1,2,3,4 -- I Declare IM War
Unless you're living under a rock, chances are you've heard that Google announced their Google Talk service recently, which ties into ones Gmail account. Once they acquired a company with an instant messaging solution, speculation was flying about them releasing something, but no one really knew what form it would take.
They've branched out into native Windows software recently, with apps like Google Desktop Search or the Google Toolbar, while at the same time focusing on more services via your browser, like Google Maps. What was announced earlier was a native Windows client that communicates via the Jabber protocol, but we'll only touch upon that in a cursory way.
Rather, we'll be focusing on what wasn't announced, and how Google's competition in the space -- Microsoft, AOL and Yahoo -- will be reacting.
Once you start developing a network of sources, or rather a network of sources starts developing around you, you learn pretty quickly how rumors and tips work. There are three base classes of rumors: Those that are real, those someone wants to be real, and those that someone wants others to think are real.
Something that has to be kept in mind is that even when a rumor is 'real', what it was based on can change. Decisions inside a company can change, and you have to account for rumors being embellished or altered by the time they reach your ear. I mention this because I don't really talk about rumors much unless I'm hearing similar things from multiple sources, where the specifics might change but the core facts line up.
On a scale of 1 to 10, I'd give the solidity of what I'm about to tell you a 9, but things can change, especially once the info hits the web. My sources are anonymous, and will stay that way even if someone tries to pull an Apple, but they're solid. I think what we're going to talk about is important, and in everyone's best interests to be aware of because it's going to directly affect you.
When it comes to instant messaging, there are basically three players...
- Microsoft, with their MSN Messenger service.
- Yahoo, with Yahoo Messenger
- AOL, with AOL Instant Messenger (AIM) and ICQ. ICQ is old school, but still has a substantial following -- especially outside of North America.
There are others, but their share amounts to rounding errors, and I'll just refer to them as the Triumvirate. Chances are you may be aware of this, but what you may not be aware of is just how chummy all three of these guys are. This isn't necessarily a bad thing, but for better or worse, instant messaging has become a 'comfortable' space. It's growing quickly, but no one sees any real way to make money on it, and no one is pushing too hard against the others.
In fact, at various times, some of the companies have looked at letting one of the others take over their IM service. It'd still be theirs, and they'd still have their client, but as an example the server side may be routed through an MSN service, instead of Yahoo bulking up their servers for something which is necessary but is rapidly becoming a commodity -- if it didn't actually start out as one. Much of this ties into MSN and Yahoo and AOL all trying to transition to SOA's (service-oriented architectures) where what's being provided is more important than how.
Anywho, as time has gone on, these transitions have been underway, and these types of talks have been underway, but while each wants users for their service it's only a loosely competitive space. No one is kicking over the table to get users to switch, they're simply trying to tie the users they have into their other services while sucking in newcomers as they come along. They all have these carefully laid plans for the future, and how they're going to tie x into y to benefit z.
All of this stuff is starting to weave together, or at least they're trying to. I.E., your Apple ID is tied to your forum account, which is tied to your iTMS account which is tied to your Apple Store account which is tied to... Everyone is trying to weave their services together to dig in.
This competitive situation isn't very abnormal, as it takes a large amount of infrastructure -- and with it cash -- to run these types of services. With only a few players in the game, who are all growing comfortably, no one is seriously wanting to rock the boat and change the game right now. I wouldn't hold it against you if the IM situation seems eerily similar to the state of free email services before Google entered the game.
Word had spread pretty early to the Triumvirate that Google had something in the works for instant messaging. They get their own share of rumors, but while this was a concern it wasn't really a surprise, and no one really knew what they had planned.
Possibilities were mapped out, and surprises were allotted for, but without really knowing what Google's secret project would look like there was little to do except continue with business as usual.
However, not long after it was announced on the web that Google admitted to having acquired an IM company -- without saying what their plans were -- word got out to the Triumvirate that not only was Google planning a native chat client for Windows, they were planning on having a multi-protocol chat client for Windows.
They completely wigged out.
The network effect says that some devices are more valuable as more people have them. If you buy a hammer, it doesn't become more valuable when your neighbor also buys one, excepting that he's less likely to want to borrow yours. However, when your neighbor buys a telephone, your telephone is more valuable because there's one more person it can reach.
The network effect can work in different ways, or rather at different levels. For example, you probably have an email address. As someone else gets their first email address, yours is more valuable because there's another person you can send to and vice versa. With IM networks -- as they currently exist -- it works a little differently. If you have an AIM screen name, the value of it increases only when someone else gets an AIM screen name. If they are using MSN or Yahoo or something weird, they can't talk to you.
The clear difference is that (for the most part) email is a standard, and the clients all interoperate. It doesn't matter what platform or client you use, but with instant messaging it very much does.
Multi-protocol IM clients allow you to sign onto multiple services via one application, instead of several. They exist for every platform that I'm aware of, and while you generally have to make some concessions in forgoing some features of the "native" clients, they bring with them their own benefits.
