HTML5 Video Tuner - disable HTML5 video and/or MSE per domain

[JustOff]

Will you update this extension to work with PM 27.4?

It will work as before, no updates are required. But I hope that with the new MSE code you will not need it unless you are using it to ban HTML media on specific sites completely.

[Tomaso]

Do you still use the extensions?

I had some problems with MSE when using PM v27.3:
https://github.com/MoonchildProductions ... ssues/982/
https://github.com/MoonchildProductions ... -300892662
https://github.com/MoonchildProductions ... -304488253
..but so far, everything seems to work just fine with PM v27.4.
Still, 'HTML5 Media Tuner' is good to have if I should run into problems again.

BTW; If you want to install the latest PM version, the final v27.4 builds for Windows are already available on the server.
Simply change the 3 to a 4 in the official download links.


--

It will work as before

You're right.
I simply forgot that the icon kinda "hides" itself in the lower right corner of the Status Bar by default!

[LL25255252]

BTW; If you want to install the latest PM version, the final v27.4 builds for Windows are already available on the server.
Simply change the 3 to a 4 in the official download links.

Thank you @Tomaso
As to MSE Issues I tried the links you've posted but they seem to work OK for me (PM 27.3 32bit Win 7 x64).

Will you update this extension to work with PM 27.4?

It will work as before, no updates are required. But I hope that with the new MSE code you will not need it unless you are using it to ban HTML media on specific sites completely.

Can you give us an example where and why you use it "to ban HTML media" (which sites for example) ?
Thanks in advance.

[Tomaso]

I tried the links you've posted but they seem to work OK for me

Yeah, like I've pointed out before, the issues indeed seemed to be HW dependent somehow.
If you have a Nvidia graphics card, pay notice to what I wrote under "Workaround #3" here:
https://github.com/MoonchildProductions ... -293193601

--

which sites for example

The sites that I mentioned in the reports that I linked to in my previous post.

[Moonchild]

This extension will very likely lose most of its use as MSE compliance is being completed in the browser itself and MSE can simply be *on* for all sites.

[Tomaso]

Yeah, well, whilst MSE do appear to work just fine under "usual" circumstances now, PM 27.4 still fails plenty of the MSE Conformance Tests here:
https://yt-dash-mse-test.commondatastor ... /2017.html
..but I guess perhaps some of these issues will be remedied once WebM support is in place for PM's new MSE code?

[LL25255252]

@Tomaso, thanks again for all the links but for me everything still works fine (yet I Don't have Nvidia graphics ) .
PM 27.3 (32 bit) win7 x64 with MSE ON.

Will you update this extension to work with PM 27.4?

It will work as before, no updates are required. But I hope that with the new MSE code you will not need it unless you are using it to ban HTML media on specific sites completely.

Can you give us an example where and why you use it "to ban HTML media" (which sites for example) ?
Thanks in advance.

@Tomaso, my question a few posts before was directed to @JustOff

[trava90]

PM 27.4 still fails plenty of the MSE Conformance Tests here

If you run the 2015 version the current unstable build passes all but one test. The later versions include tests for new/changed parts of the MSE spec that Pale Moon doesn't support yet.

[JustOff]

Can you give us an example where and why you use it "to ban HTML media" (which sites for example) ?

Personally, I do not use it for this purpose, but someone may prefer flash over HTML5 media or allow it only on selected domains.

[Moonchild]

PM 27.4 still fails plenty of the MSE Conformance Tests here

If you run the 2015 version the current unstable build passes all but one test. The later versions include tests for new/changed parts of the MSE spec that Pale Moon doesn't support yet.

The funny thing is, the 2015 tests also pass in 27.3.0 except one. How reliable are those tests?

[LL25255252]

PM 27.4 still fails plenty of the MSE Conformance Tests here

If you run the 2015 version the current unstable build passes all but one test. The later versions include tests for new/changed parts of the MSE spec that Pale Moon doesn't support yet.

The funny thing is, the 2015 tests also pass in 27.3.0 except one. How reliable are those tests?

@Moonchild, do refer to the tests in the link @Tomaso posted - https://yt-dash-mse-test.commondatastor ... 9852623510 ?
Can you please make it easier for me and tell me the number/name of "the one test that failed in 27.3.0." ?

[Tomaso]

tell me the number/name of "the one test that failed in 27.3.0." ?

