The comment position seems a little off-kilter here. Maybe add a newline or tab? Most likely a newline.

Fixed to match other instances.

Assignment inside of the call is attempt to save some size?

Yes, but it's not new. Saves 7 bytes here.

With the

function adoptValue( value, resolve, reject ) {

    // Check for promise aspect first to privilege synchronous behavior
    if ( value && jQuery.isFunction( value.promise ) ) {
        value.promise.call( value ).done( resolve ).fail( reject );

    // Other thenables
    } else if ( value && jQuery.isFunction( value.then ) ) {
        value.then.call( value, resolve, reject );

    // Other non-thenables
    } else {

        // Support: Android 4.0 only
        // Strict mode functions invoked without .call/.apply get global-object context
        resolve.call( undefined, value );
    }
}

There is not diff in the size?

Promises/A+ requires accessing .then only once; if ( jQuery.isFunction( value.then ) ) { value.then.call(…); } is prohibited (cf. Notes 3.5).

Why we don't fail any tests in this case then?

@markelog: Commenting it out won't fail any tests because Promises/A+ only specifies then, and we only define a promises-tests adapter for jQuery.Deferred. I can define a jQuery.when adapter, put in a QUnit test of jQuery.when( pathologicalThenable ), or just let it be, but the line MUST take that form if we guarantee proper treatment of any thenable conformant with Promises/A+ (which, as of 3.0, we do).

And to clarify: if ( jQuery.isFunction( value.then ) ) { value.then.call(…); } is prohibited because it accesses value.then twice, and the results of those accesses won't be identical in the presence of a pathological getter like https://github.com/promises-aplus/promises-tests/blob/0d409df278345753c9527539a5558db67d42bb02/lib/tests/helpers/thenables.js#L29 .

or just let it be, but the line MUST take that form if we guarantee proper

I think such changes should be supported and confirmed by the tests, otherwise all of it could be seen as speculation or subjective interpretation, it seems this discussion revealed luck of it, so i'd say either in this pull or right after this one, we need to add those tests for jQuery.when.

Done.

hmm, why?

Funny you should ask... the test uses jQuery.when( elem ).done(…), and this PR updates the single-argument form of jQuery.when to use .then() for unwrapping secondary thenables (e.g., jQuery.when( jQuery.Deferred().resolve( thenable ) )). Single-argument jQuery.when becomes necessarily asynchronous as a side effect, and we therefore require an extra clock tick to invoke the callback.

ooo, so we have animate(..., 50), but because when.done is now async we need value more then 50, interesting, you probably were surprised a bit :)

Yes.

Single-argument jQuery.when becomes necessarily asynchronous as a side effect

It feels to me like placing a distinction between single argument and multiple argument is a really bad idea. I opened #3100 about this.