Official LimeWire 4.1 requests thread

[verdyp]

Originally posted by Juggalo15
For more info try out the program it is up for download either at UTC or my site. This would be a milestone addition seeings how LW has has the had the original play right through no options to fast forward media player for it seems forever.

Although Limewire integrates a player, the main purpose of its presence is to allow easy preview of audio files downloaded. For best performance, there are tons of other players available. LimeWire has integrated an open-sourced Java player which is upgraded each time it gives additional performance benefits or reduced memory stress, but its development focus is not in adding CPU-intensive operations to detect spots in songs.
In some future, Limewire may include interfaces to interoperate with platform native players (that was done with iTunes on MacOS, but Windows Media Player or Real Player ot WinAmp interfaces may be possible, if there are some open-sourced interfaces that complies with the Limewire GPL.

Developing player components requires very specific and complex knowledge on codecs, and hardware acceleration support. This would be too much work for now for the LimeWire developers.

If AquaLime developers want to help LimeWire, they can contribute within the open-source project and share their experience with their component integration (but LimeWire requires now many compatibility and performance tests, and active support by the contributors to help solve issues, before such new code can be finally integrated into LimeWire and released; we don't want to include unstable and poorly tested components).

Some contributions (including my own ones) will not pass the tests phase or will be integrated in a unscheduled future release once the current planned priorities have advanced, and Limewire has understood how to support this new code.

In some future, my next works in Limewire will be to find the performance bottlenecks in the downloader classes (this part is one of the most complex classes with lots of cases to support interoperability with various servents, so I did not invest my time in optimizing it as development was quite complex; the automated tests suite for this part in the code is the longest one in LimeWire, and I prefer let LimeWire work on it), or to reduce again the memory footprint and VM stress when handling incoming Gnutella messages.