Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True

Committed: https://crrev.com/bd3dda6c8666b29641b941ffd92be01afca6647e
Cr-Commit-Position: refs/heads/master@{#14150}

[hbos, 2016-09-06 10:50:13]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
==========

[hbos, 2016-09-06 12:29:28]

Patchset #2 (id:20001) has been deleted

[hbos, 2016-09-06 13:05:57]

hbos@webrtc.org changed reviewers:
+ magjed@webrtc.org, tkchin@webrtc.org

[hbos, 2016-09-06 13:05:58]

Please take a look magjed. tkchin OOO...

[hbos, 2016-09-07 06:45:35]

On 2016/09/06 13:05:58, hbos wrote:
> Please take a look magjed. tkchin OOO...

[magjed_webrtc, 2016-09-07 09:42:02]

lgtm

You still need to TBR tkchin.

[hbos, 2016-09-08 12:51:44]

On 2016/09/07 09:42:02, magjed_webrtc wrote:
> lgtm
> 
> You still need to TBR tkchin.

This turned out not to be as urgent as I thought, circular dependency, so I'm
holding off adding the inclusion that surfaced this for the time being, but
sooner or later it will be visible. I'll wait for tkchin to review.

[hbos, 2016-09-09 07:25:58]

On 2016/09/08 12:51:44, hbos wrote:
> On 2016/09/07 09:42:02, magjed_webrtc wrote:
> > lgtm
> > 
> > You still need to TBR tkchin.
> 
> This turned out not to be as urgent as I thought, circular dependency, so I'm
> holding off adding the inclusion that surfaced this for the time being, but
> sooner or later it will be visible. I'll wait for tkchin to review.

No, I'm hitting this again, I'll TBR-land so I can address the circular
dependency in a follow-up.

[hbos, 2016-09-09 07:26:34]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
==========

[hbos, 2016-09-09 07:26:57]

The CQ bit was checked by hbos@webrtc.org

[commit-bot: I haz the power, 2016-09-09 07:27:04]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[hbos, 2016-09-09 07:27:07]

The CQ bit was checked by hbos@webrtc.org

[commit-bot: I haz the power, 2016-09-09 07:27:14]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[commit-bot: I haz the power, 2016-09-09 07:37:44]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-09-09 07:37:44]

Try jobs failed on following builders:
  ios_arm64_rel on master.tryserver.webrtc (JOB_FAILED,
http://build.chromium.org/p/tryserver.webrtc/builders/ios_arm64_rel/builds/12676)

[hbos, 2016-09-09 08:05:49]

The CQ bit was checked by hbos@webrtc.org

[commit-bot: I haz the power, 2016-09-09 08:05:53]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[hbos, 2016-09-09 08:27:04]

The CQ bit was unchecked by hbos@webrtc.org

[hbos, 2016-09-09 08:27:32]

On 2016/09/09 08:05:53, commit-bot: I haz the power wrote:
> CQ is trying da patch. Follow status at
>  
>
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

android_dbg bot offline, everything else green, landing with NOTRY

[hbos, 2016-09-09 08:27:48]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True
==========

[hbos, 2016-09-09 08:28:03]

The CQ bit was checked by hbos@webrtc.org

[commit-bot: I haz the power, 2016-09-09 08:28:13]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.webrtc.org/...

[commit-bot: I haz the power, 2016-09-09 08:36:30]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True
==========

[commit-bot: I haz the power, 2016-09-09 08:36:31]

Committed patchset #2 (id:40001)

[commit-bot: I haz the power, 2016-09-09 08:36:40]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True

Committed: https://crrev.com/bd3dda6c8666b29641b941ffd92be01afca6647e
Cr-Commit-Position: refs/heads/master@{#14150}
==========

[commit-bot: I haz the power, 2016-09-09 08:36:41]

Patchset 2 (id:??) landed as
https://crrev.com/bd3dda6c8666b29641b941ffd92be01afca6647e
Cr-Commit-Position: refs/heads/master@{#14150}

[tkchin_webrtc, 2016-09-14 10:00:20]

