thaunknown / miru Goto Github PK

no clue how to reproduce, so maybe attempt to load the file again when it 404's?

check for compatible video files inside torrent file, throw error if none found

add support for batch files with nice GUI or smth, this could also make next video button work nicely i guess, augh parsing galore

when a video file is loaded for the first time it 404's, refreshing and downloading the torrent again fixes this

only store subtitles that are in the currently played buffer range since subtitles are streamed constantly anyways, and only in the currently selected track, this can potentially free up a lot of RAM.

get buffer amounts based on torrent file not video element

ditch bootstrap, improve the layout and general feel of UI

look into autoPIP, fucking documentation sucks peen

look into the play/pause display/status issue while buffering

store currently playing media data

its a fucking mess.... needs to pass media object/look for media object instead of passing string, would then remove the selected object too

only make requests for missing thumbnails instead of overwriting existing ones

maybe add an option for store to persist across sessions with drive caching? but should clean it up better.

-||-

show replay button when vid finished, show buffering icon

from best to worst to improve initial torrent fetch times

for latest releases

could possibly parse the subtitles after a video is downloaded and also display it from a blob for better performance?

since I'd need to get updated based on RSS feeds not notification servers idk how to get it to work...

fix downloading multiple torrents at once via local server

thumbnails get created when the video is buffering, probably shouldnt

add an option to store torrents in indexedDB as an alternative

scale,
AL sync

within first 10 seconds of video skip directly to 1:30
based on interaction for first 5 seconds skip

add an option to manually set anime names with anilist

nowPlaying breaks when autoname resolve fails

burn in subtitles could potentially render directly on the video image canvas instead of merging the video image canvas with the subtitle display canvas

now with oauth a thing, preferedname could be used, also a single function for requesting anime name would be better.

a preview on the seek/progress bar would be pog, smth like youtube has, but idk if its possible

update once to 0,0

add automatic synonym lookups if romaji name lookup fails

yeah, fuck knows how, either WebRTC peers or some sort of MediaStream but idk how

add torrent downloading

By overlapping a transparent canvas over the video and then streaming its content to a PiP it might be possible to display subtitles in PiP!

cuz this shit is a fucking mess

instead of constantly changing video time when moving the slider, change it once when the user lets go of the slider

support for .ass subtitles embedded in video files, probably should ditch the special effects and make it into normal subtitle tracks via .addCue(new VTTCue(timestart, timeend, "text"));

switch to halfmoon's progress bars instead of using the garbage i have currently xD

remove the video element and re-create it every time the source changes

figure out if its possible to detect what video data is currently loaded in the video element, then delete according torrent pieces, while making the torrent think they are downloaded, this could potentially have issues as I'm not sure if the browser unloads video pieces after some time [but I don't think it does], this would most likely kill seeding.... a good alternative would be storing the torrent inside the video itself? but thats kinda crazy

add an option for a custom RSS and regex

mostly css based mobile controls would b pog

-||-, would be nice

because why not

add advanced desc to settings on hover

only download torrent pieces needed for playback instead of the entire torrent [full video streaming]
this could greatly reduce RAM usage and allow for a lot of other things, but could also potentially kill seeding!

might not be possible tho

implement a home page > PWA.... this is gonna be painful....

generate the thumbnails based on video buffer instead of current video playback

look into the possibility of background seeding using SW