Results tagged “skype” from Emerging Communications Blog

Chair: All right, Skype and Asterisk, what does it mean for business communications? Founder and CTO of Digium, Mark.

Mark: Thank you. Asterisk and Skype is probably not a pairing that a lot of you guys might have thought about a year ago. You would have thought about, "Wow, it would be nice," but didn't seem much like a reality. It is exciting that we have it now getting closer to being a reality.

First of all, what is Skype for Asterisk? You're used to SIP and SIP let's you talk to SIP devices, and Skype let's you talk to Skype devices, which pretty much is, of course, just Skype. It is a generic channel driver for Asterisk, so it supports most of the things that Asterisk channel drivers already support. There is no special programming or anything; you just install it an you can use it.

It supports, of course, the usernames, encryption, end points, and it supports both talking to regular Skype names, any arbitrary Skype name, as well as talking to the SkypeIn, SkypeOut services.

It's really, the first practical Skype gateway from a PBX platform. It allows you to connect this really broad user base of people that are already using Skype, with Asterisk. If you think about Asterisk as a very pragmatic and practical platform for telephony, for business phone systems, Skype has been incredibly successful in the Voice over IP space because it's been a very pragmatic solution for customers to be able to use.

I've often said that the lesson of Skype is really that when you look at open source, open standards, and all those things, their value is really only as strong as their ability to deliver on ease of use, performance, and low cost to the consumer. Even though Skype wasn't open source, wasn't open standards, or any of that stuff, because they were able to deliver value to the customer they were able to get such a tremendous amount of following and a number of users who are obviously accessible to you, through Asterisk.

There have been some other very ugly hacks that involve virtual machines and emulating sound hardware and stuff like that, but the Skype for Asterisk is literally just native Skype code running within the Asterisk environment so it removes this huge, ugly hack factor. It's actually scalable and it integrates nicely in Asterisk with behaviors you would expect.

There are several use cases to go over. The business call centers, of course, is one. Now, anybody who has Skype can contact you and you could register a Skype name for your business and have people call directly into there. It integrates very nicely in and gives you a low-cost way for your people to call you without having to run up 800 number minutes. A lot of customers already have Skype and you could have click-to-dial, and all that kind of stuff.

The other thing, of course, is business PBX. If you like to use Skype for your business communications, which a lot of customers do, even though the IT people typically don't; you can use SkypeIn and SkypeOut minutes to associate with your PBX and you can also have your Skype username that both comes to your Skype client natively, and rings over to the Asterisk PBX. For example, you could have everything unified on the same voice mailbox. When someone calls you, it can ring both your hard phone and your Skype soft phone.

Of course for end users, any application that you have that you would want to expose to a wide variety of users, you could do so via Skype. If you wanted to make a call for the weather, or whatever it is, any kind of IVR, you could connect all that in and allow for calling people back via Skype, as well.

The status of the Beta - because of the complexity of the integration, we decided to start with a closed Beta. We were only going to open it up to a certain number of people, initially, so that we could have engineers working directly with the customers. That started in January of this year. We have over 100 people, but less than 1,000, so far, in the Beta. We will be going to a public Beta, hopefully, very soon. So far, it's been working reasonably well.

There are some big caveats that are very important. First of all, Skype is also in the process of releasing something called the Business Control Panel. Although it's not implemented in the current Beta, Skype is requiring that the usernames you use to register your device with Skype, in other words, the ones you use with the Skype for Asterisk, will all have to be business control panel accounts, which I believe means you are not going to be able to use existing accounts unless you are somehow able to make them part of the business control panel.

This is something that Skype has demanded, so feel free to go tackle him, over there, if you have any concerns about that. There is not much we can do on it, right now. However, there are areas that we are interested in hearing more about, like chat and video. We would definitely like to get your input on priorities about how those features would want to be integrated in, in later revisions of the product.

You dial it kind of as you would expect. It's just another channel driver and you do Skype slash the username. Presence is supported. We do have some AMI events that are generated, to give you a little bit more ability to kind of hook in for some of the Skype-specific stuff that wouldn't be present in other telephony interfaces.