What's the timeline for the new spec-compliant bits? It might have been good to
wait till the C++ bits settled before doing this change so the new bits could've
been added the same time the old bits were deprecated. Now we cannot release
ObjC cocoapod until the new API is in because we are guaranteeing an API break
for clients.

https://codereview.webrtc.org/2313943002/diff/40001/webrtc/sdk/objc/Framework...
File webrtc/sdk/objc/Framework/Classes/RTCPeerConnection+Stats.mm (right):

https://codereview.webrtc.org/2313943002/diff/40001/webrtc/sdk/objc/Framework...
webrtc/sdk/objc/Framework/Classes/RTCPeerConnection+Stats.mm:15: #import
"RTCLegacyStatsReport+Private.h"
alphabetize

https://codereview.webrtc.org/2313943002/diff/40001/webrtc/sdk/objc/Framework...
File webrtc/sdk/objc/Framework/Headers/WebRTC/RTCPeerConnection.h (right):

https://codereview.webrtc.org/2313943002/diff/40001/webrtc/sdk/objc/Framework...
webrtc/sdk/objc/Framework/Headers/WebRTC/RTCPeerConnection.h:26: @class
RTCLegacyStatsReport;
alphabetize

[hbos, 2016-09-14 12:41:02]

On 2016/09/14 10:00:20, tkchin_webrtc_OOO_till_Sept_12 wrote:
> What's the timeline for the new spec-compliant bits? It might have been good
to
> wait till the C++ bits settled before doing this change so the new bits
could've
> been added the same time the old bits were deprecated. Now we cannot release
> ObjC cocoapod until the new API is in because we are guaranteeing an API break
> for clients.

The new stats API (new stats classes, PeerConnectionInterface::GetStats) is
being added with CLs currently under review, but the new stats objects are going
to be added incrementally, starting from scratch - meaning until then you will
not get ahold of enough useful stats with the new API to be wanting to switch
from the old API. We will not want to switch from "legacy stats" to
"spec-compliant stats" in months, but in the meantime there are name conflicts.

Is renaming the current stats a problem for other projects?
Can naming conflicts be avoided by other means? Such as placing the new stats in
a different namespace?

[tkchin_webrtc, 2016-09-15 11:05:09]

On 2016/09/14 12:41:02, hbos wrote:
> On 2016/09/14 10:00:20, tkchin_webrtc_OOO_till_Sept_12 wrote:
> > What's the timeline for the new spec-compliant bits? It might have been good
> to
> > wait till the C++ bits settled before doing this change so the new bits
> could've
> > been added the same time the old bits were deprecated. Now we cannot release
> > ObjC cocoapod until the new API is in because we are guaranteeing an API
break
> > for clients.
> 
> The new stats API (new stats classes, PeerConnectionInterface::GetStats) is
> being added with CLs currently under review, but the new stats objects are
going
> to be added incrementally, starting from scratch - meaning until then you will
> not get ahold of enough useful stats with the new API to be wanting to switch
> from the old API. We will not want to switch from "legacy stats" to
> "spec-compliant stats" in months, but in the meantime there are name
conflicts.
> 
> Is renaming the current stats a problem for other projects?
> Can naming conflicts be avoided by other means? Such as placing the new stats
in
> a different namespace?

Gotcha. There shouldn't have been naming conflicts in ObjC land though? Since
only C++ ifaces where one is being deprecated and renamed, and a new one is
incoming?

Renaming the current stats required changes in at least one project that I know
of - cc-ed you there. It's not hard to rename, just inconvenient, and I thought
we wanted to transition official public headers more smoothly than causing build
breaks on updating WebRTC version. We can probably still release the pod with an
extra comment that new stats APIs are coming.

There are no namespaces in ObjC unfortunately so this is the best we can do
right now if we want to alert clients that the API is going away now. It's also
a special case of sorts since we are reimplementing a class and keeping the
name.

