The new converter (#1151) uses member objects to operate on data from the .dat files. While these provide the necessary functions for conversion, they are also significantly larger than the original dat file entries (20x - 400x the size). This results in a comparatively high memory usage, especially for DE2 which uses ~1.6 GB of memory for all dat file entries. Note that the converter does not

I think there is an error in the FrameLimiter implementation logic for the Sleep strategy, unless I'm misunderstanding something.

The docs say (emphasis mine):

Sleep will call thread::sleep with a duration of 0 milliseconds until the frame duration has passed.

In the code, the Sleep strategy is called with a zero duration:

const ZERO: Duration = Duration::from_millis(0

When working on a fix for #2755 I noticed that Git tag parsing is sub-optimal performance wise - to determine changeset/date for each tag a new command is spawned via Executor in buildTagEntry(). For repositories with large set of tags (e.g. Linux kernel) this could be taxing.

Either convert this to parallel execution (thread pool) or find a way how to get all the data in single command.

A game's code is supposed to have its own error type (that can wrap over ggez::GameError) but you are forced to unwrap/expect all your errors in EventHandler's draw and update methods (because you can't turn them into GameError), right? Is this documented somewhere?

Related to ggez/ggez#421

hi,
as you know, in SoLoud, the number of filters are limited
we should implement more like different reverbs, fir and irr filters, (these could be used to implement HRTF support), Chorus, One Poll, One Zero, Pole Zero, Two Pole, Two Zero, etc
a library exists called stk under zlib license which already implemented these maybe we can implement some of these out