Jan 3

Tim O'Reilly

Tim O'Reilly

Asterisk an under-appreciated Open Source Success Story

Sparked by a recent posting to Dave Farber's IP list about a Texas university dumping Cisco's VoIP solution for Asterisk, and Asterisk-maker Digium's new round of financing, I wrote to Farber's list myself, extolling O'Reilly's own switch to Asterisk. The functionality has been great, with voice now just another internet application for us, with new features like voicemail to email forwarding and the like.

I was talking with CJ Rayhill, our CIO, the other day, and she pointed out that when evaluating new PBX alternatives, we'd gotten several $200,000+ proposals for proprietary systems. Asterisk was not only free, the VoIP switch has allowed us to increase our network bandwidth fourfold to accomodate the voice traffic (and more data traffic) yet still save $5000/month.

I've been puzzled why there isn't more focus on asterisk in the open source world, as it seems to me to be one of the really big new open source success stories.

It seems a bit like the early days of things like Perl and Linux, when they were happening under the radar, known to all the hands-on practitioners in the industry, but not covered much by the mainstream press.

Book sales are similar too :-) For the 15 months since it was published, Asterisk: The Future of Telephony has been #12 on our list of O'Reilly bestsellers. So like Perl and Linux before it, Asterisk is getting uptake ahead of its recognition by the mainstream.

Speaking of Asterisk, Mark Spencer, the founder of Digium, will be keynoting at our Emerging Telephony Conference, which will be held February 27-March 1 at the San Francisco Airport Marriott. The early registration discount ends January 8, so if you're interested, now's the time to register.

tags: open source  | comments: 10   | Sphere It

Previous  |  Next

2 TrackBacks

TrackBack URL for this entry:

Responding to Tim's comments:   Yes, Asterisk is the under the radar open source success story.  But what is great is people are taking notice one installation at a time and it will take a little more time.  The indicators I... Read More

Je viens de parcourir un article de O’Reilly à propos d’Asterisk. Cela me rappelle une expérience que j’ai menée avec Thierry l’été dernier. Il m’avait parlé de ce PABX pratiquement entièrement logiciel et gratuit. ... Read More

Comments: 10

  Mike White [01.04.07 04:31 AM]

Hey Tim- Have you seen

  adamsj [01.04.07 04:38 AM]

Possibly one reason it's not taking off like it might've could be the rapid increase in costs for the hardware. I'm still kicking myself for not getting the development kit, back when you could buy it for a couple hundred dollars, and the cards were in the hundred dollar range each. Now, just to try it out, you're looking at an outlay of way more. This decreases the number of people fooling around with it, at home or at work. If Asterisk is an ecosystem, this makes their algae awful pricey.

  roderickm [01.04.07 08:15 AM]

The cards have been significantly improved since the days of the developer kit, adamsj, so the modest price is understandable. I know one Asterisk user that recovered the cost of their analog interface within three months because of the radical cost savings over traditional telephony.

Still, you don't really need a hardware interface to get started with Asterisk. AsteriskNOW is an easy way to test drive Asterisk in 30 minutes or less, and you can use softphones to make your calls.

Others use Asterisk as a voice application server, like a sidecar to an existing phone system, to provide services such as MeetMe conferencing or complex IVRs that would cost a ton (or are simply unavailable) from a legacy PBX vendor.

I got started with Asterisk when I needed telephony capability in a web application. Asterisk has very useful APIs for programmatic call control, allowing Asterisk to interact with your existing databases and installed applications. Can you tell I'm a fan?

  Jim Van Meggelen [01.04.07 08:41 AM]

While the cost of hardware for an Asterisk system can get expensive, two things should be kept in mind:
1) Compared to traditional telecom hardware, Asterisk components are very inexpensive. The reason the price has increased is due to an increase in quality. You can still buy a zaptel-compatible FXO card for about $20 on eBay, but no one would recommend putting such a thing into production. Asterisk has become more serious; so has the hardware.
2) It is possible to build a system that does not require any hardware other than a Linux server. This will require a network that is up to the task, and a VoIP provider who is asterisk-friendly, but it is very much do-able.

What may be one of the biggest barriers to entry is that there are a whole new set of skills to be learned, and this means many hours of R&D before something workable can be achieved. Having said that, I think there is a lot more Asterisk going on than is immediately obvious. Much of the value of asterisk is in the corporate world, and they will not always publicize their results.

The other thing that keeps asterisk off the radar is that phones just aren't that sexy. We use them to make phone calls, and that's pretty much it. I love asterisk, but let's face it; YouTube is a lot more fun.

  adamsj [01.04.07 09:09 AM]


I'm not doubting that the newer technology is worth the money.

I am saying, given a slightly higher ground-level barrier to entry (not a problem with Tim's other two examples, Perl and Linux), it's harder to create the groundswell effect Tim's looking for.

  roderickm [01.04.07 12:29 PM]

True, the cost of a hardware interface raises the barrier to entry relative to Perl and Linux. That's why I pointed out that the hardware interface is entirely *optional*.

With AsteriskNOW, a free softphone, and an internet connection, you can go from a bare system to a fully functional, GUI-driven PBX in under 30 minutes. With no hardware interface. Making real calls to the PSTN. That's every bit as accessible as Perl and Linux.

I agree with Jim -- to most folks, phones (even those connected using Asterisk) just aren't sexy tech relative to YouTube. But Asterisk will continue to shift the cost structure, ability, and extensibility of phones anyway. It won't stay quiet forever.

  Brian West [01.04.07 12:37 PM]

Asterisk makes a great PBX but not a great SoftSwitch.


  Thomas Lockney [01.08.07 05:30 PM]

It seems to me the biggest reason Asterisk isn't getting as much notice is simply because of the marginality of PBX tech and telecom in general. I mean, even in a company where the IT staff might be entirely made up of linux and opensource software fanatics, it's still likely only one or two people on the team ever deal with the phone systems.

Granted a lot of this is changing as VoIP is becoming more and more ubiquitous, but unless Asterisk finds it's way into usage by everyday opensource developers and users, it's just not as likely to get much more notice.

That said, I am quite a fan of the project myself and have dreamed up (though not implemented) many a project that could take advantage of it.

  Dal [03.04.07 12:01 AM]

Responding to the "The other thing that keeps asterisk off the radar is that phones just aren't that sexy." comment.

Last time I checked Polycom IP Phones are pretty sexy.

  Alexander Rose [04.19.08 04:05 AM]

Tim, when do the hash discussions begin? What about cannabis, will that get a mention too?:p

Post A Comment:

 (please be patient, comments may take awhile to post)

Type the characters you see in the picture above.