You actually have access, as it turns out, to a lot of the variables in the Skype call. For example, you could use the language that is provided by the user in their configuration to give them IVR in their native language of choice. Some of the others, I guess you could wish them a happy birthday if they called on their birthday. I don't know exactly what you would do with that. You can see there is a lot of demographic information that may be able to be helpful, if for nothing else, from a logging and statistic gathering point of view.

There is a shortened URL, if you want to sign up for the Beta for when it does go to the public Beta. I will save the rest of the time for questions. Your questions can be about Asterisk generally, or specific to the Skype for Asterisk.

Audience 1: Is this going to use the Asterisk jitter buffer, or Skype's jittering, fancy de-jittering technology?

Mark: It's going to use the Skype de-jittering, but it will use the Asterisk native codecs. Essentially, all the existing codec work you do with Asterisk, already, will be used. That is all native. It doesn't have to get retranslated.

Audience 2: What caller ID will it actually pass?

Mark: The caller ID will be based on the Skype username that you are using to place the call. You can register multiple Skype usernames with it and then say, "I want this call to come from this particular Skype username". On inbound, it will obviously be the phone number associated with the caller. Typically, it would just be their Skype name.

Audience 3: So if you put it under the corporate name...

Mark: Right, it would go out, for example if it was Digium, the word Digium would be the caller ID and then the name would be whatever name you had associated with it.

Audience 4: You can use a mobile number as your Skype caller ID, instead of your regular Skype ID. Will the...

Mark: I'm not really sure if we've set it up for that. I hadn't thought about that, in terms of allowing your mobile number, whatever you have listed in Skype as your actual phone number...

Audience 4: It becomes handy for getting into things like CauliFlower, and so on.

Mark: Okay, I will have to get back to you on that one. You stumped me on that one.

Audience 5: Mark, if you're phoning someone like 1-800-go-fedex through Asterisk to Skype, does early media work, and can we get the DTMF into that?

Mark: The early media is only unidirectional, right now, in Skype. I don't believe you can transmit media before the call is through. FedEx, as a specific example, probably wouldn't work, although now that you mention it, I don't know how you do that through SkypeOut anyway. That's a good follow up question.

Audience 6: There was a question about the jitter buffer. Can you elaborate a little bit about the media exchange, going from a SIP endpoint to Skype? It's starting out on RTP, and then are you guys doing a virtual driver, or are you writing a wav file to a file and then playing it out?

Mark: The audio comes in over RTP, into Asterisk, and Asterisk routes it out the Skype interface.

Audience 7: I have two questions. One has to do with security. At what point does the Skype call cease to be encrypted? The other question is will you be allowing Skype-to-Skype re-routing, as opposed to a PSTN to Skype? What can I do, through this application, transfer a call to another Skype user?

Mark: The first question is when does it get unencrypted. It gets unencrypted when it hits Asterisk, obviously it has to be to be able to convert to other media. I think the second question was can you have a Skype call in and then send a Skype call back out. The answer is yes, you can. I don't know if it's going to be out in the first release, but there is a way to transfer that call off. Otherwise, you would be in the middle of that call and you would essentially be getting the media, decrypting it and re-encrypting it on the other leg of the call.

Audience 8: Mark, is there a charge for the [0:11:56.7 unclear] Skype? Has that been determined, yet?

Mark: The current plan is that they would be sold, more or less, like the G.729 licenses, but there hasn't been any kind of formal price announcement or anything like that. Hopefully, as we get closer to the final Beta, we'll be able to confirm what the terms of that will be.

Audience 9: I'm just curious; there is something I heard about, called SkyHost. Is that part of this solution?

Mark: That's basically a name for the API that's being used, yes.

Thank you very much.

Chair: Great job, thanks Mark.

Skype has kindly sponsored the 2009 conference. We are grateful for their support. 

I asked Jonathan Christensen an hour ago why Skype decided to back the event. He responded:

I had to follow temptation by following that up by asking if we can expect Skype to debut something at the 2009 conference? His response was:

So there you have it; Skype sponsored eComm 09 and plan to have a "big year" come 2009. Can't wait. Now we can start the guessing game...

