[Grml] Speech-Dispatcher and SMP (again)

Keith Hinton keithint1234 at gmail.com
Mon Mar 16 03:47:56 CET 2009


Hi,
I have previously in the last year or so posted on this subject of
Speech-Dispatcher and SMP.
I had previously back when testing my friend Shaun's system thought
that it was only to AMD Athlon X2 3800 plus dual-core processors.
I found out, that I was wrong.
About four weeks ago, I tested the latest stable GRML using the grml
swspeak speechd boot option at the Isolinux boot prompt on my Intel
laptop. My Fujitsu laptop is pretty new. Well, shortly there after,
speech crashed, again.
So, I did this:  Speechd-up -v
Grml uses Speechd-up 0.1, wich is extremely old and out-of-date. Even
the  last stable release contains severe bugs. The CVS builds contain
latest bug fixes.
In edition to this, I discovered that GRML was using an RC version
(not stable!) version of Speech-Dispatcher!
Why is this?
The issues seem to be caused when using Espeak-generic, and speechd-up
0.1, and Speech-Dispatcher. whether it is Speech-Dispatcher itself
though, or simply Speechd-up, I don't know. In any case, I strongly
suggest the espeakup interface over the Speechd-up interface (to the
point of you folks ripping it completely out)
It is odd though, because I thought  that at least Speech-Dispatcher
would contain some improvements? They were fixing SMP bugs last time I
checked. Although if it is an RC version, I am not surprised. Although
it may very well be an old version of Speechd-up itself. I will
perform another test, on March 20, when I have some time. This time,
I'll attempt to execute an instance of the Speechd-up software by
typing speechd-up again, to see if it is the Dispatcher process
crashing. I will also find out if there are any speech-dispatcher log
levels generated through output of log contents (if any).
Because Speech-Dispatcher is an interface wich uses the TCP/IP
protocol, and is a client/server model, if it was Speechd-up crashing,
and not Speech-Dispatcher, then it will probably generate some  form
of client/module disconnection in the log files.
I will report back hear with more test results.
Regards, --Keith


More information about the Grml mailing list