Since Google Talk currently just talks to itself and other Jabber clients, you'll be forgiven for being confused. It's unknown whether they planned to initially release it as a multi-protocol application and have scaled it back, or whether it was always planned to roll it in later, but they are planning to do it.
The problem for Google is that while they have massive mindshare, and can get a bunch of geeks to sign up and play, most already have a healthy IM rolodex -- all on services that aren't Google Talk.
Initially, this isn't a huge problem for Google. They're going to have to work out kinks, and their mindshare is such -- and how they release software is such -- that a lot of people are going to want to get it and play around. These are primarily going to be geeks and early adopters and those that want to be both, but at some point Google is going to need the normal users.
Normal users often don't run more than one chat client -- they generally gravitate towards one service because it's where all their friends happen to be. They may well have names on more than one service, but they're generally only regularly signing into one. They can be talked into signing up for Google Talk by a friend, and they'll have that friend on the service and perhaps a few more, but they're still left with a ton of people who are on the one of the Triumvirate's services.
This is where growth will sputter, because the other services have a massive amount of inertia on their side -- the network effect -- and Google Talk is only as valuable as the amount of people you're able to talk to.
As mentioned, multi-protocol chat apps exist on just about every platform. Mac OS X has three that I'm aware of, and Windows has its own, Trillian being the most popular. These are primarily just nuisances to the major networks, because:
- All of them are just hitching a ride on the network, but just to chat, they aren't trying to tie it into other services, and hence aren't serious competition.
- None of them have the mindshare, nor resources, to be a credible threat. They're primarily just viewed as small parasites -- or remoras, depending upon your view.
However, Google is both of the above, and then some. Every person they tie into their Google Talk service is tied into their Gmail service and whatever else they have planned for the future, and they have a hefty pot of gold from their IPO and, you know, actual revenue is doing well by Google. However, the killer is mindshare.
Not everything Google has touched has turned to gold, but for the majority of their initiatives their entrance into a market has been staggering for those already making a go of it. Gmail is just now coming out of beta, although everyone I know already has an account and its completely changed the landscape of service-based email accounts.
They're going to make a splash with Google Talk, and people are going to want to download it and play, but eventually their mindshare is going to crash on the wall of momentum the existing services have built for themselves. People are going to have to keep their other chat clients to talk to their friends for now, and that gives the services breathing room to work out how to respond. However, by having a multi-protocol app, Google is wiping out their momentum in one swoop, and you can understand why they're freaking out.
There'll still be issues with things like A/V and file transfer, which is good for the Triumvirate, but they'll often work in favor of someone talking to a Google Talk user just picking up a Google Talk account instead. In these situations, Google has zero to lose and everything to gain.
Within the last few weeks, there appears to have been a meeting between MSN, Yahoo and AOL. They'd all been talking amongst themselves -- and sparsely with each other -- about how to respond to Google, but were still trying to make up their minds.
This meeting made it official: It's three against one, indirectly.
Specifically, sometime in the near future, all three services are going to be releasing new versions of their software and "Sunsetting" all older versions of their Windows clients. From what I can tell, Microsoft is spearheading the initiative, but the other two are going to follow.
Generally, each service has its own protocols and they do well to keep older versions in mind when they release a new client, because they want as many people on the network as possible. However, that's going to be going away -- you'll use the new client, or you won't chat on their networks.
This is far from ideal, and is going to cause them a ton of headaches, but they feel they have little choice as they can't allow Google to augment their massive incoming mindshare with their service's own momentum.
The fallout from this is going to have a dramatic impact, at least initially, and I'm not just talking about tens of millions of people having to upgrade their chat clients to the most recent version. At least one of the services isn't planning on filtering anything but Windows traffic, which means Gaim and AdiumX and Fire and Proteus and such will be ok for that service.
However, while I don't know about alternative platforms, I do know anything on Windows is in serious trouble, including Gaim and Trillian.
Unless Cerulean Studios (authors of Trillian) can negotiate access with the individual services, their client will be gone. Each service will probably handle this in a different way, and while I've tasted breadcrumbs like "SSL" left about, it's unclear how each one will do it and will most likely be a combination of different techniques.
There are two sides to this question:
- Can they do that technically?
Oh yes, they can most certainly make third party access all but impossible, providing they take the drastic action of sunsetting all earlier versions. At the very least, they can make it so difficult for third parties to keep up -- and reverse engineer what they're doing -- that its such a nightmare few if any will bother.
Historically they haven't had much luck in keeping others off their networks, but they haven't really tried that hard, either. Backwards compatibility was always paramount, so they only had a small technical window to operate in, but by mandating new clients and filtering for anything coming in via Microsoft Windows it's Game On. Certificates, SSL, the whole bit.
- Can they do that legally?
The only one of the services I was curious about was AOL, as back when they merged with Time Warner there was concern a monster might be being created when it came to the internet and distribution. The FCC gave AOL a choice: Allow third parties on the network, or forgo integrating A/V into your client.
AOL chose the former, and created a separate way for clients to access their IM network. It doesn't contain all of the features that the official client has access to, so the majority of third parties eschew it in favor of reverse engineering the protocols AOL itself uses. Theoretically, this would allow someone like Google to come along and tap into it, and have a legal right to demand it.