Transcript below from Jonathan's opening keynote at the inagural eComm 2008. The corresponding slides can be found here. As a side note the complete comference video and audio is pretty much ready to go but a few minor niggles are holding it up.

Chair:    We are going to start this conference with Jonathan Christensen who is the general manager of audio and video at Skype. Jonathan is going to kindly warm this conference up by speaking about the history of IP communications over the past 10 years. So, I would like you to welcome Jonathan please.

(Applause)

Jonathan:    Thank you. So, I am going to try and give the 30-minute version of his 15-minute intro. I am Jonathan and I have been with Skype for about 2 years. I lead up our audio-video research and development that is spread across a couple of continents and a lot of time zones. I'm building world class software that we hope for next generation communications. I am just going to give a little bit of an overview to talk about the last 10 years from my perspective. The things that I have seen, the major inflection points in Voice over IP and then hopefully try to move in to the "what's next?" stuff'.

A lot of people get credit for the first VoIP instance. But I think that around the industry, most people give credit to VocalTec and their Internet phone for being the first usable network-based Voice over IP client and this was a desktop phenomenon. Then came a lot of people rushing into the market once the concept was proven, to build gateways. And, we are going to talk about these in a little bit. Then, there were some carriers that came along and they established the gateways and POPs and they connected it to the PSTN and they created a market for moving VoIP minutes around.

And along with that, there are a bunch of other lower level innovations that started happening. Companies like Global IP Sound building next generation Codecs that were Internet aware that offered wide band and that did interesting things like that, sort of changing the nature of the communications a little bit.

And some use cases started to gell, first there was this PC to PC, this kind of the ham radio users. Jeff Pulver, a name that is sort of synonymous with Voice over IP and the whole VoIP revolution was a ham radio operator in his youth. He clung on to this idea when he first saw it and he saw that there was something really interesting here.

The next use case that really drove the early days of VoIP was tandem trunking. I would go into some detail about tandem trunking and what is interesting and what is not so interesting about it.

Along with that a couple of mainstream apps emerged, you had Netmeeting from Microsoft - how many people in the audience ever used Netmeeting? How many people in the audience have ever used Skype? Well, Netmeeting was first introduced in the late 90's. It went through three revs and it was last updated in 1999. I was involved in that work at Microsoft. Amazingly, that application is still widely deployed and used especially inside large organizations today without a single update to the basic application in almost a decade.

The other use case that started to emerge was two-staged dialing. This is a notion that you call a number, usually an 800 number, you enter a pin and it gives you access to the network to make cheap calls. Most of that was built on top of VoIP infrastructure because at the middle of the call, we could use as tandem trunking and you could avoid a bunch of the cost. So how this worked in the early days is, if you are on the West Coast here and you wanted to make a call to the East Coast, your call would be connected to a class 5 switch or line side switch and that call would be routed to a tandem switch - and this is an over simplification of course - and the call would go over TDM, a traditional network to another tandem switch to class 5 switch and the phone on the East Coast would ring.

VoIP tandem switching replaced that center part with VoIP gateways. And the interesting thing about this is that it really was the death of physical distance. But much more than that, it became the death of the business model behind distance. You have some bandwidth efficency and that was kind of interesting. The local exchange carrier could avoid the LD carriers along haul lines and that was kind of interesting. The incumbent long distance player could then avoid the termination fees at the LEC because the call did not arrive over long distance or telephony service but it arrived over an information service that was unregulated and this is a sort of regulatory trick. And what we really saw was a whole era of arbitrage and many would say inferior service.

Suddenly, you are transcoding or re-encoding using the network that maybe was not suited for real time communications and all sorts of things like that. Vagarious of the network that were introducing inferior qualities.

Most of the time, these offerings were offered under alternative brands. Some of the big guys did not want to marginalize their brand. You could hear a pin drop or their brand had been around for a couple of decades and in some cases, 100 years. And, they did not want that to be associated with these services so they came up with wacky names and wacky marketing schemes and prepaid calling cards and what not. So, those services were marginally interesting if you were trying to save money and if you are a college kid or that kind of thing.