[hbos, 2016-09-15 20:20:43]

> On 2016/09/14 12:41:02, hbos wrote:
> > ...
> 
> Gotcha. There shouldn't have been naming conflicts in ObjC land though? Since
> only C++ ifaces where one is being deprecated and renamed, and a new one is
> incoming?
> 
> Renaming the current stats required changes in at least one project that I
know
> of - cc-ed you there. It's not hard to rename, just inconvenient, and I
thought
> we wanted to transition official public headers more smoothly than causing
build
> breaks on updating WebRTC version. We can probably still release the pod with
an
> extra comment that new stats APIs are coming.
> 
> There are no namespaces in ObjC unfortunately so this is the best we can do
> right now if we want to alert clients that the API is going away now. It's
also
> a special case of sorts since we are reimplementing a class and keeping the
> name.

The naming conflict was introduced when I included new stats-classes in
peerconnectioninterface.h and there are ObjC files that use both that and
RTCStatsReports.
Now the CL looks like this:
https://codereview.webrtc.org/2331373004/diff/160001/webrtc/api/peerconnectio...
In hindsight I wound up only needing a pointer to a callback class, which could
be declared without including stats stuff. We may very well be able to avoid the
conflict like that.

Sooner or later this name conflict will emerge, but it might be possible to
revert this CL and wait until we are ready to deprecate the old stats API. Or we
add support for both stats APIs from the start, as the intention is to do in C++
and Chromium. Thoughts? (I'm OOO Friday)

[hbos, 2016-09-15 20:22:17]

Ah, cut already done :)

[tkchin_webrtc, 2016-09-20 17:38:19]

Description was changed from

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True

Committed: https://crrev.com/bd3dda6c8666b29641b941ffd92be01afca6647e
Cr-Commit-Position: refs/heads/master@{#14150}
==========

to

==========
Renamed RTCStatsReport to RTCLegacyStatsReport in objc files.

This is to avoid a naming conflict with webrtc::RTCStatsReport that is
surfaced if you try to include it in peerconnectioninterface.h.

Background: The current stats is very much non-spec-compliant. A new
stats collection API is underway that is meant to be spec-compliant.
Some classes in Chromium and webrtc/sdk/objc have spec-compliant names
but non-spec-compliant behavior. These are being renamed to "Legacy" so
that new spec-compliant classes can be added with the correct names.

BUG=chromium:627816
TBR=tkchin@webrtc.org
NOTRY=True

Committed: https://crrev.com/bd3dda6c8666b29641b941ffd92be01afca6647e
Cr-Commit-Position: refs/heads/master@{#14150}
==========

[tkchin_webrtc, 2016-09-20 17:38:19]

tkchin@webrtc.org changed reviewers:
+ pthatcher@webrtc.org

[tkchin_webrtc, 2016-09-20 17:39:12]

On 2016/09/15 20:20:43, hbos wrote:
> > On 2016/09/14 12:41:02, hbos wrote:
> > > ...
> > 
> > Gotcha. There shouldn't have been naming conflicts in ObjC land though?
Since
> > only C++ ifaces where one is being deprecated and renamed, and a new one is
> > incoming?
> > 
> > Renaming the current stats required changes in at least one project that I
> know
> > of - cc-ed you there. It's not hard to rename, just inconvenient, and I
> thought
> > we wanted to transition official public headers more smoothly than causing
> build
> > breaks on updating WebRTC version. We can probably still release the pod
with
> an
> > extra comment that new stats APIs are coming.
> > 
> > There are no namespaces in ObjC unfortunately so this is the best we can do
> > right now if we want to alert clients that the API is going away now. It's
> also
> > a special case of sorts since we are reimplementing a class and keeping the
> > name.
> 
> The naming conflict was introduced when I included new stats-classes in
> peerconnectioninterface.h and there are ObjC files that use both that and
> RTCStatsReports.
> Now the CL looks like this:
>
https://codereview.webrtc.org/2331373004/diff/160001/webrtc/api/peerconnectio...
> In hindsight I wound up only needing a pointer to a callback class, which
could
> be declared without including stats stuff. We may very well be able to avoid
the
> conflict like that.
> 
> Sooner or later this name conflict will emerge, but it might be possible to
> revert this CL and wait until we are ready to deprecate the old stats API. Or
we
> add support for both stats APIs from the start, as the intention is to do in
C++
> and Chromium. Thoughts? (I'm OOO Friday)

I defer to pthatcher here. Likely iOS APIs will follow same migration plan as
C++ APIs. Peter, wdyt?