Ezra
Since the current metadata situation sucks, try adding in ICY metadata.
Really weird timestamps currently screw up kawa badly (way too slow/fast). We should probably switch to approximation via samples/sample rate, maybe only if we detect something really weird though.
Current status of client in terms of features - [x] General - [x] Sequential/rarest picking - [x] Downloading - [x] Download queue - [x] Custom location - [x] Initial hash...
As noted in #150 it seems like some form of event based scripting system would be useful to have and close out some feature requests. Current thought is to have...
Test coverage is dismal to say the least. Essentially only protocol parsing/encoding are tested (and poorly at that) and a few auxiliary function as well. Improving this will require many...
Looks like mio finally removed its restriction on passing FDs between polls on threads, we should migrate from amy now.
This is needed to remove sycli's dependency on a websocket framework which thus removes its dependency on a hardcoded TLS library.
There's a limited number of unit tests currently implemented for synapse, however these aren't nearly sufficient to catch significant regressions in behavior or simulate certain types of unusual situations. To...
DHT works currently, but not very well, or consistently. General investigation needs to be done into the DHT code to ensure that it can actually find peers more reliably.