[orca-list] sound on feisty

[Michael Whapples <mwhapples aim com>]

If what you put in the bash script is one line (or could be done on one
line), then you could put all that in the location="" string, and if
there are multiple commands you could do it with the && between
commands. Also if my technique was a patch then it assumes that the user
would want to use aoss (why else would they be applying the patch). It
just seems tidier not to have extra files floating around on the system
(and I remember from when I used slacware, one of the file systems you
needed to specify the block size, as larger block sizes is more
efficient with large files, but if there are many small files then you
may reach a maximum number of files but not have a full disk, whereas
smaller block sizes lowers the chance of the number of files being
reached but is less efficient for large files). Also how does that extra
bash call (the one to execute the script, as well as the one to do the
call from gnome speech affect performance? Probably isn't noticable on
modern machines, considering that these calls are infrequent (only when
speech is to be produced, which is likely to be less than one a second).

From
Michael Whapples
> Message: 1
> Date: Mon, 23 Apr 2007 17:22:44 +0200
> From: Halim Sahin <halim sahin t-online de>
> Subject: Re: [orca-list] sound on feisty
> To: orca-list gnome org
> Message-ID: <20070423152244 GA32123 halim local>
> Content-Type: text/plain; charset=us-ascii
>
> Hi,
> On Mo, Apr 23, 2007 at 02:24:34 +0100, Michael Whapples wrote:
> > Don't forget there is the other system by modifying the bonobo
> > activation file. To list the instructions for that precisely, here they
> > are.
> > 1. Open the GNOME_Speech bonobo activation file for the synth of your
> > choice (eg. for espeak on my ubuntu system I
> > open /usr/lib/bonobo/servers/GNOME_Speech_SynthesisDriver_Espeak.server). NOTE: you will need to open 
> > this with read/write access which probably means that you will need administrator/root rights.
> > 2. Modify line 3 which contains something like
> > type="exe" location="/usr/bin/espeak-synthesis-driver">
> > and add aoss to the beginning of the location string, so the above
> > becomes
> > type="exe" location="aoss /usr/bin/espeak-synthesis-driver">
> > Now save the file and logout and login again.
> >
> > How do these two systems compare? The above could even be put into a
> > patch file (if I knew how to create a patch file or someone else could
> > make the patch).
> A disadvantage of this method is that user need
> allways the alsa-oss package installed.
>
> Modifieing the synthesis-driver in this way I discribed falls back to 
> use oss if the libaoss.so isn't installed.
> Users with soundcards like sblive don't need the alsa-oss package 
> installed.
>
> HTH.
> Halim