, a recently launched “blogging apprentice”.

Currently only a Firefox plugin ( IE on the way, if anyone cares ), running on Blogger, Typepad and WordPress < 2.5. Its concept is simple: It gets the images from Wikimedia Commons, Flickr and various stock photo providers, and lets you pick them out and place them into your post. They are suggested based on contextual comparison of your text and the image’s title, description and articles they were used in originally.

From what i used it, i was satisfied. Looking forward for the WordPress 2.5 update.

Update:
Zemanta now works with WordPress 2.5, and i seem to have played my part in it =D

After couple of hours, here is it: A libcaca extension to Mjpg-Stream[singlepic=141,300,200,,][singlepic=140,300,200,,]

You need img2txt from the libcaca-utilities and /dev/shm to be mounted and writable (or change to another location)

/dev/shm is present on most systems and is a mounted filesystem located on the main ram.

The source is in past the page brake.

(more…)

The past few days i looked at ways could use my Logitech Quickcam Pro 5000 USB Webcamera. Skype 2 did a nice job using all its features, including to my surprise, the integrated microphone.

Installing the driver/module was a piece of cake, goto http://linux-uvc.berlios.de/ pickup the latest stable and do the usual compile&install, then load the module.

You may need to reboot at this point.

Now some testing was in order, pickup the luvcview viewer (you need libsdl-dev packages to compile). Make it, then you can run it just by ./luvcview ( if you have more video devices installed, you may need to set the according device it needs to read, see readme ) or install it with a quick ‘sudo cp luvcview /usr/local/bin/’ for later reuse.

Ok, i got the cam running, i can see myself, and i can chat to people with Skype, what next ? (If it didnt work for you at this stage, the next steps wont help you.)

(more…)

A new version of Inkscape, the SVG editor is available for Linux. Although a windows version is still in production, lets check out a bunch of new stuff in this version:

Top of the list is a neat perception tool, you can see its results in the top right of this web page’s design. The new, and long awaited bucket fill works just like any other bucket tool you might have used in any other program, but its advanced filling options might just give it the edge.

The tweak tool gives you a quick way to morph, shrink/grow attract/repel shaped without worrying about nodes.

The usual horizontal and vertical guidelines have become angled at 0/90 degrees. To change to a different angle, just double-click the guideline you want to change and enter the values.

Skeletal brushes are now possible with the new path on part effect. Anyone for engraving ?

Also, new built in stroke patterns and a interface to OpenClipArt.org.

As the last post was about the size of bash.org, this one is about xkcd, the famous comic site, a simple set of scripts and you get the whole set and a few stats:
Use script wisely, it’s a strain on servers.

#!/bin/bash
echo "Downloading 395 pages."
for i in `seq 1 395`;
do
	if [ -s "xkcd/$i" ]; then
		continue
	else
		echo -n "`date +%H:%M:%S`: Trying $i ..."
		lynx --source "http://xkcd.com/$i" > "xkcd/$i"
		echo -n " Done. Image:.. "
		wget -q -p "comics" -nH "http://imgs.xkcd.com/comics/"`awk 'BEGIN{FS="<img src=\"http://imgs.xkcd.com/comics/";RS="\" title="}/<img/{print $2}' "xkcd/$i"`
		echo " Done."
		sleep 2s
	fi
done
echo "All done."

This piece of code does sometihng special, it takes the name of the image and uses wget to download it.

$n=1;
$vse=0;
while ($n < 410) {
	unset ($fajl);
	$fajl=file_get_contents("original/".$n);

	preg_match_all("|
<p class=\"quote\">(.*)<b>#(.*)</b>(.*)
<p class=\"qt\">(.*)

|Us", $fajl, $out);
	$i=0;
	while (isset($out[0][$i])) {
		echo '('.$out[2][$i].")\n".$out[4][$i]."\n";
		echo $out[2][$i]."\n".$out[4][$i]."\n";
		$i++;
		$vse++;
	}
	$n++;
}
echo "\n(".$vse.")";

And a parser that makes the final big file of everything, coincidentally also making the comments easy to read.
Comics make the most part of the download, with ~22 MB.

And as usual, the download link: LINK (22mb), or email me for the data.

I spent the last few hours on a simple question, how large is the worlds largest irc quote database (bash.org) ?
Thinking specifically of the quotes themselves.

So first i had to get them all, a simple bash script was sufficient.

#!/bin/bash
echo "Downloading 409 pages."
for i in `seq 1 409`;
do
if [ -s "original/$i" ]; then
continue
else
echo -n "`date +%H:%M:%S`: Trying $i ..."
lynx --source "http://www.bash.org/?browse=$i" > "original/$i"
echo "Done."
sleep 10s
fi
done
echo "All done."

Please, do not use that script, it is a strain on the bash servers, instead you can grab the original files at the end of the article.
After a couple of hours that was done, and i had my next script ready as well;

$n=1;
$vse=0;
while ($n < 410) {
unset ($fajl);
$fajl=file_get_contents("original/".$n);

preg_match_all("|
(.*)#(.*)(.*)
(.*)|Us", $fajl, $out);
$i=0;
while (isset($out[0][$i])) {
echo '('.$out[2][$i].")\n".$out[4][$i]."\n";
echo $out[2][$i]."\n".$out[4][$i]."\n";
$i++;
$vse++;
}
$n++;
}
echo "\n(".$vse.")";

The last line is to make sure i got all of them, 20440 at the time.
Ran it with shell, and piped to “final”: php parser.php > final

So, the conclusion was, the size of bash.org is ~5 MB
This are the files if you want them: link. (or email me)