mirror of
https://github.com/tuxbox-fork-migrations/recycled-ni-neutrino.git
synced 2025-09-02 10:21:04 +02:00
0973d86ff9
* Various EPG related strings (language- and country keys, audio descriptions) come with a pretty low variance. Cache them, and use an index instead of allocating redundant strings for each and every SIevent.
* Storing content classifications in native format instead of std::string removes quite some memory overhead, too.
This saves about 250 bytes of RAM for every event cached, resulting in a considerably reduced sectionsd memory footprint.
Conflicts:
src/eitd/xmlutil.cpp
Origin commit data
------------------
Commit: 0bcff8b930
Author: martii <m4rtii@gmx.de>
Date: 2014-09-14 (Sun, 14 Sep 2014)