Re: [orca-list] recommended text to speech engine other the ne-speak



Thanks Luke, that makes things clearer for me, and hopefully I won't
make that mistake again.

Michael Whapples
On Tue, 2008-06-24 at 09:40 +1000, Luke Yelavich wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, Jun 23, 2008 at 10:31:00PM EST, Michael Whapples wrote:
When you are running i386 software on you x86_64 machine I believe that
it is through some sort of compatibility mode? Does this mean that you
need to be running an i386 distribution or can the i386 support work
from inside an x86_64 distribution (IE. can you have a mix of x86_64 and
i386 software running on the one system)? Does running i386 software on
your x86_64 machine mean that you loose some of the advantages of your
x86_64?

An X86_64 compatible CPU can run either an X86_64 operating system or an i386 operating system. In terms of 
Linux, you can have things set up a few ways. First, you can have a completely X86_64 operating system, i.e 
all programs are built for X86_64. Secondly, you can have an X86_64 kernel, with an i386 userspace. This 
means that the kernel will allow the accessing of 4GB or more of RAM, etc. I386 userspace means that 
everything from the runtime libraries, to all your every day applications are built for use with i386. 
Since you may have 4GB or more of RAM, the system can fully utilize this RAM, however applications can 
still only use 2GB of RAM at a maximum. It is also possible to have a complete X86_64 system, with 
libraries installed for i386 applications to be used, most distros do things this way.

To use ttsynth/voxin on an X86_64 system, you are going to need i386 versions of the libraries that 
ttsynth/voxin depends on.  The names of these packages varies from distro to distro, but on Ubuntu/Debian, 
the ia32-libs package should install what you need. The only other issue is installing the gcc 2.95/glibc 
2.2 library that ttsynth/voxin needs. Since for Debian/ubuntu at least, this is an i386 deb, it cannot be 
installed onto an amd64/X86_64 Ubuntu/Debian install. You are going to have to unpack the deb manually, 
copy the library to where it should go, and create symbolic links etc as needed.

Once ttsynth is set up, you will then need to work out how it will be linked and will communicate with the 
various speech backends.  This is an exercise for the reader, as while I've given it thought, I don't have 
a solution in mind, however given demand, I may consider coming up with something relatively easy for 
Ubuntu users.

Hope this helps clear up a few things.

Luke
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIYDRyjVefwtBjIM4RApi8AKC/1y1dAAhtaSRzM2AVgiwk/eMDGgCcD9cI
L77Wfbik8IDtL8V3Asq6pPw=
=S2OA
-----END PGP SIGNATURE-----





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]