From yitzhakbg at gmail.com Fri Dec 14 05:02:25 2007 From: yitzhakbg at gmail.com (Yitzhak Bar Geva) Date: Fri, 14 Dec 2007 12:02:25 +0200 Subject: [Telegraph-users] Adearsion/Telegraph cooperation instead of competition Message-ID: <3c0677320712140202v7626a057wd5916c3cc77ea569@mail.gmail.com> Wondered whether we're not going to end up with a duplication of effort between Telegraph and Adhearsion whereas cooperation between Adhearsion and Telegraph as a single project could result in an immensely improved framework. What benefit does two seperate projects bring while we could be concentrating the efforts to a better end in the context of a single project? Feedback from the user community is warmly welcomed -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/telegraph-users/attachments/20071214/5df227f4/attachment.html From jpalley at gmail.com Fri Dec 14 06:39:02 2007 From: jpalley at gmail.com (Jonathan Palley) Date: Fri, 14 Dec 2007 19:39:02 +0800 Subject: [Telegraph-users] Adearsion/Telegraph cooperation instead of competition In-Reply-To: <3c0677320712140202v7626a057wd5916c3cc77ea569@mail.gmail.com> References: <3c0677320712140202v7626a057wd5916c3cc77ea569@mail.gmail.com> Message-ID: <81A57B3A-5A84-4769-879E-A07C0D141F97@gmail.com> Hi Yitzhak - Jay and I discussed this a while ago. At the time it was pretty clear that the use cases/syntaxes/DSLs were very separate. Telegraph is designed for adding voice functionality to Rails application and the key design goal was to make it really tightly integrated into Rails. Actually, Telegraph is an extraction from what we use at my company and has suited as well for doing all the voice functionality on top of the very large Rails system we run. I can't speak for the Adhearsion team but it certainly seems that the goals and ideal use cases of Adhearsion are for making a framework for building voice applications independent from the web. From an architectural and design standpoint there is a large difference. Having said that, there is a large duplication that I would like to see remedied: the core AMI/AGI interface libraries can certainly be shared. I think this would be a good thing. For instance, I noticed that Adhearsion has listed as one of their TODO's that the AMI module needs a rewrite. I've offered this before (not sure why it wasn't picked up), but the AMI module we've been using in Telegraph is ugly but very stable. We've been running it basically nonstop for a year now and the only problem was a memory leak we fixed a few months ago. I believe this is a good example: the AMI stuff Asterisk sends back is messy enough that there is no reason to have multiple libraries doing the same thing. Having said that, we are working on refactoring it and the AGI stuff again to using EventMachine as we can achieve much higher throughput/ lower CPU usage and cleaner code this way. We are also abstracting out so we can drop FreeSWITCH support in. We'll release it when we have a chance to clean it up and get it out. Again, turning this libraries into a shared library would certainly be ideal and I would encourage it. My main prerogative is of course writing stable and effective code/ DSL's that suit my companies Rails based system (which is a complex distributed call center/oral language learning/international calling system). If someone can champion and shepherd a proper extraction of these core libraries we will certainly contribute and run it with Telegraph on our main systems. I would love to see it happen. Thanks for starting the discussion. Jonathan On Dec 14, 2007, at 6:02 PM, Yitzhak Bar Geva wrote: > Wondered whether we're not going to end up with a duplication of > effort between Telegraph and Adhearsion whereas cooperation between > Adhearsion and Telegraph as a single project could result in an > immensely improved framework. What benefit does two seperate > projects bring while we could be concentrating the efforts to a > better end in the context of a single project? > Feedback from the user community is warmly welcomed > _______________________________________________ > Telegraph-users mailing list > Telegraph-users at rubyforge.org > http://rubyforge.org/mailman/listinfo/telegraph-users From mike at infoanywhere.com Fri Dec 14 08:57:38 2007 From: mike at infoanywhere.com (Mike Clark) Date: Fri, 14 Dec 2007 08:57:38 -0500 Subject: [Telegraph-users] Telegraph & Realtime Asterisk Message-ID: <47628BD2.9000107@infoanywhere.com> Does Telegraph play nicely with Realtime Asterisk? We are using rails to build a GUI interface to Realtime Asterisk and Telegraph does look very interesting as an additional tool. However, I wanted to make sure it did not disrupt anything in Realtime or a standard rails environment. Thanks, Mike Clark From jpalley at gmail.com Fri Dec 14 09:08:55 2007 From: jpalley at gmail.com (Jonathan Palley) Date: Fri, 14 Dec 2007 22:08:55 +0800 Subject: [Telegraph-users] Telegraph & Realtime Asterisk In-Reply-To: <47628BD2.9000107@infoanywhere.com> References: <47628BD2.9000107@infoanywhere.com> Message-ID: <45180F54-77E3-478C-A6ED-718E56814906@gmail.com> You have nothing to worry about. Telegraph doesn't touch Asterisk Realtime (to get Rails integration just point Asterisk to your rails db and add the appropriate models to rails). It also won't disrupt your standard rails environment at all (at least 1.2...hasn't been tested with 2.0 yet..will soon). JP On Dec 14, 2007, at 9:57 PM, Mike Clark wrote: > Does Telegraph play nicely with Realtime Asterisk? We are using > rails to > build a GUI interface to Realtime Asterisk and Telegraph does look > very > interesting as an additional tool. However, I wanted to make sure > it did > not disrupt anything in Realtime or a standard rails environment. > > Thanks, > > Mike Clark > _______________________________________________ > Telegraph-users mailing list > Telegraph-users at rubyforge.org > http://rubyforge.org/mailman/listinfo/telegraph-users From timuckun at gmail.com Sat Dec 15 03:13:59 2007 From: timuckun at gmail.com (Tim Uckun) Date: Sat, 15 Dec 2007 21:13:59 +1300 Subject: [Telegraph-users] Adearsion/Telegraph cooperation instead of competition In-Reply-To: <3c0677320712140202v7626a057wd5916c3cc77ea569@mail.gmail.com> References: <3c0677320712140202v7626a057wd5916c3cc77ea569@mail.gmail.com> Message-ID: <855e4dcf0712150013l46431dafn48fef8e94b18a53c@mail.gmail.com> I think Jonathan pretty much nailed it. These are two different projects with very different aims but some code can (and should IMHO) be shared. I should also point out that there are two other frameworks. Rastman: http://rastman.rubyforge.org/ Asterisk Ruby Framework: http://asterisk-ruby.rubyforge.org/docs/1.0.0/ Obviously a lot of effort has been duplicated by these projects and it would be nice if somehow the projects could co-ordinate with each other.