So moving on, the introduction of consumer VoIP really happened when Jeff Pulver and some other industry luminaries started getting together and thinking about what they could do here. In the early inspirations this "minex" thing that Jeff got together was the idea that minutes were commodities and they could be traded. He was a trader at Cantor Fitzgerald early in his career. He thought, "Hey! There is something here that we could build this trading desk for these minutes and we could get something going there." He met Jeffrey Citron in that process and he was also toying with a thing called Free World Dial-up. How many people are familiar with Free World Dial-up? Pretty much, that is remarkable. I was a user. I had a one-port gateway installed and people made calls over my Seattle local line and those calls were flat rate billed. So I did not get billed for them and other people out of LATA [Local Access and Transport Area] got free calls on my infrastructure. It is a pretty interesting little application but did not really scale. It was hard to set up and what not.

The Komodo was an interesting device, the first little ATA that enabled this Free World Dialup scenario. Jan Fandriato built this little company and I like to joke that it turned any perfectly good analog phone into an inferior IP handset. Cisco acquired Komodo and really did a lot to validate the market. Suddenly, people were like, "What just happened?" Out of this mix, out of this soup, Vonage was born and the Voice over Broadband revolution began.

Vonage really set the early pace. They had lots and lots of cash, they were extremely aggressive and they went to market direct to the consumer. The long distance players entered to get into local service markets with unregulated services. We saw offerings like this cropping up. And then the MSO's, the cable companies responded. Generally, all of the people who entered into this arena were the people who did not have anything to lose. They did not have something to cannibalize.

So what did we get out of all this? Well, we got these attractive pricing plans I guess, these bundles, great savings - maybe I am not sure. New features, well you have call log, voice mail and a whole bunch of stuff that I think does not have much to do with the fact that it was Voice over IP. And many of these features existed before these services were launched. Still, it is a question like, what really happened here? My hypothesis is that for and as Lee [Dryburgh] said, it is a sort of foregone conclusion that the phone is dead. This is another foot in the grave, another beginning to the end. So the incumbents of course come in and they matched the prices and they start up a price war and they do that on TDM and some mash up with some web stuff if you want it. The landline telephony game becomes cheap, boring, stagnant and at the same time losing the mobile.

A couple of years ago I called my mom. She has had the same phone number since I was little kid. I got the recording, "This phone has been disconnected." And I thought, "What happened?" We have had this phone number since I was a little kid. So I wrote her an email and she responded and said, "What do I need that for? The only phone I use anymore is my mobile phone." When your mom moves into that segment, things are changing for sure. I think that that is a good sign that the commoditization of that market is total and complete at this point. And I think that the other thing to look at is, where is the next wave of innovation is going to come from?

I will go into some more of that but one of the interesting things that came out of this, it was not really by design, was VoIP export networks that started to crop up. People would order a couple of Vonage sets or choose your provider. And they would put one in a box and they would send it overseas and those two sets would be set up with the same numbers. So suddenly, you have phones ringing in India on a Silicon Valley number, right? And, they could call their relatives and create these ad hoc networks within a community and essentially export the numbers.

This is important because this is really one of the final nails in the death of distance, right? Not only does it destroy distance with the Internet being the pipe that is carrying this traffic but it also totally fragments the numbers. The numbering plans have really been used by the operators to discriminate both on price, politically and geo-physically. All of those things are just part of the old world and the numbering plan is a big part of keeping it all together. This again is further deteriorating the traditional voice business.

In 2003, I was leaving Microsoft, I was standing on a piece of land in California that I was thinking about moving to and I got a call from Jeff Pulver. I had called him in a beautiful sunny afternoon. I totally remember it. And he said, "Hello Jonathan. What is up?" This is like June of 2003. I said, "Jeff, we have to build a client that is super lightweight and simple and just works. It gets over the NAT traversal issues. It just works. People can just download this thing and make it work." And Jeff said, "Well you know, Free World Dialup works and the clients are getting better all the time and there are third party providers who are billions." I said, "Jeff, Jeff, Jeff, look, it is way too hard to use. We just need to build the client that works."

