Browsers which don't support CyTube's limited subset of
generally-supported codecs probably aren't worth warning about.
1Mbps is way too low of a threshold to warn about bandwidth, but even if
the threshold for warning were raised, it's probably still not that
useful.
The use of the channel library as a cache for metadata to avoid
re-requesting metadata for known media is an optimization that dates
back to 1.0. However, it doesn't have any TTL, is prone to bugs, and is
of dubious value.
This commit ignores the results of the library check when queueing a new
video, opting to always re-request the metadata. This fixes a few bugs:
* Google Drive metadata being lost when storing in library
* Streamable metadata being lost when storing in library
* Videos in the channel library that are now unavailable on their
source website being queueable and then failing to play (e.g. deleted
YouTube videos).
In its place, a small fail-open check is left behind to emit metric
counters on how many queues would have been cache-hits, to provide
insight into whether a proper caching solution (i.e. one not tacked on
top of the library) would be worth pursuing or not. This will be
removed eventually.
Fixes#681. Technically, resending the entire userlist is not
necessary; it would be sufficient to resent setUserMeta, but there's not
currently a bulk frame for that so sending the userlist is probably more
efficient.
Camo: https://github.com/atmos/camo. This has a couple advantages over
just allowing images to be dumped as-is:
- Prevents mixed-content warnings by allowing the server to proxy HTTP
images to an HTTPS camo instance
- Protects users' privacy by not exposing their browser directly to
the image host
- Allows the camo proxy to intercept and reject bad image sources
(URLs that are not actually images, gigapixel-sized images likely to
DoS users' browsers, etc.)
Whitelisting specific domains is supported for cases where the source is
known to be trustworthy.
User.prototype.refreshAccount was responsible for multiple race
condition bugs as well as inefficient duplication of DB queries in an
attempt to correct such race conditions.
It has now been replaced by a more reasonable model:
* Global user account information and aliases are fetched in parallel
on socket connection
* Channel rank is fetched when the user tries to join a channel
The previous commits do not handle all of the edge cases of #583
appropriately. This is a short term solution that will work, but is not
as efficient as it could be. The whole refreshAccount function needs to
be reconsidered and replaced with a more sane way of handling atomic
updates to the user's account state.
Instead of emitting frames to each individual socket, group them into
socket.io rooms of people who can see hidden poll results and people who
can't, then just do 2 broadcasts.
If clients call it quickly in succession with large playlists, it can
cause node to get stuck stringifying socket.io frames and cause an out
of memory crash.
CyTube has been crashing recently due to things attempting to release
the reference after the channel was already closed (apparently the
uncaughtException handler isn't called for this?). This newer
implementation keeps track of what is ref'ing and unref'ing it, so it
can log an error if it detects a discrepancy.
Also changed the server to not delete the refCounter field from the
channel when it's unloaded, so that should reduce the number of errors
stemming from it being null/undefined.
If two people tried to play the same playlist item, before the playlist updated, it would delete instead of playing.
The same would also happen if the play button was double-clicked instead of single-clicked.
Also, the active item's play button functioned as a delete button.
Fully tested. Still removes the item (if it was added as temporary) when it finishes playing, or if the play button of a *different* item is clicked.
