Tuesday, August 23, 2005

Google Talk: Whats the big deal?

Deepak asked a very good question when I posted about Google Talk earlier today:
But what is there to be extremely happy about?! So far, all we've seen is a Jabber server! Okay, maybe a Jabber server which can handle more load than any other out there.. but still.. nothing to get all worked up about!
Why am I so excited about GoogleTalk? If you see my original post, you can see that I have no dearth of instant messaging services to connect to, and thanks to software like Gaim, AdiumX and so on I don't even need to have different clients. So, why get excited about an offering from Google, especially when it represents absolutely nothing new? (In case you didn't know, Jabber - the server on which Google Talk is based - has been around for years and anyone in the world can download and host a server). The answer is simple. Two sentences, four words.

Open source. Mass scale.

Whats the most advanced feature of the latest instant messaging services that you've subscribed to? Voice chat? Webcam? Online games? Great stuff. Is it perfectly common to use these things? Unfortunately, no.

Why not? Because the barriers are too great. You use Windows (95% of my blog readers do), I use a Mac or maybe a Linux box. Even if I do use Windows, maybe I'm using Windows Messenger 4.0 and you're using MSN Messenger v6.0. Do you use a Mac? Lets do a videoconference... but wait... Apple decrees that iChat can only accept videoconference through the incredibly beautiful, but impossibly expensive, iSight webcam which costs more than Mac OS X Tiger. There are hacks to get other webcams to work, but they're painful. So the bottom line is - unless you're using the same software client on the same platform on the same messaging service - your interaction on an instant messenger is limited to one thing - text. Exceptions exist, notably Skype, which has done an amazing job of simultaneously enabling near-perfect telephony and voice chat on all three major (well the one major and two not so minor) platforms today.

Why am I excited that Google's offering is "just" setting up a Jabber server? Because its open. Go see the bug list at Adium or Gaim pages. Its filled with stuff like "When will Microsoft's WINK feature be available on Gaim?". "Why doesn't the :R emoticon work on AIM?". Or even, "My Yahoo account doesn't work anymore!" Because these services are continuously changing their data formats, standards and protocols and they aren't obliged to reveal them to anybody - leaving tons of talented open source developers doing reverse engineering round the clock to figure things out. With Google's jabber service, this won't be a problem. Even though Google has (once again) deplorably left out Mac OS X and Linux in the lurch for voice chat, clients on both can connect to the instant messaging service and if (as I hope) the voice-streaming is using a standard protocol, we might soon have some clients that can actually handle them.

Another part I like is that Google is not trying to monopolize the IM market. If I am reading the Jabber spec right, as a jabber user with a gmail ID I can as easily talk to someone using a Jabber account on some other server. Ever asked someone, "Are you on MSN?" and hear "Oops... I only use AOL"? Spread the open IM standard far enough, and these things will be past. I am thankful its not Google's intention (as some predicted) to have a client where it would do something like generate adwords based on IM conversations, or have a client like MSN Messenger or Yahoo with a zillion tabs in the software interface solving your every problem from astrology predictions to dating troubles.

Development of IM applications, extensions and clients can also take place much faster given that there is a community that is behind the Jabber software and standard. Instead of begging and waiting for a feature - go and code it yourself, if you want it badly enough. And if Google's history has anything to show, they don't intend to just have a messenger in place and be done with it. I'm sure their truckloads of PhDs and summer interns are going to be involved in taking this initiative further.

Of course, all this existed before - Jabber has been all of this and existed for a long, long time. Why get worked up over it? Because, as someone on Slashdot (where else) insightfully mentioned: "If Google ties their messenger to GMail accounts, it will have gone from zero to a zillion accounts in one day". Bingo. Heck, people call me a geek, and I don't have a Jabber account with any public server (I had one when we were playing around with implementing Jabber on PDAs in SMA). When I enthusiastically emailed all my friends that Google Talk was live on any Jabber client, the first responses were, inevitably, "Jabber? Whats that?" So here you go - Google Talk is a much more compelling reason for the average man (and I'm included in that category) to go Jabber. Conversely, its that much more incentive for developers to start taking an interest in Jabber - and thereby increasing the chances of vastly improving the IM experience.