-
Notifications
You must be signed in to change notification settings - Fork 605
EXPECTING_SPEECH + SPEAK directive simultaneously on multi-turn conversation. #44
Comments
Hello Juliano, It sounds as though the handler for the Speak directives is not being handled with a BLOCKING policy in DirectiveSequencer or that the ExpectSpeech directive is behaving as though it has a different dialogRequestId than the Speak directives. Can you confirm that your handler for Speak directives is registered as having a BLOCKING policy? Can you provide the dialogRequestId values for the Speak and ExpectSpeech directives in an instance where this problem occurred. There is a good change we will be able to see all that from the log of a DEBUG build. Can you post one to this thread? Thank you, |
HI, Thanks for quick response. The SPEAK which is saying that I have X items on my shopping list has same DialogID as Recognize event: But all SPEAK directives that are arriving after that, one for each ITEM on shopping list, does not have dialodID. Same thing happens for EXPECTED_SPEECH directive, it also does not have DialogID Thank you, |
Hi Juiano, |
Sure. Here is the x-amzn-requestid: x-amzn-requestid: 121315fffec65dfa-00006abf-0003fcbd-cfdea69fb1329a83-d98415b8-3 Thank you, |
Hi, Do you have any feedback on that? Thank you in advanced |
Hi Juliano, |
Hi Juliano, |
OK. Thank you for the feedback. |
Hello Juliano, A client side fix has been implemented in the Directive Sequencer library. It will be included in the next release. -SWH |
Thank you for keeping me posted. -Juliano |
Changes in this update - Added `AudioPlayer` capability agent. - Supports iHeartRadio. - `StateSynchronizer` has been updated to better enforce that `System.SynchronizeState`is the first Event sent on a connection to AVS. - Additional tests have been added to `ACL`. - The `Sample App` has been updated with several small fixes and improvements. - `ADSL` was updated such that all directives are now blocked while the handling of previous `SpeechSynthesizer.Speak` directives complete. Because any directive may now be blocked, the `preHandleDirective() / handleDirective()` path is now used for handling all directives. - Fixes for the following GitHub issues: - #44. - A bug causing `ACL` to not send a ping to AVS every 5 minutes, leading to periodic server disconnects, was fixed. - Subtle race condition issues were addressed in the `Executor` class, resolving some intermittent crashes.
Hello Juliano, The fix for this was included in the 1.0.0 release. Thank you, |
Hi,
My name is Juliano Ribeiro and I work for Qualcomm.
I notice that the current behavior from AVS SDK for EXPECTING_SPEECH Directive is to stop MediaPlayer in case any SPEAK directive is being playing out.
So, when I ask "Alexa, What is on my shopping list?" If shopping list is bigger then 5 items response from AVS will be 6 SPEAK directives (one for each item + 1 question asking if I would like to listen the rest of the items) + 1 EXPECTING_SPEECH ( to capture my response).
The problem with this use case is that EXPECTING_SPEECH directive is forcing MediaPlayer to STOP SPEAK directive and the question is interrupted as soon as it starts. In other words user can not listen to the question and does`t know what is happening.
*Note that, last SPEAK and EXPECTING_SPEECH are reaching device almost at same time.
Said that, here are my questions:
Can you please help me to review that?
Multi-turn conversation where we have just 1 SPEAK + 1 EXPECTING_SPEECH is working fine.
Thank you in advanced.
Juliano Ribeiro
The text was updated successfully, but these errors were encountered: