1239242 - [Presentation WebAPI] support start session with designated device

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

For browser UI or system-level Gaia app, they usually have following UX flow for launching presentation request:

  1. browser UI displays available devices
  2. user picks a device
  3. user picks an action to perform

Since browser UI already have the privilege to know device list via XPCOM [1] or PresentationDeviceInfo API [2], we should provide an `PresentationRequest.startWithDevice(deviceId)` for skipping the device selection prompt.

[1] https://dxr.mozilla.org/mozilla-central/source/dom/presentation/interfaces/nsIPresentationDeviceManager.idl
[2] https://wiki.mozilla.org/WebAPI/PresentationDeviceInfoAPI

Comment 1

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

Attachment: start-with-device-api.patch

For any code that can access PresentationDeviceInfo API should be able to use PresentationRequest.startWithDevice().

Comment 2

[Olli Pettay [:smaug][[e-mail]]]

Comment on attachment 8707394 [details] [diff] [review]
start-with-device-api.patch

>+  while(NS_SUCCEEDED(enumerator->HasMoreElements(&hasMore))
>+        && hasMore){
>+    nsCOMPtr<nsISupports> isupports;
>+    rv = enumerator->GetNext(getter_AddRefs(isupports));
>+
>+    nsCOMPtr<nsIPresentationDevice> device(do_QueryInterface(isupports));
>+    MOZ_ASSERT(device);
>+
>+    nsAutoCString id;
>+    if (NS_SUCCEEDED(device->GetId(id))
>+        && id.Equals(utf8DeviceId)) {
Nit, I'd put && at the end of previous line

>@@ -38,24 +38,27 @@ interface nsIPresentationService : nsISu
update uuid of this interface


>+function testStartConnectionWithDevice() {
>+  return new Promise(function(aResolve, aReject) {
>+    gScript.addMessageListener('device-prompt', function devicePromptHandler() {
>+      gScript.removeMessageListener('device-prompt', devicePromptHandler);
>+      ok(false, "Device prompt should not be triggered.");
Should we still remove the listener even if this doesn't get called?
Or do we somehow manage to not leak here.

>+   * A chrome page uses startWithDevice()
chrome page, or page which has presentation-device-manage permissiongs

Comment 3

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

(In reply to Olli Pettay [:smaug] from comment #2)
> Comment on attachment 8707394 [details] [diff] [review]
> start-with-device-api.patch
> 
> >+  while(NS_SUCCEEDED(enumerator->HasMoreElements(&hasMore))
> >+        && hasMore){
> >+    nsCOMPtr<nsISupports> isupports;
> >+    rv = enumerator->GetNext(getter_AddRefs(isupports));
> >+
> >+    nsCOMPtr<nsIPresentationDevice> device(do_QueryInterface(isupports));
> >+    MOZ_ASSERT(device);
> >+
> >+    nsAutoCString id;
> >+    if (NS_SUCCEEDED(device->GetId(id))
> >+        && id.Equals(utf8DeviceId)) {
> Nit, I'd put && at the end of previous line
I realized that there is enough space if we put it in one line.
fixed by merging it into one line.
> 
> >@@ -38,24 +38,27 @@ interface nsIPresentationService : nsISu
> update uuid of this interface
fixed.
> 
> 
> >+function testStartConnectionWithDevice() {
> >+  return new Promise(function(aResolve, aReject) {
> >+    gScript.addMessageListener('device-prompt', function devicePromptHandler() {
> >+      gScript.removeMessageListener('device-prompt', devicePromptHandler);
> >+      ok(false, "Device prompt should not be triggered.");
> Should we still remove the listener even if this doesn't get called?
> Or do we somehow manage to not leak here.
gScript.destroy() at teardown procedure will remove the listener.
> 
> >+   * A chrome page uses startWithDevice()
> chrome page, or page which has presentation-device-manage permissiongs
fixed.

Comment 4

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

Attachment: start-with-device-api.patch

rebase to m-c and update according to review comments, carry r+.

Comment 5

[Shih-Chiang Chien [:schien] (UTC+8) (use ni? plz)]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=fe0280331c94

Comment 6

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/802b5a1ad701

Comment 7

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/802b5a1ad701