Hello all, 

my name is Pedro. Im having such a hard time using MRL xD, im trying to run the latest stable version which i couldn't even understant which one is it.

I have been facing severall dificulties and probably some other people are facing the same dificulties than me!

 

So first one, is it only needed a myrobotlab.jar file to install all the dependencies for the MRL?

Second, what freaking version to install? Gael talks 1.107, which im not beeing able to install in linux, the others i try (1.0.119 and "latest build"), every script i try run give different errors.

Third, how to use acapela service? in the latest build we have that option but it seems to give some errors for the snippet of create.blockspeech("Hello World") for exemple.

Fourth, the scripts Gael have uploaded (minimal) use mouth serviçe i think, instead of capela which would be the latest service for talking. Damn im so confused xD

Fifht, in order to make my arduino be recognized i use superuser when runing .jar, but then when i run webgui for talking we will go to chrome as default to listen, however since im superuser he will try open chrome as superuser which is not allowed

 

sorry guys i hope im not to much confusing with my problems...

 

 

 

juerg

8 years 5 months ago

Hi Pedro

You are right, you only need to download myrobotab.jar

move it into a folder e.g. home/mrl

you need to have java installed on your control computer (java -version should show you what you have)

cd into the mrl folder

type in: java -jar myrobotlab.jar

You might like to try this with the version 107 of myrobotlab.jar. This is a rather old version but it will reduce the chance of getting a latest with some flaws and it is known to be stable.

let us know how far you can get with this.

GroG

8 years 5 months ago

Hello and Welcome Pedro,

This issue that you reported (Acapela Speech not working) is because Wiki Data Fetcher had a conflicting library.

I fixed this..

Kwatters verified it ...

You could take the latest jar you were using, put it into a different directory, install, and try again..

 

Acapela is already working i confirm :)

 

Something is making me confused, when u start a serviçe, he says loud for exemple "starting mouth" where is it defined ? i mean maybe in a deeper header ???

Also in the early minimal scripts for exemple you say something and he replyies for exemple "did you ask open hand", i always got confuse because again i cant understand where is the code for what mp3 he fetch, and etc

myrobotlab.jar is a Java application - the code is here https://github.com/MyRobotLab/myrobotlab 

"starting mouth" - is defined here :
https://github.com/MyRobotLab/myrobotlab/blob/master/src/org/myrobotlab…

Sphinx was not a very good speech recognizer, so Gael wanted confirmation on actions - It would attempt to recognize speech, then use a SpeechToText system to generate a confirmation.

The Speech in the old system was primarily handled by GoogleSpeech, which no longer works because Google has broken that interface.  Now it can be handled by AcapelaSpeech or MaryTTS depending on the script details.

Most of the logic you are asking about is all in https://github.com/MyRobotLab/myrobotlab - the "scripts" are like glue which bind the logic together...

thanks grog, 

i just like to have full freedom :) and understand all the code allow that!!

Im sseing the code now but, is it in the directory of the MRL we create with the jar? or is it online and cant be changed! Maybe this is a stupid question sorry for that

Getting Started