Free 0.9a software for linux

Borealis provides a "better feel" through use of the subtle aural notifications.
The goal of this sound theme is to provide "better feel" through use of the subtle aural notifications. The test I always have in my mind while designing the desktop sounds is to imagine having half-dozen (or more) computer stations in the same office or studio and consider whether their simultaneous use (and resulting desktop sounds) would pose as distraction on users and their productivity. Therefore, in order to minimize the "distraction factor" this theme associates many of the common desktop events with brief, mostly non-pitched, but nonetheless pleasing sound sources (a.k.a. effects).
All of the sounds included have been composed (if applicable), recorded, and mastered by me (although a few of them bear resemblance to other OSs sounds). Some of them are leftovers from my previous compositional and production endeavors, while others are entirely new creations.
All sounds are designed and mixed at CD-quality (stereo, 16-bit, 44.1KHz). Ogg version is distributed as 192Kbit stereo files.
Main features:
- Includes sounds for:
- KDE WM
- KDE System Notifications
- Kdevelop
- Kopete
- KDE Sysguard
- KMail
- Konsole
- Proxy configuration
- K3b (missing 1 sound)
- 4 startup sounds
- 2 exit sounds
- Automated installer and uninstaller
Enhancements:
- Couple improvements to the install and uninstall scripts (thanks to Kirk Strauser for the heads-up!)

Video Contact Sheet *NIX or vcs for short, is a script that creates a contact sheet (preview) from videos by taking still captures distributed over the length of the video.

The output image contains useful information on the video such as codecs, file size, screen size, frame rate, and length.

Notes:

Using -i or -n with values that lead to the same number of vidcaps doesnt necessarily yield exactly the same vidcaps, youll have to play with them for best results.

Example: With a video that lasts for 22min 14secs, -n 6 will create the six vidcaps:

3:22, 6:44, 10:04, 13:28, 16:50 and 20:12, while -i 3m will create
3:00, 6:00, 9:00, 12:00, 15:00 and 18:00.

As you can see, -n distributes the vidcaps all over the video while -i N starts at N and goes in increments of it up to the end.

You can set the aspect ratio of the video with -a. It is only required for anamorphic material and discouraged for the rest.

It can take either a floating point number (like 1.33 or 1.778) or a fraction (like 4/3 or 16/9). The latter is slightly preferable.

Text::Bloom can evaluate Bloom signature of a set of terms.

SYNOPSIS

my $b = Text::Bloom->new();
$b->Compute( qw( foo bar baz ) );
my $sig = $b->WriteToString();
$b->WriteToFile( afile.sig );
my $b2 = Text::Bloom::NewFromFile( afile.sig );
my $b3 = Text::Bloom->new();
$b3->Compute( qw( foo bar barbaz ) );
my $sim = $b->Similarity( $b2 );
my $b4 = Text::Bloom::NewFromString( $sig );

Text::Bloom applies the Bloom filtering technique to the statistical analysis of documents.

The terms in the document are quantized using a base-36 radix representation; each term thus corresponds to an integer in the range 0..p-1, where p is a prime, currently set to the greatest prime less than 2^32.

Each quantized value is mapped to d integers in the range 0..size-1, where size is an integer less than p, currently 2^17, using a family of hash functions, computed by the HashV function.

Each hashed value is used as the index in a large bit vector. Bits corresponding to terms present in the document are set to 1; all other bits are set to 0.

Of course, collisions may cause the same bit to be set twice, by different terms. It follows that, if the document contains n distinct terms, in the resulting bit vector at most n * d bits are set to 1.

The resulting bit string is a very compact representation of the presence/absence of terms in the document, and is therefore characterised as a signature. Moreover, it does not depend on a pre-set dictionary of terms.

The signature may be used for:

testing whether a given set of terms is present in the document,
computing which fraction of terms are common to two documents.

The bit representation may be written to and read from a file. Text::Bloom prepends a header to the bit stream proper; moreover, whenever the package Compress::Zlib is available, the bit vector is compressed, so that disk space requirements are drastically reduced, especially for small documents.

The hash function is obviously a crucial component of the filter; the reference implementation uses a radix representation of strings. Each term must therefore match the regular expression /[0-9a-z]+/.

There are quite a few viable alternatives, which can be pursued by subclassing and redefining the method QuantizeV.