I love Microsoft products (except DRM-crippled products... more on that later). So, it hurts me when MS doesn't play its full potential. Area of this time: VoIP.

I've been doing VoIP full-time for over a year and a half. Not exactly a long time, but not too short in this market. I think a lot of people miss the most important part of VoIP. The use of packet-switched versus circuit-switched goes way beyond being able to use cheaper transit and hardware. The real power that IP brings is that the shift can move from hardware to the software. As VoIP platforms become available, the power shifts because voice is finally something every developer can control. This goes back well before the current wave of VoIP-craziness and super-inflated company valuations.

Microsoft was here quite some time ago. NetMeeting was released in 1996 and had a decent API (Well, I've not used it, but the docs seem to indicate a lot of power). Messenger continued this (well, except I don't think the API allows you to do anything interesting). So Microsoft was certainly there back then; this isn't a completely new territory.

Software means no needing to spend $$$$ on all sorts of features. For instance, many vendors sell a “voicemail module“ at a decent price. What exactly is a voicemail module? Isn't it nothing more than a simple IVR + recording function? Sigh.

Skype
Hence, you might now understand why I think Skype is silly, or more to the point, MS screwed up again. Messenger (and NetMeeting before that) has had all Skype's features for years (including PSTN connectivity). The *only* thing Skype did was allow audio to be proxied. Yes, that's their only interesting feature. Had MS simply done that, perhaps Skype would never have surfaced. I remember the first time I heard about Skype. Someone IM'd to tell me to download it so we could voice chat instead of IM. I was really puzzled and pressed the Audio button in Messenger and asked “Like this?”.

Apart from the audio proxy problem, many users just didn't know that audio chat in Messenger exists. I don't know people can miss this, but whatever... Just like some folks are pretty amazed at what Netmeeting can do... even more so when I say it's from 96. Microsoft simply failed to market these features. Now they have to play “catchup” and do announcements like “VoIP will be in Messenger Live!”... That's just.... lame.

Speech Server
Speech Server seems pretty cool. Indeed, multi-modal app development and, well, speech recognition, are pretty cool. However, instead of making a killer product that allows you to build whatever app your mind dreams of, they tied themselves to horrid hardware, essentially rendering this product way more expensive and inaccessible. Intel Dialogic? Please. SIP is only available as a 3rd party addon. I asked around to find out more about what MS's plans for VoIP application development are, and there's nothing too interesting (say, the level of what I can easily do with Asterisk and C# today) for at least another year. Maybe MS has some secret surprises, especially in light of the rising VoIP hype. Even so, I hate seeing MS have to react in these scenarios instead of leading.

Where Microsoft should be looking
People tell me that VoIP isn't MS's core market. I don't know about that. Voice communications are a huge hub of daily life. Up until now, this was relatively inaccessible to Microsoft. But with Windows 2003 having great reliability, CPUs now fast enough to easily handle several T1s of voice, packet-switched voice, cheap TDM interfaces -- it's now completely accessible. (Asterisk has proven this.)

Business
On the business side, Exchange and Live Communications Server seem to want to jump right into this market. I see no reason why there shouldn't be a Microsoft PBX product. The biggest hurdle is getting a decent hardphone vendor to play along. (Although, Microsoft absolutely rocks when it comes to interface hardware, so I'm sure a hardphone from MS would win awards.) Really, why shouldn't our IM/email/voice/etc. systems be all nicely integrated? I'm quite sure MS will get into this space, even if it takes several years and an acquisition to do so.

(Heck, if MS wasn't so damn late to the scene, they could have designed a good VoIP protocol. As is, we'll probably end up having to live with that abomination named of SIP for the next two decades or so. )

Home
Seeing how MS is so desperately trying to get into things such as your living room (Xbox and Media Center), I don't see how taking over voice seems so far fetched (I heard a rumour about Xbox offering VoIP anyways). Voice fits in perfectly with the “eHome“-digital-lifestyle that MS is trying to own.

Developers
Of course, everyone knows the path to victory lies with developers. And everyone recognizes that Microsoft is king in that arena (save the few Delphi fans who don't realise Anders now works for Microsoft). Microsoft needs to have a very comprehensive and complete set of VoIP (both Video and Voice) APIs and *services* that make it easy to integrate into both Windows and web applications.

For instance, it should be trivial for someone to write a program that allows customers to call in and check their order status via phone (trivial in dev effort, upfront costs, and service setup). Connecting a user via voice, be it via PSTN or direct-VoIP, should be easy. The littlest of VB developers :) should be able to do this without spending much time at all. And then, MS should market the heck out of it so people realise this functionality is there...

Let's see how long it'll take Microsoft to get to that point.