Two 2015 tests fails in PM v27.3:
#45: AudioWithLargeGap
#57: XHROpenState

In PM v27.4, only one 2015 test fails:
#57: XHROpenState

If you click on "Run All", all test will be executed one after another.

[Moonchild]

45 fails for me "45:AudioWithLargeGap: Failed with "Assert failed: Source buffer number is (1) which should be (2)" "
Seems you're correct on the XHROpenState failing too (on both!) without any specific error, though.

Even funnier is that the 2017 tests give the exact same result in release as they do in unstable, yet unstable is much more compliant and in line with the spec (as evidenced by solving real site playback issues)... How do you explain that?

[Tomaso]

How do you explain that? :)

You're probably right about these tests not being totally reliable.
At least for me, MSE seems to work just fine for practical use in PM v27.4, and I haven't encountered a single issue yet!
trava90 has done a great job with it!! :)
Still, it probably wouldn't be such a bad idea to try and get PM to pass as many of those 2017 tests as possible.

[Moonchild]

Why would I use obviously unreliable tests as a criterion?

Also, we tend to be reactive to what is actually in use, and not blindly implementing whatever a spec writer comes up with (because that's the #1 cause of code bloat)

[Tomaso]

Why would I use obviously unreliable tests as a criterion?

Well, for one thing, there's another possible explaination for the identical unstable vs stable results..
Despite all the changes in PM's MSE code, perhaps those specific issues simply hasn't been addressed yet?

Also, those tests are commonly refered to in both Chromium's and Mozilla's dev. pages, and seems to be used as a standard reference.
Most browsers doesn't seem to have much of a problem running any of those tests at this point.
For instance, SeaMonkey v2.46 only fails one of the 2017 tests (#92: AppendMultipleInitOpusAudio).

[Moonchild]

Also, those tests are commonly refered to in both Chromium's and Mozilla's dev. pages, and seems to be used as a standard reference.
Most browsers doesn't seem to have much of a problem running any of those tests at this point.
For instance, SeaMonkey v2.46 only fails one of the 2017 tests (#92: AppendMultipleInitOpusAudio).

Mozilla follows Chrome anyway. it's what they do. So of course they will refer to Google's tests and implement whatever is in them.

Still, of course I'm not averse to improving spec compliance (quite the opposite), I just doubt that it's actively in use -- with all our "failures" we are still able to stream just about anything MSE that isn't DRM, as it is now. So I have to wonder if it's not being given disproportionate priority.

[Tomaso]

Mozilla follows Chrome anyway. it's what they do.

Yeah, I agree.
My point though, is that deviating from the accepted standards for those major browsers these days (right or wrong) might be proven futile over time.

[adesh]

Moonchild has already said it, I'd just add this:
We should not try to pass some ridiculous test but rather fix real problems at hand to create a better browsing experience. If spec* is followed and the issues with the websites are resolved, passing the tests will come naturally.
Also we need not implement each and every feature, just the right set. That is in line with the philosophy of Pale Moon.

* As everyone knows Google is using it dominance to dictate the spec these days.

[T2001]

Been a while since I posted last, but I have a request for a new toggle option for the HTML5 tuner.

Would it be possible to add an option that toggles media.mediasource.enabled between true and false?
I've found that ever since YouTube killed off Flash, altering this option is the only way I can disable DASH playback on YouTube, and thus force it to fully buffer HTML5 video.
Previously, I'd been using the option to toggle DASH playback in YouTube Center to disable it, but that hasn't worked for me in quite some time.
There is a FireFox add-on (YouTube without DASH Playback) that is supposed to add a toggle button to the toolbar to which toggles media.mediasource.enabled between true and false, but unfortunately, it's an SDK based add-on, and won't install on PaleMoon.

I would just leave media.mediasource.enabled set to false all the time, but I've found that this breaks HTML5 video playback on some other sites.
Seeing as how I use your add-on a lot to effectively white-list which sites can use HTML5, I was hoping maybe you could add the option for enabling/disabling media.mediasource.enabled on a per domain basis, as this would allow to have it disabled on YouTube, but enabled on other sites if it causes video playback problems.
If it's not possible, then no worries, as I can just jump in and set media.mediasource.enabled back to true on the odd occasions I need to, but having a toggle option in the toolbar as I do with HTML5 would be much more convenient.

Thanks in advance, and Merry Christmas