Re: [asterisk-users] LCR in Asterisk

2008-03-17 Thread Ex Vito
On Wed, Feb 13, 2008 at 6:49 PM, Tilghman Lesher
[EMAIL PROTECTED] wrote:

  Macros are deprecated.  Gosubs are the way forward, and yes, they have
  local variables.  Simply define them once as Set(LOCAL(foo)=bar) and foo
  will be gone when the innermost stack is removed (either by Return or
  StackPop).


  I was keeping up with the list traffic when I found this... Hmmm, maybe to
  my surpirse.

  Are macros deprecated in 1.4 or 1.6 ? What, if any, is the replacement for
  the M() option in the Dial() application ?
--
 exvito

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-03-17 Thread Tilghman Lesher
On Monday 17 March 2008 21:55:10 Ex Vito wrote:
 On Wed, Feb 13, 2008 at 6:49 PM, Tilghman Lesher

 [EMAIL PROTECTED] wrote:
   Macros are deprecated.  Gosubs are the way forward, and yes, they have
   local variables.  Simply define them once as Set(LOCAL(foo)=bar) and foo
   will be gone when the innermost stack is removed (either by Return or
   StackPop).

   I was keeping up with the list traffic when I found this... Hmmm, maybe
 to my surpirse.

   Are macros deprecated in 1.4 or 1.6 ? What, if any, is the replacement
 for the M() option in the Dial() application ?

1.6.  U().

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-17 Thread Thomas Kenyon
Tilghman Lesher wrote:
 
 Macros are deprecated.  Gosubs are the way forward, and yes, they have
 local variables.  Simply define them once as Set(LOCAL(foo)=bar) and foo
 will be gone when the innermost stack is removed (either by Return or
 StackPop).
 
Yeah, but I like macros.
stamps feet.

Still nice to know that the , | () nonesense is being cleaned up.

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-17 Thread Tilghman Lesher
On Sunday 17 February 2008 03:41:09 Thomas Kenyon wrote:
 Tilghman Lesher wrote:
  Macros are deprecated.  Gosubs are the way forward, and yes, they have
  local variables.  Simply define them once as Set(LOCAL(foo)=bar) and foo
  will be gone when the innermost stack is removed (either by Return or
  StackPop).

 Yeah, but I like macros.
 stamps feet.

Not that I would ever encourage such dialplan abuse, but Gosub will permit
going far deeper, even in recursion than Macro.  With Macro, you are limited
from anywhere from 5 to 10 levels deep, because it exhausts the call stack,
whereas Gosub will permit something like 100,000 nested calls deep.

 Still nice to know that the , | () nonesense is being cleaned up.

Thank you.  I realize this is going to be painful initially, but it's quite
necessary to build even more complexity into the dialplan in the future.

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-17 Thread John
On Feb 13, 2008 12:33 PM, Tilghman Lesher
[EMAIL PROTECTED] wrote:

 In the same way that a PHP programmer should not attempt write Python the
 way she writes PHP, I would agree with you.  However, if you're willing to
 adapt to the ways the dialplan works, you can create dialplans which aren't
 obfuscated at all.  Tcl and Lisp are close cousins to the dialplan in terms of
 how they do things.  Not everybody is a Lisp programmer, and some people
 absolutely detest it.  That doesn't make it any less of a good language.

 --
 Tilghman


Having done significant work in Tcl,  and very interested in this LCR
debate, can you shed light on the similarities between Tcl and the
dial plan?   Thx!

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-17 Thread Tilghman Lesher
On Sunday 17 February 2008 10:33:18 John wrote:
 On Feb 13, 2008 12:33 PM, Tilghman Lesher

 [EMAIL PROTECTED] wrote:
  In the same way that a PHP programmer should not attempt write Python the
  way she writes PHP, I would agree with you.  However, if you're willing
  to adapt to the ways the dialplan works, you can create dialplans which
  aren't obfuscated at all.  Tcl and Lisp are close cousins to the dialplan
  in terms of how they do things.  Not everybody is a Lisp programmer, and
  some people absolutely detest it.  That doesn't make it any less of a
  good language.

 Having done significant work in Tcl,  and very interested in this LCR
 debate, can you shed light on the similarities between Tcl and the
 dial plan?   Thx!

I'd say the most striking resemblance between the two involves the use
of components, which Osterhout describes in his 1994 book on the Tcl/Tk
environment.  Rather than creating a singular monolithic language which
is able to do everything, the dialplan relies in great part on the ability to
add new components, whether they be dialplan applications or dialplan
functions, to supplement various tasks.  These components act in a
cooperative fashion for whatever the pbx administrator needs to do.

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Tilghman Lesher
On Tuesday 12 February 2008 23:14:58 Alex Balashov wrote:
 Rizwan Hisham wrote:
  Hi all,
  I am planning to implement LCR routing on my already running asterisk
  server. Uptill now i have found out that asterisk has no support for
  lcr, i have to do something about it myself, for example using the AGI.
  Im looking for ideas here. Whats the best way to start implementing lcr
  in asterisk. Should i use agi and start implementing my own lcr script
  or is there any plugin available which can be used with asterisk.

 If you are interested in prebuilt solutions, you may consider
 TransNexus's NexOSS product (www.transnexus.com).  The Open Settlement
 Protocol (OSP) they implemented can be used with Asterisk - they have a
 module.  In fact, I am not sure about the commercial status of the OSP
 module as such;  it may be possible to get it free of charge.  Not sure.
   But OSP is an open protocol, so perhaps it's possible.

 Otherwise, I would think that the best way to approach this would be to
 make it fully outboard and divest it of Asterisk.  Implement a SIP proxy
 that forwards to providers using LCR decisionmaking, and just have
 Asterisk send calls to it.  OpenSER can be used for this - and indeed,
 there is an OSP module for it as well, if you wanted to go that route.

 If you're dead-set on doing it in Asterisk and don't want to do OSP, I
 would suggest FastAGI.  Definitely don't implement the logic in the dial
 plan, at any cost.

Uh, why not?  You can do LCR quite easily in the dialplan, by using func_odbc
for each of the provider lookups, then use SORT() to get the lowest cost.
It's quite easy, and you do not need to resort to AGI.

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Alex Balashov
Tilghman Lesher wrote:

 Uh, why not?  You can do LCR quite easily in the dialplan, by using func_odbc
 for each of the provider lookups, then use SORT() to get the lowest cost.
 It's quite easy, and you do not need to resort to AGI.

You can do almost anything in the dial plan with enough spiritual 
commitment in about the same way that you can do just about anything you 
need to do with a bash script, as opposed to Perl, Python, or any 
toolkits or frameworks.

It's not syntactically terse, balloons quickly in semantic complexity, 
is objectively less efficient as the dial plan *is not a programming 
language* (despite having variables, control structures and other things 
characteristic of an execution environment of such), and otherwise 
unnecessarily complicated.  In implementing and extending the logic 
going forward (beyond naive lookups) in accordance with evolving 
requirements in the business rules, you will find that you run into the 
limits of the algorithmic complexity that the dial plan can provide, and 
that whatever the approach, it's overly obfuscated.  The dial plan 
limits meaningful modularisation and functional decomposition that is 
available with outboard runtime environments.

So, it's not that you couldn't - it's that you shouldn't.

-- 
Alex Balashov
Evariste Systems
Web: http://www.evaristesys.com/
Tel: (+1) (678) 954-0670
Direct : (+1) (678) 954-0671
Mobile : (+1) (706) 338-8599

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Jay R. Ashworth
On Wed, Feb 13, 2008 at 11:33:19AM -0600, Tilghman Lesher wrote:
 On Wednesday 13 February 2008 09:57:59 Alex Balashov wrote:
  Tilghman Lesher wrote:
   Uh, why not?  You can do LCR quite easily in the dialplan, by using
   func_odbc for each of the provider lookups, then use SORT() to get the
   lowest cost. It's quite easy, and you do not need to resort to AGI.
 
  You can do almost anything in the dial plan with enough spiritual
  commitment in about the same way that you can do just about anything you
  need to do with a bash script, as opposed to Perl, Python, or any
  toolkits or frameworks.
 
  It's not syntactically terse, balloons quickly in semantic complexity,
  is objectively less efficient as the dial plan *is not a programming
  language* (despite having variables, control structures and other things
  characteristic of an execution environment of such), and otherwise
  unnecessarily complicated.  In implementing and extending the logic
  going forward (beyond naive lookups) in accordance with evolving
  requirements in the business rules, you will find that you run into the
  limits of the algorithmic complexity that the dial plan can provide, and
  that whatever the approach, it's overly obfuscated.  The dial plan
  limits meaningful modularisation and functional decomposition that is
  available with outboard runtime environments.
 
 Like any other language, you certainly can write in an obfuscated way, and
 the dialplan does not discourage it.  That said, you certainly can write in a
 modularized way.  I would guess that you simply aren't familiar with the
 dialplan enough to make those decisions, but it is quite possible and doable.
 
  So, it's not that you couldn't - it's that you shouldn't.
 
 In the same way that a PHP programmer should not attempt write Python the
 way she writes PHP, I would agree with you.  However, if you're willing to
 adapt to the ways the dialplan works, you can create dialplans which aren't
 obfuscated at all.  Tcl and Lisp are close cousins to the dialplan in terms of
 how they do things.  Not everybody is a Lisp programmer, and some people
 absolutely detest it.  That doesn't make it any less of a good language.

Having programmed in about 8 different languages over the last 25
years, I can see both points of view.  And admittedly, I haven't tried
to do non-trivial things with dialplan.

That said, my view of this interaction is that Tilghman has drunk the
Kool-Aidtm, and that Alex's view of the situation is much closer to
objective.

dialplan appears to have jes' growed, and that never makes for a good
language design.  Ask the Python 3 team.  :-)

Cheers,
-- jra
-- 
Jay R. Ashworth   Baylink  [EMAIL PROTECTED]
Designer The Things I Think   RFC 2100
Ashworth  Associates http://baylink.pitas.com '87 e24
St Petersburg FL USA  http://photo.imageinc.us +1 727 647 1274

 Those who cast the vote decide nothing.
 Those who count the vote decide everything.
   -- (Joseph Stalin)


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Tilghman Lesher
On Wednesday 13 February 2008 09:57:59 Alex Balashov wrote:
 Tilghman Lesher wrote:
  Uh, why not?  You can do LCR quite easily in the dialplan, by using
  func_odbc for each of the provider lookups, then use SORT() to get the
  lowest cost. It's quite easy, and you do not need to resort to AGI.

 You can do almost anything in the dial plan with enough spiritual
 commitment in about the same way that you can do just about anything you
 need to do with a bash script, as opposed to Perl, Python, or any
 toolkits or frameworks.

 It's not syntactically terse, balloons quickly in semantic complexity,
 is objectively less efficient as the dial plan *is not a programming
 language* (despite having variables, control structures and other things
 characteristic of an execution environment of such), and otherwise
 unnecessarily complicated.  In implementing and extending the logic
 going forward (beyond naive lookups) in accordance with evolving
 requirements in the business rules, you will find that you run into the
 limits of the algorithmic complexity that the dial plan can provide, and
 that whatever the approach, it's overly obfuscated.  The dial plan
 limits meaningful modularisation and functional decomposition that is
 available with outboard runtime environments.

Like any other language, you certainly can write in an obfuscated way, and
the dialplan does not discourage it.  That said, you certainly can write in a
modularized way.  I would guess that you simply aren't familiar with the
dialplan enough to make those decisions, but it is quite possible and doable.

 So, it's not that you couldn't - it's that you shouldn't.

In the same way that a PHP programmer should not attempt write Python the
way she writes PHP, I would agree with you.  However, if you're willing to
adapt to the ways the dialplan works, you can create dialplans which aren't
obfuscated at all.  Tcl and Lisp are close cousins to the dialplan in terms of
how they do things.  Not everybody is a Lisp programmer, and some people
absolutely detest it.  That doesn't make it any less of a good language.

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Douglas Garstang
- Original Message 
From: Jay R. Ashworth [EMAIL PROTECTED]
To: asterisk-users@lists.digium.com
Sent: Wednesday, February 13, 2008 9:45:34 AM
Subject: Re: [asterisk-users] LCR in Asterisk

On 
Wed, 
Feb 
13, 
2008 
at 
11:33:19AM 
-0600, 
Tilghman 
Lesher 
wrote:
 
On 
Wednesday 
13 
February 
2008 
09:57:59 
Alex 
Balashov 
wrote:
 
 
Tilghman 
Lesher 
wrote:
 
 
 
Uh, 
why 
not?  
You 
can 
do 
LCR 
quite 
easily 
in 
the 
dialplan, 
by 
using
 
 
 
func_odbc 
for 
each 
of 
the 
provider 
lookups, 
then 
use 
SORT() 
to 
get 
the
 
 
 
lowest 
cost. 
It's 
quite 
easy, 
and 
you 
do 
not 
need 
to 
resort 
to 
AGI.
 

 
 
You 
can 
do 
almost 
anything 
in 
the 
dial 
plan 
with 
enough 
spiritual
 
 
commitment 
in 
about 
the 
same 
way 
that 
you 
can 
do 
just 
about 
anything 
you
 
 
need 
to 
do 
with 
a 
bash 
script, 
as 
opposed 
to 
Perl, 
Python, 
or 
any
 
 
toolkits 
or 
frameworks.

Is that nasty little problem of no local variables in macros fixed yet? That's 
a pretty big pain in the ass. You have to prefix your variables with the name 
of the macro it's in to avoid stepping all over yourself.

Doug.






  

Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Philipp Kempgen
Douglas Garstang wrote:
 - Original Message 
 From: Jay R. Ashworth [EMAIL PROTECTED]
 To: asterisk-users@lists.digium.com
 Sent: Wednesday, February 13, 2008 9:45:34 AM
 Subject: Re: [asterisk-users] LCR in Asterisk
 
 On 
 Wed, 
 Feb 
 13, 
 2008 
 at 
 11:33:19AM 
 -0600, 
 Tilghman 
 Lesher 
 wrote:
 On 
 Wednesday 
 13 
 February 
 2008 
 09:57:59 
 Alex 
 Balashov 
 wrote:

 Tilghman 
 Lesher 
 wrote:


 Uh, 
 why 
 not?  
 You 
 can 
 do 
 LCR 
 quite 
 easily 
 in 
 the 
 dialplan, 
 by 
 using


 func_odbc 
 for 
 each 
 of 
 the 
 provider 
 lookups, 
 then 
 use 
 SORT() 
 to 
 get 
 the


 lowest 
 cost. 
 It's 
 quite 
 easy, 
 and 
 you 
 do 
 not 
 need 
 to 
 resort 
 to 
 AGI.



 You 
 can 
 do 
 almost 
 anything 
 in 
 the 
 dial 
 plan 
 with 
 enough 
 spiritual

 commitment 
 in 
 about 
 the 
 same 
 way 
 that 
 you 
 can 
 do 
 just 
 about 
 anything 
 you

 need 
 to 
 do 
 with 
 a 
 bash 
 script, 
 as 
 opposed 
 to 
 Perl, 
 Python, 
 or 
 any

 toolkits 
 or 
 frameworks.

Could you fix your e-mail client please?

Regards,
  Philipp Kempgen

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Tilghman Lesher
On Wednesday 13 February 2008 11:45:34 Jay R. Ashworth wrote:
 Having programmed in about 8 different languages over the last 25
 years, I can see both points of view.  And admittedly, I haven't tried
 to do non-trivial things with dialplan.

 That said, my view of this interaction is that Tilghman has drunk the
 Kool-Aidtm, and that Alex's view of the situation is much closer to
 objective.

Or maybe I'm just the architect of the dialplan moving forward, which is why
I advocate that if you really don't need to use AGI, you don't.  ;-)

 dialplan appears to have jes' growed, and that never makes for a good
 language design.  Ask the Python 3 team.  :-)

I'm specifically working on removing misfeatures from the dialplan, to make
it much easier to use and more predictable.  1.6 will be a huge improvement
towards this goal.

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Ira
At 09:33 AM 2/13/2008, you wrote:
In the same way that a PHP programmer should not attempt write Python the
way she writes PHP, I would agree with you.  However, if you're willing to
adapt to the ways the dialplan works, you can create dialplans which aren't
obfuscated at all.  Tcl and Lisp are close cousins to the dialplan in terms of
how they do things.  Not everybody is a Lisp programmer, and some people
absolutely detest it.  That doesn't make it any less of a good language.

Look, I've done lots of cool stuff in the dial plan and other have 
done stuff way beyond me, but I defy you to call the dial plan 
language good or well designed. It works, it gets the job done but 
it's always harder than it needs to be.

Ira 


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Tilghman Lesher
Doug-

Please fix your email client.  One line per word in quoting is a little
excessive.  Better yet, turn off HTML.

On Wednesday 13 February 2008 12:17:30 Douglas Garstang wrote:
 Is that nasty little problem of no local variables in macros fixed yet?
 That's a pretty big pain in the ass. You have to prefix your variables with
 the name of the macro it's in to avoid stepping all over yourself.

Macros are deprecated.  Gosubs are the way forward, and yes, they have
local variables.  Simply define them once as Set(LOCAL(foo)=bar) and foo
will be gone when the innermost stack is removed (either by Return or
StackPop).

-- 
Tilghman

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Jay R. Ashworth
On Wed, Feb 13, 2008 at 12:52:42PM -0600, Tilghman Lesher wrote:
 On Wednesday 13 February 2008 11:45:34 Jay R. Ashworth wrote:
  Having programmed in about 8 different languages over the last 25
  years, I can see both points of view.  And admittedly, I haven't tried
  to do non-trivial things with dialplan.
 
  That said, my view of this interaction is that Tilghman has drunk the
  Kool-Aidtm, and that Alex's view of the situation is much closer to
  objective.
 
 Or maybe I'm just the architect of the dialplan moving forward, which is why
 I advocate that if you really don't need to use AGI, you don't.  ;-)

I'm not sure those aren't equivalent.  :-)

  dialplan appears to have jes' growed, and that never makes for a good
  language design.  Ask the Python 3 team.  :-)
 
 I'm specifically working on removing misfeatures from the dialplan, to make
 it much easier to use and more predictable.  1.6 will be a huge improvement
 towards this goal.

Well, this should be interesting.  :-)

Cheers,
-- jra
-- 
Jay R. Ashworth   Baylink  [EMAIL PROTECTED]
Designer The Things I Think   RFC 2100
Ashworth  Associates http://baylink.pitas.com '87 e24
St Petersburg FL USA  http://photo.imageinc.us +1 727 647 1274

 Those who cast the vote decide nothing.
 Those who count the vote decide everything.
   -- (Joseph Stalin)


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Jay R. Ashworth
On Wed, Feb 13, 2008 at 07:49:36PM +0100, Philipp Kempgen wrote:
 Douglas Garstang wrote:
[ ... ]
  do 
  with 
  a 
  bash 
  script, 
  as 
  opposed 
  to 
  Perl, 
  Python, 
  or 
  any
 
  toolkits 
  or 
  frameworks.
 
 Could you fix your e-mail client please?

I dunno; his message comes out fine here, though Mutt and lynx --dump.

I grow less impressed with T-bird by the day...

Cheers,
-- jra
-- 
Jay R. Ashworth   Baylink  [EMAIL PROTECTED]
Designer The Things I Think   RFC 2100
Ashworth  Associates http://baylink.pitas.com '87 e24
St Petersburg FL USA  http://photo.imageinc.us +1 727 647 1274

 Those who cast the vote decide nothing.
 Those who count the vote decide everything.
   -- (Joseph Stalin)


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-13 Thread Alex Balashov
Tilghman Lesher wrote:

 Like any other language, you certainly can write in an obfuscated way, and
 the dialplan does not discourage it.  That said, you certainly can write in a
 modularized way.  I would guess that you simply aren't familiar with the
 dialplan enough to make those decisions, but it is quite possible and doable.

The dial plan certainly does lend itself to this to some degree, no 
argument, but not to the extent that fully developed programming / 
scripting languages do.


-- 
Alex Balashov
Evariste Systems
Web: http://www.evaristesys.com/
Tel: (+1) (678) 954-0670
Direct : (+1) (678) 954-0671
Mobile : (+1) (706) 338-8599

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] LCR in Asterisk

2008-02-12 Thread Alex Balashov
Rizwan Hisham wrote:

 Hi all,
 I am planning to implement LCR routing on my already running asterisk 
 server. Uptill now i have found out that asterisk has no support for 
 lcr, i have to do something about it myself, for example using the AGI. 
 Im looking for ideas here. Whats the best way to start implementing lcr 
 in asterisk. Should i use agi and start implementing my own lcr script 
 or is there any plugin available which can be used with asterisk.

If you are interested in prebuilt solutions, you may consider 
TransNexus's NexOSS product (www.transnexus.com).  The Open Settlement 
Protocol (OSP) they implemented can be used with Asterisk - they have a 
module.  In fact, I am not sure about the commercial status of the OSP 
module as such;  it may be possible to get it free of charge.  Not sure. 
  But OSP is an open protocol, so perhaps it's possible.

Otherwise, I would think that the best way to approach this would be to 
make it fully outboard and divest it of Asterisk.  Implement a SIP proxy 
that forwards to providers using LCR decisionmaking, and just have 
Asterisk send calls to it.  OpenSER can be used for this - and indeed, 
there is an OSP module for it as well, if you wanted to go that route.

If you're dead-set on doing it in Asterisk and don't want to do OSP, I 
would suggest FastAGI.  Definitely don't implement the logic in the dial 
plan, at any cost.

-- 
Alex Balashov
Evariste Systems
Web: http://www.evaristesys.com/
Tel: (+1) (678) 954-0670
Direct : (+1) (678) 954-0671
Mobile : (+1) (706) 338-8599

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


[asterisk-users] LCR in Asterisk

2008-02-12 Thread Rizwan Hisham
Hi all,
I am planning to implement LCR routing on my already running asterisk
server. Uptill now i have found out that asterisk has no support for lcr, i
have to do something about it myself, for example using the AGI. Im looking
for ideas here. Whats the best way to start implementing lcr in asterisk.
Should i use agi and start implementing my own lcr script or is there any
plugin available which can be used with asterisk.

-- 
Best Regards
Rizwan Hisham
Software Engineer
Axvoice Inc.
www.axvoice.com
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users