YouTube Captions Search, 2. Comment Search #917

[ImprovedTube]

from #917

Eugene-Gubar/YCS <3 MIT

Firefox version: https://github.com/pancevac/ytsc-extension

hi @chirag127 🧢👋

Captions search: https://greasyfork.org/en/scripts/440153-youtube-ctrl-f-support/code

---

2. Generally, lets also check some of chirag's closed issues again(, who likes to close later) We can respect that but mark them "#unresolved" i https://github.com/code-charity/youtube/issues?q=is%3Aissue+sort%3Aupdated-desc+author%3Achirag127+is%3Aclosed

[raszpl]

for the comment search, Im not a fan of code that downloads something form YT servers https://github.com/pancevac/ytsc-extension/blob/b671e915b7cf51dd7c6a2923d3c83bbc1d0fd54d/src/utils/CommentResource.js#L15 replies.comments.map((comment) => (new ReplyResource(comment)).fetch).reverse() : in case of old video there might be thousands of comments and using this will trigger a storm of fetches.

Caption one is easy, small and to the point. btw caption/subtitles, one nice feature is automatically translating subtitles if not available in selected language. This is a one line of code to implement and saves 5 clicks I just noticed https://github.com/code-charity/youtube/issues/1647

[ImprovedTube]

Caption one is easy

👍 the extra specification in #1647 doesnt make your sentence wrong.

a storm of fetches. (hi! @raszpl!)

There is more to it. (Comments are big. The world spends ~ 1/4 of all time spent on websites on youtube.com alone. Comments can be filtered in more ways. Time-marks are very relevant: https://github.com/code-charity/youtube/issues/1565

• While users often only care in top comments/results, we should give them this choice ("load all" button. The "YCS Choice" - currently only avaiable to less than 100 000 people) Yet Youtube/Technnology is ready. Hopefully @sonigy (or @pancevac ) come here still.

• We could use 3rd party API / Archive for bulk fetches, maybe i.e. all comments from subscribed channels per user, to store locally permanently. We could even 'crowd/mesh-cache' (or run servers) & provide regular updates to archive.org (compare: https://docs.invidious.io/api/ & https://archive.org/details/archiveteam_youtube?tab=collection - both of which we can use and/or contribute to.)