How Google Could De-Throne AIM
by Ragnar Schierholz
Here is an interesting article, bringing up the idea that Google might start an IM service based on the open Jabber protocol.
I am a happy Trillian Pro user, which allows me to be online in MSN, AIM, ICQ, YahooIM!, Jabber and Sametime (oops, IBM Lotus Instant Messaging) from one single (and IMHO neat) user interface. Admitted, the downpart is, that it is only available for Windows so far.
I know there's quite a number of IM users around here, so what do you think? Joogle?
Comments
I wonder if there might be a conversion of technology. For my part, I use skype much more than icq or the like. Added Voice-over IP Technology and an icq-like client. Waiting for video and it would be complete... . So if google provide all this, I would agree. World domination :-)
on my mac i use proteus. which is - for me - the sleekest im-app out there. as i gmail user i must say that the idea of "joogling" kinda attracts me. 1 GB space to archive all the "important" conversation :-)
How about "Audium" for the Mac?
but I suppose most of you guys may know it already :-)
Sorry, it's actually spelled Adium and can be found here:
http://www.adiumx.com/
Is AIM really »the king«? Many people I know still stick with ICQ, at least in DE; ok, AOL owns that but you need the newest client for some kind of »interoperability« and in my experience many Windows users don't bother with updating anything ever ... On the other hand I know a growing number of people who learn about IM via Microsofts unholy MSN Messenger, simply because newer versions of Windows ship with it.
For now I'm sticking with the quite excellent Proteus which cost about 15 € but is very well done and supports a large number of IM services including Sametime and something called Gradu Gradu.
I see similar hassles with all the afore mentioned (already available) clients:
- They all merely combine multiple IM protocols in one UI (which is undoubtedly a great enhancement but still slices the IM world)
- They are mostly (if not all) only available for a single platform
The latter could be solved by a "Joogle" service, being web-based it should be platform-independent. The prior is a little tougher, but the cited article suggests that Google would have the power (development resources and market power) to a) improve the Jabber protocol to suit almost any IM requirements and to b) unite the IM users under a single roof.
I, for one, could well imagine this ro happen. Even though I must say, that I kind of dislike the idea of a purely web-based IM client. I certainly like the option to have a fixed window with my buddy list stuck to the screen edge or a minimized system tray notifying me of events. I don't see that being implemented as neatly in a web-app.
I use jabber only, the jabber server is offering gateways to icq/aim, so I don't need to be connected to icq - one just needs a valid icq account. I'd love to see more people use jabber since the jabber protocol is quiet flexible, useful and of course open.
I like the idea to have someone big in the jabber world like Google _if_ and only if Google links it's server(s) to the main jabber networks so we all don't need an extra client to communicate with google IM users. BTW, 'joggle.[com|net] seem to to be owned by someone else but Google...
@Ragnar: I think we already do have a high degree of platform independence in the IM segment. Just take a look at Adium and Proteus as well as Fire for the Mac. All of them are based on libgaim, which in fact is the core library behind the IM software Gaim. Now, Gaim is available for Windows and Linux so I don't see any major platform not being covered by this very same technology. As for the Web based IM service: I for one wouldn't want to use it. The major benefit of IM besides it being almost immediate is that it's a direct link. No easy eavesdropping or logging there.
@Stefan: Yes, Jabber is a nice product and I already implemented some servers using Jabber, mostly for internal corporate usage. That's where you can get the most out of Jabber, imho. Besides that, there's not so much difference between using a Jabber server as a central hub to my different IM accounts compared to using a piece of standalone software that accomplishes the same thing. The only thing Jabber offers in addition is multi-user conferencing and creation of conferences at will (if enabled on the server) but even that's not included in the standard package.
What's really missing here are defined gateways between the various systems in use today. I would love to be able to communicate with any IM user, no matter whether he's on Yahoo, ICQ, IRC, AIM or even a corporate IM network. Technically, there's no reason why I shouldn't be able to contact vowe@aim.vowe.net nor that I couldn't be reached using stefan@jabber.whocares.de. (Don't stick to my use of email-like adressing, maybe an approach like jabber://stefan@whocares.de would be better). The reasons why this isn't possible today are only political for every major IM service tries to lock in their respective users.
Ok, point taken, the libgaim gives a good level of platform independence, but still, each of the products based on this can, and to my knowledge do, add additional features. In a competitive market that's just fair, but one standardized set of features would be nice.
Not that the gaming is an important feature to me, but I've had friends contacting me via MSN (I use Trillian, as mentioned above) and asking why I don't join their gaming invitations. Well, I just don't see any of them, since Trillian doesn't support the feature. Sure, being based on the same libraries as the "original" MSN client, it would be an easy bite to support that, I know.
But as you said, the major point in this is, that I'd like to see a way to contact anyone in the IM world. And for that there's a need for either a gateway mechanism. And for that, a common protocol as a least common denominator would be required.
For email that already came into place with SMTP and the related protocols long before there were commercial interests in the technology, thus a lot of the political issues weren't affecting the development (proprietary clients such as Lotus Notes have learned to talk the SMTP protocol and its fellows).
For IM though that is still lacking and if I understand the cited article properly it kind of suggests that Jabber should be a base for such a least common denominator protocol. It says it doesn't cover all requirements yet, but suggests that Google has the power to advance it technically and to drive it into the market.
