Remove extension automation support that was used only by CEEE.

Remove extension automation support that was used only by CEEE.

BUG=none
TEST=all automated tests pass

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=80626

[Jói, 2011-04-05 15:58:31]

Hi guys,

Amit: Please review chrome_frame/...
Siggi: Please review everything else.

Mostly code removal.  Amit, look out for one unit test where I changed the
functionality (it used to send a couple of sync messages, now it sends only one
- looked to me like that was OK but I may be missing something about what it was
intended to test).

Cheers,
Jói

[Sigurður Ásgeirsson, 2011-04-05 16:57:16]

lgtm

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:118: [uuid(812C601D-D787-49f2-BF05-6A460B44D2B3)]
I don't think this is necessary. As the interface here is a dispinterface.
There's no harm, that I can think of, however.

[Jói, 2011-04-05 17:04:17]

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:118: [uuid(812C601D-D787-49f2-BF05-6A460B44D2B3)]
On 2011/04/05 16:57:16, Ruðrugis wrote:
> I don't think this is necessary. As the interface here is a dispinterface.
> There's no harm, that I can think of, however.

Yeah, I didn't think so either.  Just going by the book, so to speak :)

[Paweł Hajdan Jr., 2011-04-05 17:26:32]

Drive-by with a tiny nit, no need to wait for me.

I'm very very happy this cleanup has happened. Thanks!

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
File chrome/common/automation_messages_internal.h (right):

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
chrome/common/automation_messages_internal.h:1136: // DEPRECATED MESSAGE - But
we must leave this comment and message so as
Could you add "TODO(phajdan.jr): Remove when the reference build is updated" to
all deprecated messages?

[amit, 2011-04-05 18:00:24]

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:14: uuid(25248626-A306-4857-A13C-4AD84FF15E10),
Tommi pointed out that any existing users embedding GCF will break with this
change.  Maybe we keep the methods (perhaps rename them) and return E_NOTIMPL?

[Jói, 2011-04-05 18:47:10]

Thanks guys, will upload a new version soon to address the comments.

Amit, see one question below just to confirm.

Cheers,
Jói

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
File chrome/common/automation_messages_internal.h (right):

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
chrome/common/automation_messages_internal.h:1136: // DEPRECATED MESSAGE - But
we must leave this comment and message so as
On 2011/04/05 17:26:33, Paweł Hajdan Jr. wrote:
> Could you add "TODO(phajdan.jr): Remove when the reference build is updated"
to
> all deprecated messages?

One of the messages' comments didn't have room for that, so instead added the
following to just one of the messages I deprecated:


// TODO(phajdan.jr): Remove when the reference build is updated (this and
// all others marked "DEPRECATED MESSAGE").

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:14: uuid(25248626-A306-4857-A13C-4AD84FF15E10),
On 2011/04/05 18:00:24, amit wrote:
> Tommi pointed out that any existing users embedding GCF will break with this
> change.  Maybe we keep the methods (perhaps rename them) and return E_NOTIMPL?

OK, will address this and upload a new version soon.

Just to confirm:  This is only an issue for IE, right?  Just removing the
Firefox methods should be fine, it will generate an error only for those who try
to use them.

[amit, 2011-04-05 18:50:17]

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:14: uuid(25248626-A306-4857-A13C-4AD84FF15E10),
On 2011/04/05 18:47:10, Jói wrote:
> On 2011/04/05 18:00:24, amit wrote:
> > Tommi pointed out that any existing users embedding GCF will break with this
> > change.  Maybe we keep the methods (perhaps rename them) and return
E_NOTIMPL?
> 
> OK, will address this and upload a new version soon.
> 
> Just to confirm:  This is only an issue for IE, right?  Just removing the
> Firefox methods should be fine, it will generate an error only for those who
try
> to use them.
Yes, that's perfect. Also, please double check that the other artifacts like
typelib are still compatible.
Thanks!

[Jói, 2011-04-05 19:24:45]

PTAL, I've uploaded the new version that addresses review comments and run
chrome_frame_tests and chrome_frame_unittests locally.

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl
File chrome_frame/chrome_tab.idl (right):

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:14: uuid(25248626-A306-4857-A13C-4AD84FF15E10),
On 2011/04/05 18:50:17, amit wrote:
> On 2011/04/05 18:47:10, Jói wrote:
> > On 2011/04/05 18:00:24, amit wrote:
> > > Tommi pointed out that any existing users embedding GCF will break with
this
> > > change.  Maybe we keep the methods (perhaps rename them) and return
> E_NOTIMPL?
> > 
> > OK, will address this and upload a new version soon.
> > 
> > Just to confirm:  This is only an issue for IE, right?  Just removing the
> > Firefox methods should be fine, it will generate an error only for those who
> try
> > to use them.
> Yes, that's perfect. Also, please double check that the other artifacts like
> typelib are still compatible.
> Thanks!

I reverted the IID change here and on the dispinterface, and left the three
deprecated methods in (they return E_NOTIMPL in chrome_frame_activex_base.h).

http://codereview.chromium.org/6756044/diff/7002/chrome_frame/chrome_tab.idl#...
chrome_frame/chrome_tab.idl:118: [uuid(812C601D-D787-49f2-BF05-6A460B44D2B3)]
On 2011/04/05 17:04:17, Jói wrote:
> On 2011/04/05 16:57:16, Ruðrugis wrote:
> > I don't think this is necessary. As the interface here is a dispinterface.
> > There's no harm, that I can think of, however.
> 
> Yeah, I didn't think so either.  Just going by the book, so to speak :)

To be on the safe side (to leave the type library compatible with any clients
out there) I reverted the change of IID.

[Paweł Hajdan Jr., 2011-04-05 19:47:02]

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
File chrome/common/automation_messages_internal.h (right):

http://codereview.chromium.org/6756044/diff/7002/chrome/common/automation_mes...
chrome/common/automation_messages_internal.h:1136: // DEPRECATED MESSAGE - But
we must leave this comment and message so as
On 2011/04/05 18:47:10, Jói wrote:
> // TODO(phajdan.jr): Remove when the reference build is updated (this and
> // all others marked "DEPRECATED MESSAGE").

That's fine, thanks. LGTM.

[amit, 2011-04-06 02:18:36]

lgtm