Bug 146132 – Support releasing media sessions

• Status: RESOLVED
• Resolution: FIXED
• Component: Media

[Matt Rajca]

Implement the release method described in section 5.1.3 of the Media Session spec, as well as its helper algorithm from section 4.6 "Releasing a media session".

[Matt Rajca]

Created attachment 255155 [details]
Patch

[Darin Adler]

Comment on attachment 255155 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=255155&action=review

> Source/WebCore/Modules/mediasession/MediaSession.cpp:84
> +    while (!m_activeParticipatingElements.isEmpty()) {
> +        HTMLMediaElement* element = m_activeParticipatingElements.takeAny();
> +        element->pause();
> +    }

Could write this without the local variable:

    while (!m_activeParticipatingElements.isEmpty())
        m_activeParticipatingElements.takeAny()->pause();

> Source/WebCore/Modules/mediasession/MediaSession.cpp:98
> +    if (m_activeParticipatingElements.size() >= 1)

I think !isEmpty() would be better. Sometimes it’s better not to take specification language too literally! One or more doesn’t have to turn into >= 1.

[Radar WebKit Bug Importer]

<rdar://problem/21464030>

[Matt Rajca]

Committed r185885: <http://trac.webkit.org/changeset/185885>

[Matt Rajca]

(In reply to comment #2)
> Comment on attachment 255155 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=255155&action=review
> 
> > Source/WebCore/Modules/mediasession/MediaSession.cpp:84
> > +    while (!m_activeParticipatingElements.isEmpty()) {
> > +        HTMLMediaElement* element = m_activeParticipatingElements.takeAny();
> > +        element->pause();
> > +    }
> 
> Could write this without the local variable:
> 
>     while (!m_activeParticipatingElements.isEmpty())
>         m_activeParticipatingElements.takeAny()->pause();
> 
> > Source/WebCore/Modules/mediasession/MediaSession.cpp:98
> > +    if (m_activeParticipatingElements.size() >= 1)
> 
> I think !isEmpty() would be better. Sometimes it’s better not to take
> specification language too literally! One or more doesn’t have to turn into
> >= 1.

Thanks for the feedback. Fixed.