691394 - Implement overloads in idl

Status: RESOLVED WONTFIX

(Core :: XPCOM, defect)

Description

[Boris Zbarsky [:bzbarsky]]

See thread at http://groups.google.com/group/mozilla.dev.platform/browse_thread/thread/23fd83abca2601db#

Comment 1

[Bobby Holley (:bholley)]

So, assuming we don't want to invent our own mangling scheme, it seems like the VC++ reordering of vtable overloads makes this unfeasible to do in XPConnect.

Should we wait for the new DOM bindings?

Comment 2

[Benjamin Smedberg]

It shouldn't be a problem to use different names in C++, e.g.

OverloadedMethod1
OverloadedMethod2
etc...

Similar to [binaryname] already, although presumably automatic.

Comment 3

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Yeah, we could do.

Alternatively, I presume whatever reordering it does do is not completely insane.  We could try to determine what the compiler does and match it in typelib generation.  What I've found on the Internet suggests that the compiler just groups all of the overloads together, which isn't that bad (in fact, that's how I think we should write our IDL).

Comment 4

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

http://support.microsoft.com/kb/131104

seems to explain what it does, and it's only about 50% insane.

Comment 5

[Bobby Holley (:bholley)]

Realistically, I don't think this is likely to get done early enough to make it worth doing before we switch over to the new DOM bindings. Once we have a better idea of what they'll look like, we can decide how best to support this.

Unassigning for now.

Comment 6

[Boris Zbarsky [:bzbarsky]]

We might need syntax-level support for this as a prereq to the DOM bindings work.

Comment 7

[Bobby Holley (:bholley)]

(In reply to Boris Zbarsky (:bz) from comment #6)
> We might need syntax-level support for this as a prereq to the DOM bindings
> work.

My understanding was that we hadn't yet determined whether we're going to use XPIDL or a new WebIDL compiler for the new DOM stuff. Is that correct, or am I missing something?

Comment 8

[Boris Zbarsky [:bzbarsky]]

In the end, we're going to use something that takes WebIDL as input and outputs both C++ headers and whatever binding stuff we need, no?

Comment 9

[Bobby Holley (:bholley)]

(In reply to Boris Zbarsky (:bz) from comment #8)
> In the end, we're going to use something that takes WebIDL as input and
> outputs both C++ headers and whatever binding stuff we need, no?

Right. And since WebIDL != XPIDL, building syntax-level support for overloads into XPIDL isn't necessarily what we want.

Comment 10

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

I think we're going to drop a WebIDL parser into the tree, distinct from the XPIDL stuff.

Comment 11

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Our WebIDL binding support this.  XPCOM will continue to live in the dark ages.