74792 - helper to make code that asks for proxies more readable

Status: RESOLVED FIXED

(Core :: XPCOM, defect)

Description

[Dan Mosedale (:dmosedale, :dmose)]

It's a pain to have to get the proxy object manager yourself every time you want
a proxy object in a different function.  Since most users of proxies already
have a link-time dependency on libxpcom, I'm attaching a patch that exposes an
NS_GetProxyForObject.

Comment 1

[Dan Mosedale (:dmosedale, :dmose)]

Attachment: expose NS_GetProxyForObject, also update Contributors: per cvs logs

Comment 2

[Brendan Eich [:brendan]]

- * Contributor(s): 
+ * Contributor(s): Doug Turner <[e-mail]> (Original Author)
+ *                 Dan Mosedale <[e-mail]>

Isn't the MPL term of art "Original Contributor"?  I've seen others call out the
O.C. first, followed by a possibly-empty Contributor(s): list.

Why not put each Contributor on his own line?

I see yet more impl. in an IDL file, wahhh.  Is there no implementation-specific
header file that ties the interface to the one true impl.?  OTOH, if it's truly
the one impl (there can be only one), why am I whining.

Address these nits, and r=[e-mail].

/be

Comment 3

[Dan Mosedale (:dmosedale, :dmose)]

> Isn't the MPL term of art "Original Contributor"?  I've seen others call out 
> the O.C. first, followed by a possibly-empty Contributor(s): list.

I've seen Original Author format used far more often, and since it's part of the
Contributors list, it still conveys both contribution and authorship, which
seems clearer.

> Why not put each Contributor on his own line?

Huh?  They are... in my patch there is exactly one Contributor per line.

> I see yet more impl. in an IDL file, wahhh.  Is there no 
> implementation-specific header file that ties the interface to the one true 
> impl.?  OTOH, if it's truly the one impl (there can be only one), why am I 
> whining.

That's not the implementation in the IDL file; it's the only possible C++
forward declaration, given the IDL signature for the function being wrapped. 
The wrapper implementation is in the second file patched, nsProxyEventManager.cpp.

Comment 4

[Dan Mosedale (:dmosedale, :dmose)]

Filed bug 74792 on Orig Contrib vs. Orig Author stuff.  Brendan sez that I
should now just find an sr= and checkin.

Comment 5

[Dan Mosedale (:dmosedale, :dmose)]

Er, bug 75813, I meant.

Comment 6

[Chris Waterson]

sr=waterson

Comment 7

[Dan Mosedale (:dmosedale, :dmose)]

Fix checked in.