A few weeks later, a friend of mine, Allan Duric who was working at Global IP Sound at that time got me an early beta account, pre-release account of Skype and the lights went on. And that was in the summer of 2003. By Fall VON, there were 500,000 registrations. I was in Boston and everybody was talking about Skype. And most people were pretty angry about it especially the "SIP die-hards". Skype had made some early noise about, "Oh well, we do not use SIP because we do not need to and that really ticked off people who have been working in this space for a while.

Let us switch gears a little bit and talk about IM as the foundation of this next wave. In 1999, I was in Helsinki or Stockholm or some place like that, I was working for Microsoft at the time, I was on the MSN network and I was talking to a customer on the East Coast and one of our support people in Redmond. I was chatting. It was the middle of the night and we are having this three-way conversation on multi-chat. And just at that moment, it dawned on me that aside from being funny, I Am therefore IP, IM is the run away next generation signaling infrastructure. We were all talking at that time about what is the next signaling infrastructure. Is it AIN [Advanced Intelligent Network]? From my job perspective at the time, how is Microsoft going to participate in that next generation? What are the softswitch guys are doing so on and so forth? But here, I was having a conversation with people all around the globe. I was having that through NATs and firewalls and it worked. I was having it in real time and it was widely deployed already.

At the intersection of VoIP and IM in 2003, Skype appeared. And adding to that voice and visual sharing and rich communications, pictures, all of the other things - web links all of the other things that you want to share in a real time session is really the foundation for next generation rich communications.

How did this all come together? Well, in the summer of 2003, the solution was really 99% complete. And I think that this is why people at the Fall VON show were upset when there was so much overwhelming hype about the introduction of Skype. Because they were all saying, "Hey Vocal Tech did this years ago and we have chat in our IM client and so on and so forth. There was a robust audio stack, multi-media PC's were well-penetrated, broadband penetration was on the rise and by that point, well penetrated enough around the world to be a platform for this.

The P2P file sharing network had introduced new paradigms for low cost infrastructure. The NAT traversal techniques were well-known at that point. Maybe not deployed, but well-known. For example, in most of the voice chat scenarios of the incumbent IM players, voice chat failed 30-40-50-60% of the time if there was a NAT in the picture. And so, people did not use it, right? If it fails 2% at the time, you are not going to try it again.

The IM networks were well deployed for sure. All that Skype did was that they closed the loop. They added the other 2% at the top and they made this thing work. One simple little application, good formula, the users really liked it as I said; the industry was sort of confounded and by fall of 2003, there were 500,000 downloads and this thing was on an incredible trajectory.

Today, 276 million registered users, 30 million of which were in the last quarter of 2007. We have global appeal in reach, reliable PSTN interconnect. We have lots and lots of devices supported. Things like PSP's and N800's and cordless phones and all sorts of devices and more on the works. We're profitable. And so, I would say that we have entered the era of rich PC based Internet communications. This is a time when we are finally seeing the vision of the original SIP founders of multimodal rich communications being deployed to user's desktop in a usable, cheap, most of the time free, way. This includes real time video, data, presence, text and wide band audio. Wide band audio today is more widely deployed than ever before and the majority of those minutes are happening on Skype.

We are also talking about a paradigm shift where the smart platform on the end point is what is driving innovation as Lee [Dryburgh] talked about versus things coming from the network or the provider in the core. So what is next? I think a big part of this and Lee also talked about this, is sorting out the mobile mess. So in my view, mobility is the last anchor to the old way. It is an incredible innovation to be able to pick up your phone and take it with you. And as a result, it is the fastest growth telecom service today. But the spectrum scarcity and the other issues make it a perfect world garden. In a lot of ways, it is back to the future. It is the good old days again, closed networks, device lock in, phones with these numbers, these pesky numbers, geographically oriented contract, honor some contracts. Basically, the old world.

But maybe there is some hope here. And for many, many reasons, the flat rate pricing, the auction of 700 [Mhz] bands spectrum and where open platform conditions apply. Maybe a new game is beginning here. And, this is really one of the most exciting times in the communications, I think. There are still lots and lots of hurdles. We have to make sure that these guys do not do things to discriminate against applications, that the principals of that neutrality continue to apply and we are vigilant about that as a community. But maybe, a new game is here.

