Von Europe

2007 VoIP 2.0 Project Thomas Howe http://www.thomashowe.com

Short VoIP History • 1990 - H.320 and Px64 • Compression, packets and multimedia • PictureTel, CLI and DSP SE

• 1994 - H.323 • IP, Shared network resources, heterogeneous • PictureTel, NetMeeting, Intel

• 1997 - SIP and H.324 • Internet, applications integration, dumb networks • PingTel, Columbia, DynamicSoft

• 2001 - MGCP • Centralized control, real PSTN over IP • Softswitches, Softswitches, Softswitches, Softswitches

• 2007 - Web 2.0

Web 2.0 - 60 Second Intro • Web as Platform • • • •

Platform is no longer the server or application Not the Win32 API - the Google and Amazon APIs Double click - 2000 users. Google Adsense - 1MM Vonage - 1 million users - Skype - 75 million

• Harness Collective Intelligence

• You’re open source guys - you should know this

• Data is the new asset

• How you do it is less important as solutions vary

• End of the release cycle • Lightweight programming models • Software spans devices, networks, companies

A Fundamental Shift Applications

• • • •

• •

•

Today everybody has the same voicemail Little personalization and customization between services Tomorrow’s long tail; personalization; niche services

Architecture

Today we have stove pipe architectures...

Tomorrow we have loosely coupled web services, generic resources

Development

• •

•

• •

•

Today we have IMS architecture, large company view

Tomorrow we have disaggregated two pizza teams

Developers and Business Model

Today we have discrete and defined vendors, carriers and enterprises

Tomorrow we have open source, open opportunity, fuzzy participants

2007 VoIP 2.0 Project

A series of small projects designed to explore the integration of Web 2.0 and VoIP technology in an open source setting

Our Goals • To explore the undiscovered country • There’s a lot we don’t know • Very interesting real time issues

• To educate ourselves and our peers • Who and what you know

• To move our industry forward • So sick of replicating the PSTN • Phones suck; communicating doesn’t.

• To do some real geeking

Our Format • One interesting problem • One small group of geeks • Two pizza teams • Try to mix them up each time

• Projects no greater than 10 days total • Each geek gives a few days

• Publish the result to the community • Wiki documentation • Podcast • Full Source code

First Project : Turk Integration • Integration of Asterisk based voice mail messages with Amazon Mechanical Turks • Chop up voice mail into ten second increments • Human transcribes

• Asterisk already forwards you a copy of the voice mail… why not a transcription? • Excellent Web 2.0 Mashup • The asset? The message, not the voice mail system. • Uses the Web as the platform, not just Asterisk • Truly harnesses collective intelligence!

• Solid business driver • Transcriptions are searchable • Read them; forward as SMS? • May be start of larger applications, like conferencing

Second Project : ECC • ECC : Elastic Computing Cloud • Large supply of on demand processing

• Today’s switch engineering relies on Herr Erlang • What happens when we deploy a simple, but massive SIP application? • Use simple UA and proxy servers

• Excellent Web 2.0 Mashup • Leverage web based resources for optimal deployment • End of the release cycle? End of the whole thing!!! • Lightweight model - the network does the heavy lifting

• Solid business driver • Hardware costs lower, more manageable, better cash flow • Excellent underpinning for DDOS resiliency • Disaggregation + ECC == Infinite Linear Scalability?

For more information •

Check out the web site : • http://www.thomashowe.com • Smtp: [email protected]

•

Start to go deep on Digital Utilities • Google APIs • Amazon

•

Check in with us early next year • Still have a few spots open • Send an idea!

•

Take what we do and extend it. Please.