Assigning to QA,


Advisory:
=============================

This is a very little bugfix update:

For users who would use svox-pico with speech-dispatcher will need a conf file named "pico-generic" but this file is missing in our current package.
So this update adds this missing "pico-generic" conf file into speech-dispatcher package.



Packages in 5/core/updates_testing:
========================
speech-dispatcher-0.8-14.1.mga5.i586.rpm
speech-dispatcher-0.8-14.1.mga5.x86_64.rpm
speech-dispatcher-conf-0.8-14.1.mga5.i586.rpm
speech-dispatcher-conf-0.8-14.1.mga5.x86_64.rpm
lib64speechd2-0.8-14.1.mga5.x86_64.rpm
lib64speechd-devel-0.8-14.1.mga5.x86_64.rpm
libspeechd2-0.8-14.1.mga5.i586.rpm
libspeechd-devel-0.8-14.1.mga5.i586.rpm
python3-speechd-0.8-14.1.mga5.i586.rpm
python3-speechd-0.8-14.1.mga5.x86_64.rpm

Source RPM: 
========================
speech-dispatcher-0.8-14.1.mga5.src.rpm


How to test this update request:
========================
- Install 'speech-dispatcher' from core/release and check if it contains a file named "pico-generic.conf" in /etc/speech-dispatcher/modules/ folder: 
Normally not:
$ ls /etc/speech-dispatcher/modules/ |grep 'pico'
$ (return nothing)

- Install 'speech-dispatcher' from core/updates_testing and check again if it contains now this file.
Normally yes:
$ ls /etc/speech-dispatcher/modules/ |grep 'pico'
pico-generic.conf
$

Assignee: geiger.david68210 => qa-bugs

This stuff is a bit of a mystery to me so bear with my lack of knowledge on this subject. 

The conf exists with the update installed, but the conf sets..

GenericExecuteSynth \
"pico2wave -w $TMPDIR/pico.wav -l $VOICE \'$DATA\' && $PLAY_COMMAND $TMPDIR/pico.wav

We don't appear to have any package which supplies pico2wave, so this is of little real benefit.

Whiteboard: (none) => feedback

Confirmed with madb rpmdiff that this is the only change, so apart from the above all is good.

Claire,
pico2wave comes from svox-pico
http://madb.mageia.org/package/show/application/0/name/svox-pico

@ Claire

The only one change in this update is the adding of module "pico-generic.conf" file.

So for test we need only make sure if this module is now present. that's all :)

Correct me if I'm wrong.  Adding this change through updates depends on a package in backports - That's a no go I'm afraid.

Claire,
No, speech-dispatcher (SD) works without this add-on, and svox-pico works also without SD. The configuration file added allows that SD uses voices from svox-pico when they are installed together. SD has already a list of configuration files for other synthesizer like festival, although not all synthesizer are installed.
I hope to be clear.

Thanks for clarifying. We do need to be careful with ties between updates and backports. 

To summarise then: This update enables the backport of svox-pico to be used if it is installed, but should cause no regression to the system when the backport is not installed. Note that svox-pico has not been assigned for testing at this stage.

We'll need to investigate how to properly test speech-dispatcher to verify it causes no regression, without having the backport for svox-pico installed. It's important we do properly check for regression as the config points to binaries which are not available without the backport being installed.

I seem to remember testing it before so we should have a record of it.

Whiteboard: feedback => (none)

Having a look at this on x86_64.

Checking the system before the update to see what is what.
No pico modules installed.  Using spd-say on the command line routes synthesized speech to the bluetooth speaker.  Examining pavucontrol shows an entry in the playback section for "speech-dispatcher".
$ spd-say "Good morning QA" 
That sounds perfect but do we need to do something more complex to test it, like using it in conjunction with espeak or the other applications with config files in the modules directory (cicero, flite, ivona, festival...)?  Not clear how to do that.

Just feeling my way.

CC: (none) => tarazed25

There appear to be six speech-dispatcher process running just now.  Some remarks on an Ubuntu forum suggest that these are left lying about after an application exits and that the user should 'killall speech-dispatcher'.
$ ps aux | grep -P 'speech-dispatcher.*\.conf'
lcl        446  0.0  0.0 220384  4752 pts/6    Sl   16:46   0:00 /usr/lib64/speech-dispatcher-modules/sd_cicero /etc/speech-dispatcher/modules/cicero.conf
lcl        451  0.0  0.0 236232  5552 pts/6    Sl   16:46   0:00 /usr/lib64/speech-dispatcher-modules/sd_flite /etc/speech-dispatcher/modules/flite.conf
lcl        454  0.0  0.0 220384  4824 pts/6    Sl   16:46   0:00 /usr/lib64/speech-dispatcher-modules/sd_dummy /etc/speech-dispatcher/modules/dummy.conf
lcl        457  0.0  0.0 220400  4776 pts/6    Sl   16:46   0:00 /usr/lib64/speech-dispatcher-modules/sd_generic /etc/speech-dispatcher/modules/generic.conf
lcl        460  0.0  0.0 388896  9056 pts/6    Sl   16:46   0:00 /usr/lib64/speech-dispatcher-modules/sd_espeak /etc/speech-dispatcher/modules/espeak.conf
lcl       4670  0.0  0.0  14492  2396 pts/6    S+   17:29   0:00 grep -P speech-dispatcher.*\.conf

It looks like speech-dispatcher looks at /usr/lib64/speech-dispatcher-modules and launches these processes automatically when it starts up so a killall does not sound like a good idea.

OK.  Since there is no pico-generic.conf file shall go ahead with the update and check for it again.

I've asked diogenese if he could look at this too Len. He's more familiar with the workings of these.

Mid-air collision.  Lost my spiel.

The 64-bit update did install pico-generic.conf and with no svox-pico on the system spd-say worked fine.
Used killall to remove the five speech-dispatcher processes and ran spd-say
again.  That relaunched all the processes from the lib64/speech-dispatcher-modules directory.
So far so good.

Validating

Keywords: (none) => validated_update
Whiteboard: advisory => advisory mga5-64-ok
CC: (none) => sysadmin-bugs

An update for this issue has been pushed to the Mageia Updates repository.

http://advisories.mageia.org/MGAA-2016-0058.html

Status: NEW => RESOLVED
Resolution: (none) => FIXED