So for the next 10 years, I think that there is the possibility that we will enter the era of rich Internet communications. Again, the same list as before, but add to that new mash ups of fixed mobile convergence of web-based communication and voice and video and other real time communications perfectly mashed together in ways that developers can recreate and mix and match applications. We will have freedom to these applications on a new platform that includes mobility.

And finally, I think that we all agree, I hope, in this room that there is a natural segmentation of competencies that emerges from these, one that the industry has been fighting against for quite sometime. But that network infrastructure and pipes and the competencies associated with rolling that out are different than the competencies around application, innovation at the edge of the network. Let us let those competencies lie where they are. Let us let them blossom where they are and move forward in that way.

So I think I kept within the time. If there any questions maybe we can do that.


Jonathan:    David if you if you ask me that question I will repeat it.

Audience:    The question is, how is the Skype-eBay relationship going?

Jonathan:    Okay the question is how is the Skype-eBay relationship going? From my perspective very well, we recently got a new CEO who is a medium term eBay guy. So today, with 276 million registered users, I think that there is less focus at eBay today on finding the place where eBay and Skype intersect on the web and mash-up to create a new application paradigm for eBay or communications paradigm for eBay and more focused on Skype growing its business and eBay growing its business. So can I answer your question?

Audience:    At one point, Skype was really the future or the leading edge that went into eBay. So, what are Skype's future plans? How is Skype going to grow its business?

Jonathan:    If you did not hear it the question was, after being rolled into eBay and now, kind of looking at the landscape, what are the future plans for Skype--

Audience:    Tell us all your secrets?

(Voice overlapped)

Jonathan:    I think that in some ways, we stalled. We got wrapped into the M&A [Merger & Acquisition]. There is almost always a period of integration that a lot of weird things are tried and some work and some don't, and there is defocus. We obviously had some management turn over and what not. For me personally, this is the most exciting period for Skype going forward. The projects that I am leading in my team and that we are working on for the next 2 years or 3 years I think are ground-breaking projects. And that sense of innovation and hard work and startupness is very much alive in the company. I cannot really talk about the specifics, the things that we are going to be rolling out but I am really serious when I say that I am personally excited about what we are working on. Like almost no time before my career.

Chair:    That sounds terribly exciting. If it is okay, we have time for one or two more questions if Jonathan is good?

Audience:    Hi Jonathan. This is Irv Shapiro from IfByPhone. As a voice application provider, we would like to provide highly scaleable applications on the Skype network. At the current time, we find it complex to interconnect to the Skype network at a high volume level with hundreds or hundreds of thousands of users. What is in the works to make it easier for those end points to blossom in the Skype network infrastructure?

Jonathan:    This is a topic that is near and dear to me. I participate in the discussions internally about platform versus application and I guess it is early to make any definitive announcements or anything like that. We are still grappling with the issues there. But I would say that it is a very logical next step for us that Skype will enable these kinds of interfaces. I think that our API and developer story thus far has been sort of interesting but not in the realm of the scaleable applications that you are talking about. I and many of us in the company really recognize that. We think that there are easy solutions but they take shift in mindset to be able to get those out to the community.

And, we have to be very careful as well. The last thing that we want to happen is something that is detrimental to the health of the network.

Chair:    One last question for Jonathan please. Any last questions?

Jonathan:    This one over here.

Audience:    (Inaudible)

Chair:    Can you repeat the question?

Jonathan:    The question was about the network outage in December and what we have done to fix that. I am not involved in the core team work around the P2P network and I do'nt have specific knowledge of the fixes but I do know that they isolated the problem. They have put that patch out to the network and they are confident that things would continue to scale.

I do not know how much you have read in the blogosphere about it but it was a very interesting event. It was sort of an event at the leading edge of peer-to-peer networking and computer science in many ways. This anomaly that was caused by the restart of a lot of computer simultaneously created vulnerability in the way that the core infrastructure the network is supported. And, it was just sort of this spiral. Once that it started, it got worse and worse and worse. As far as I know, we have addressed those issues and that anomaly and would not be affected by future scenarios like that one.

Chair:    Jonathan, thank you very much for coming. Thank you very much for opening. And, thank you very much for fielding very difficult questions. Please thank Jonathan.

(Applause)

Jonathan:   Thank you.