And to be honest, your suggested username/IM-address "jabber://" looks to me as if you're thinking much in the same direction... :-)
Well, once upon a time, when the internet was still young in the sense of common awareness of its existence, there were more email systems and protocols than just SMTP. SMTP happened to be the vehicle of choice on the Internet and with the rise of the Internet SMTP became a defacto standard - together with all its shortcomings and pitfalls we experience today.
So, I'm not really sure whether a least common denominator is really what we need. I'd say we already have exactly that. If you think about it, IM is just that: instant messaging. The term messaging of course leaves a lot of leeway. Does it only cover typed messages or is a voice message still a message in the sense of IM? What about video chat?
Clearly, the original IM concept wasn't built with those audio and video applications in mind, not to speak of instant gaming and other things. Thus, every major vendor invented - no, had to invent - a system to allow for additional services, using new and mostly standards based protocols. Be it H.xxx for video-conferencing or I-dunno-what for gaming: all of them are piggyback load on top of the original protocol, the clients being patched to know what this or that flag in a message header means and what new protocols to start.
So, basically, the standardized feature as you call it in my opinion is the typed message. To bridge that between different systems shouldn't be too hard. For the additional services like video and so on are either initiated using special flags in the standard protocol or by using a different protocol suite altogether, it shoudn't be too hard to either bridge the flags, too, or to build support for the new services into - yes, still - different clients.
I believe that the major problem of IM right now is that besides the non-existent gateways the adressing schemes in use just suck. There's no easy way for a client application to decide whether 91630387 designated a user on ICQ oder in IRC or whereever. This could be solved by a new adressing scheme which would allow me to not only specify the users name but also give some information about what protocols to use (xxx://) and where to look for further information like the users current IP address (@yyyy.zzz). Using this information and having support for the different protocols built into the client I wouldn't even really need gateway services. My client would just contact the right server, and say "Hello, I'm client XXX and I want to talk to your user YYY".
And yes, currently Jabber is the only system that allows me to do almost everything depicted above. However, it still requires me to have a valid account on the central server the receiver of my messages is using (if that said server isn't connected to the Jabber network).
Common misconception: "Jabber is an IM product"
Not really. It's an open protocol based on streaming XML in a client-server model. Intrinsic functionalities are addressing, routing, location-awareness and presence.
Presence is built as a subscription model, so every online jabber client instance can freely decide who/what is able to "see" its status.
I see Jabber rather as a middleware which eventually supports the average instant messaging application right out-of-the-box. IM is not really hard to implement, if you think about it.
The design rationale behind Jabber is clearly visible when you look through the server-provided services (user directory, chat rooms, transports that proxy Jabber to AIM or Yahoo IM, and other things "that plug"). This is where its real powers are hidden :)
And, Jabber is not a "common least denominator" approach at all. Due to its extensibility through XML and name spaces it's rather the opposite. We are talking "information transport infrastructure" here.
Things like libgaim or Trillian provide instant relief[TM] to thousands of troubled IM users, but they simply can't connect them without the barriers of their respective IM service provider. Gosh.
That's where I see Google's strong point: acting as a proxy (in terms of protocol/application development and actually using the protocol) that simply operates by foot print and market share.
If Google implement their gateway services (misread: getaway services) in an open manner, reflecting native Jabber protocol onto a distributed server platform, it will be a win/win situation giving all proprietary IM users a choice they never had before:
The user community wins by IM interoperability and Google wins by mining and scoring a million personal messages each minute - but that's what PGP support in many of todays Jabber clients is for ;-)
One last note: Ever thought about what will be the next step(s) when they got their monster in place?
Corrections? Questions? Remarks? Flames? xmpp:byteborg@datasink.webmonster.de?message or aim:goim?screenname=ByteBorg
If you happen to be in the Darmstadt area: Come over for a beer and a nice chat about information infrastructure, in real-life, which might still be the richest experience, at least for now ;-)
No, not exactly a misconception... I always referred to Jabber as a protocol. Of course, there are products implementing the protocol which call themselves jabber clients which leads to a slight ambiguity.
But aren't the other IM "products" basically protocols as well? Just, they aren't as open as Jabber is. The authors of Trillian, gaim and the alike basically reengineer the protocols which clients like AIM, ICQ, MSN Messenger (speaking of the products now), etc. are using to communicate with each other. After all, once the message comes across the network, it's only network packages. What application it was sent from, who cares?
And as far as "least common denominator", I still stick to that label. Especially the extensibility of Jabber makes it a least common denominator in the sense that it says "here's the basics we think we can all agree on and here's how you can add your stuff that YOU think YOU need as well". Call it middleware (but wouldn't that be more a product than a protocol as well?) or whatever you want, it's a common base for others to build up on top of.
Um mal wieder den Kontakt mit der breiten Nutzerbasis herzustellen: "Die wichtigsten Features des Instant Messaging für User sind Foto- und Filesharing und die Benützerfreundlichkeit."
Gefunden hier: http://www.zdnet.de/news/tkomm/0,39023151,39125331,00.htm
So viel zu unseren Theorien über globales IM und wie man was machen könnte ...
