New Media Processor is Running

@Null
I post my videos here at 500K bitrate (64K for audio) (rendered using AV1 (libsvtav1)), should I up the bitrate to 2.85M or keep them at 500K going forward?
The encoders use a vbr so it should not matter what the source is.

Now this I am kinda surprised by. This is a format that will play on a website, with the right decoder.
I only ever whitelisted formats that were likely to run on browsers. I've said this before, it is a whitelist that has existed for sanity reasons. It IS NOT a limitation of the processor pipeline.
 
I only ever whitelisted formats that were likely to run on browsers. I've said this before, it is a whitelist that has existed for sanity reasons. It IS NOT a limitation of the processor pipeline.

I'm well aware. Note that I said nothing about the processing pipeline. I was just surprised that FLV wasn't in your whitelist since it's not uncommon.

@Null may I suggest the ability to control the volume of audio before hitting play because I dont want my ears to be brutally raped because it would be so awesome, it would be so cool
View attachment 8935480

You can, right click, click "Show all controls" and it will expose the volume setting before playback.
 
Original filenames are visible in the HTML. I don't remember these being exposed before, seems like a minor opsec risk for users expecting them to always be dropped.
1777570175542.png
 
Now how do we fix the bullshit where every youtube embed just blanks out with a 153 error. Admittedly that's probably just tor things but it's fucking annoying when the embed doesn't work and opening the link hits you with the looping captcha -> complete captcha -> randomly deciding that the captcha was redundant and that your network activity is suspicious -> loop that for about 5 times before the cunt works.
 
The volume slider does seem to be wrong, like from 100% to 10% it only lowers the volume 20% to 30% and the last 10% to 0% is what actually matters.
If AI did it then it must be a simple error with sound volume being calculated linearly instead of exponentially or logarithmically, in my experience AI likes to fuck up shit like that but it should be an easy fix.
 
Josh ASMR gave me fucking Nigger AIDS but the fact that I can't save the video to my phone (Firefox on a Google Pixel running stock Android) is still upsetting. Therefore, I'm joining in the chorus of phone fags who are complaining about the issue.

Other than that? Good shit! Thank you.

P.S.: I'll provide extra debug info if needed but seeing as the problem seems to be very common, I doubt it's necessary.
 
Admittedly that's probably just tor things
Yes, it is, and it cannot be fixed without just downloading every YouTube video posted on the site. Which I am of the mind to do... if... we had more money.

If the subscriber count stays high I promise you I will literally archive every single video ever posted on this site in every format and make it available through the player on Tor. But this is thousands of dollars.
 
Original filenames are visible in the HTML. I don't remember these being exposed before, seems like a minor opsec risk for users expecting them to always be dropped.
Filenames have always been available in certain contexts for all files and if you dox yourself by uploading a file literally named something that gets you doxed you're a retard faggot and science can't help you be less fucking stupid
 
I noticed that clicking the play-pause button or using the spacebar is immediate, but clicking the video outside that button does the same thing but has a short lag, maybe 100 ms.
I tested this on a different, faster computer, and it seems to be about the same. My guess is it's some JavaScript event handler that adds an arbitrary delay. Maybe I'll take a peek at the code later.

This is a very minor nitpick, but no shit is too minor.
 
Yes, it is, and it cannot be fixed without just downloading every YouTube video posted on the site.
I wouldn't say every video but any video in the main boards would make sense.
Original filenames are visible in the HTML. I don't remember these being exposed before, seems like a minor opsec risk for users expecting them to always be dropped.
I don't know exactly but that file name looks like it's from that fucking twitter media harvest extension which logs literally every download you do with it and the dev has admitted he just likes looking through the shit people download.
 
I tested this on a different, faster computer, and it seems to be about the same. My guess is it's some JavaScript event handler that adds an arbitrary delay. Maybe I'll take a peek at the code later.

This is a very minor nitpick, but no shit is too minor.
Sounds like classic webshit behavior to differentiate between single click and double click? That's what you get for using a higher level of abstraction than xlib like God intended.
 
Back
Top Bottom