Mitch Capper wrote:
You may want to look at the Intel Atom combo machines you can get a 1.6
ghz machine probably for around $100-150 USD in a very small form factor
and very powerful.
Thanks for the tip. Any link where I could check this out?
The cheapest PC's I find are 230€ (Asus'
The cheapest PC's I find are 230€ (Asus' EeePC), and with not enough room
to stick a PCI card.
Well actually a barebone with VIA C7 + 1 slot for PCI card (carefull you
have to remove some metal part otherwise Sangoma/Digium card will not
fit *LOL*)
and external PSU are around 190/215€
Thanks Antonio for the links on Acrosser and PCEngines. It seems like PCE's
alix1d is a good solution, provided 256MB is enough to hold Linux +
Freeswitch + some tiny LAMP stack. Still, it looks like an Atom-included
mobo like those from Asus or Gigabyte would be cheaper. The biggest issue is
Fred-145 ha scritto:
Thanks Antonio for the links on Acrosser and PCEngines. It seems like PCE's
alix1d is a good solution, provided 256MB is enough to hold Linux +
Freeswitch + some tiny LAMP stack. Still, it looks like an Atom-included
mobo like those from Asus or Gigabyte would be cheaper.
On May 4, 2009, at 9:25 PM, Anthony Minessale wrote:
When you are using on-hook agents, it's presumed to be under low
call volume, you can just set the agents to get popped
into the queue in nowait mode so if the caller changed his mind the
agent will get a hangup. Remember, if there are
Hello,
Anthony, I would like to provide a patch allowing having different call
distribution strategies, at least for call back agents.
Do you think the simple approach of modifying the SQL query in
find_consumers (given strategy that would be set from dialplan) would be
enough?
Thanks,
On Sun, May 3, 2009 at 11:01 PM, seven dujinf...@gmail.com wrote:
Actually, for the call back agents, because the fifo use originate to
start a new session, the new session won't hang up unless one agent answered
or timeout. Agents will hear nothing and wait(member_wait=wait) on the
queue or
I am having a problem with getting multiple Polycom IP phones to register
to my Freeswitch server. Here is my setup (IP addresses are not actual
ones, but are consistent throughout):
Freeswitch server in colo facility
IP addr: 1.1.1.1 (publicly routable)
Linux NAT firewall
BTW Brian,
Here is something that would make FS's VAD much better. The technique also
improved Sphinx-3 performance in low-SNR enviroments and made it run over 40%
faster.
http://figment.cse.usf.edu/~sfefilat/data/papers/WeBT5.3.pdf
Mark.
-Original Message-
From: Brian West
Wasn't aware Sphinx 3 was integrated into FreeSWITCH ...
/b
On May 4, 2009, at 12:00 PM, mszla...@aol.com wrote:
BTW Brian,
Here is something that would make FS's VAD much better. The
technique also improved Sphinx-3 performance in low-SNR enviroments
and made it run over 40% faster.
Nope it isn't but does that make a difference if pocketsphinx could use a
similar upgrade?
Anyway, you now have a way to make VAD better in FS.
-Original Message-
From: Brian West br...@freeswitch.org
To: freeswitch-users@lists.freeswitch.org
Sent: Mon, 4 May 2009 10:29 am
Hi,
maybe things in speex should be worth to look for too.
Just my 2 cents...
Regards,
Tamas
ps: It seems, VAD+DTX in mod_speex does not work.
Brian West írta:
VAD isn't really high on my list right now.
/b
On May 4, 2009, at 12:46 PM, mszla...@aol.com
mailto:mszla...@aol.com
VAD isn't really high on my list right now.
/b
On May 4, 2009, at 12:46 PM, mszla...@aol.com wrote:
Nope it isn't but does that make a difference if pocketsphinx could
use a similar upgrade?
Anyway, you now have a way to make VAD better in FS.
Brian West
br...@freeswitch.org
-- Meet us
13 matches
Mail list logo
