Zusammenfassung
In this talk, we examine how different approaches to sound input and output handling are currently used in the free software landscape. We discuss kernel interfaces like OSS and ALSA, sound server approaches like artsd, ASD and ESD and MAS and requirements by applications or application suites like Gnome, KDE, Wine and XMMS. We proceed to show using a simple example how different interoperability problems arise from the very heterogenous landscape, which affect both, developers and end users. Finally, we suggest solving these issues by introducing an abstraction for various sound backends, and present an implementation called CSL (common sound layer).
CSL is already supported by some application and has been intensively discussed in interoperability meetings between the GNOME and KDE community, as solution for a common sound layer standard for sorting out interoperability issues. With incompatible sound servers and solutions around, like aRts, ESounD, ALSA and MAS, the issue appears to be important to the whole free software community.
Über den Autor
Tim Janik studiert Informatik an der Universität Hamburg und ist in diversen Freien Software Projekten involviert, wie GIMP, GTK+, GNOME und BEAST. Zur Zeit konzentriert sich seine Arbeit auf Musiksynthese und Toolkitintegration.