However, for some time AOL has been arguing that they shouldn't be bound by this agreement any longer, claiming the market is still open enough to allow the entry of other players just fine, and that agreements like the one they have with Apple for their .Mac service are a better way to go. It appears the FCC has decided AOL has a serious point, and this won't be a factor.
Now, if you're a geek, your mind is automatically going to go, "Well, they can try, but it'll just get reverse-engineered eventually and..." but that's a trap. You'd be right -- any protection scheme has a habit of getting broken eventually -- but the key isn't about making it foolproof so much as so difficult no one is going to try to base anything serious on it.
This is going to be way behind reverse engineering the protocols, and they're well aware of everything going over the networks and equally aware of how to shut it down -- it just requires retiring all older clients.
Even setting aside technical issues, sunsetting all the previous clients -- and with it eschewing backwards compatibility -- means protections can be added that could well bring in clauses of the Digital Millennium Copyright Act (DMCA) if anyone tried. It wouldn't stop a hacker or cracker from giving it a go, but it would be a serious hurdle for someone trying to reverse engineer access.
One of the amusing things about this drastic action is that it's going probably going to increase Google's mindshare and momentum in the short term. If Joe Public hears that something Google is putting out is immediately putting all the others on defensive -- and in a drastic way -- it elevates Google, which has zero share, onto a mental level where it's as big as the established players.
People are more likely to be curious, and go check their new product out, but it's also the kind of mindshare boost you can't buy. From what I can tell the services are entirely aware of this, but the alternative is just so much scarier for them. Damned if you do and damned if you don't, but they're figuring that this way they might be damned a little less.
The other side to the backlash is the idea of purposefully shutting out people who want to use a different client, whether it be Google Talk or Trillian or something else. Since I use a third party client myself, I get it, however they do have a decent argument for Joe Public here: It's their network and resources, and it costs them money to run it and support it.
In the short term, there are going to be a ton of headaches going around because of this. It's going to be rough, however I don't believe Google is changing events so much as accelerating them, much as they did when Gmail entered the market. At the time there were competing services, but no one was really going balls-to-the-wall against the others.
Practically overnight, the world (Excluding users of Apple's .Mac service) went from clearing out old messages from their accounts so new ones wouldn't bounce to storage being a non-issue. Basically, we all would have had 1 gigabyte of email space eventually, as over time the services would have had to compete more heavily for users, but Google pushed forward the timetable.
Over time multi-protocol is probably going to be the future, as being tied to one service to communicate is good for the company but bad for the user. If you could only email people on MSN from your MSN account, or Yahoo accounts from your Yahoo email, you'd think it was stupid, yet here we are with instant messaging for the foreseeable future. It's sad in a way, and things like Jabber could theoretically make it a non-issue, but until Google Talk, Jabber was gaining little traction in the grand scheme of things.
Circumstances would probably lead to the Triumvirate becoming less chummy and competing harder against each other eventually, which is Good Thing™ for the user. We'd get there eventually, but while Google's entry into the IM market is initially bringing them closer together as they circle the wagons -- which is going to cause headaches -- it will hopefully speed up the timeline for serious competition between all the parties.
Posted by drunkenbatmanAugust 29, 2005, at 04:13 AM
Comments (43)
Posted by: David Appleyard at August 29, 2005 05:05 AM
Great article. I'm looking forward to seeing how this pans out.
Posted by: mikeash at August 29, 2005 05:08 AM
The proposed strategy by the triumvirate seems like it could very well backfire on them.
As things stand, Google Talk is cool but I don't see it going anywhere significant because of network effects, as you discuss.
But by breaking every existing client and forcing people to upgrade, the status quo changes. Every upgrade (and it's going to be every user upgrading) is a chance for someone to download the latest version of Google Talk instead of the latest version of Yahaolmsn. This is fairly unlikely for nontechnical users, but then technical users have a weight that goes beyond their numbers, since we're the ones giving advice to all the nontechnical people.
From a technical standpoint, locking out third parties forever is pretty simple. Force the client to provide a hash from a server-specified byte range of the executable, and you're done. Third-party software would either have to distribute a copy the official client (very much illegal), or require the user to download his own copy and piggyback on it (extremely inconvenient, possibly illegal). The only reason this hasn't already been done is because the IM providers either believe that breaking their older clients is worth the pain of having third-party clients, or they believe that third-party clients are actually a net benefit.
One thing I don't understand is how a provider could only lock out third-party Windows clients while still allowing them elsewhere. The server doesn't know what OS you're running, and any non-stupid third-party client is going to lie if asked.
With network effects, people using third-party clients with no advertising or other tie-ins can still be a net gain. Every extra person on your network increases the value of the network. As long as most users are still using the official network, there can be a net gain for the provider.
Looking at it that way, I see this benefit as being another big problem for this proposed move. By locking out third party clients, you suddenly force all of the people using them to choose between the official Yahaolmsn client or switching to Google Talk. And people using third-party clients are typically going to be savvier, and thus more flexible and carrying extra weight with their friends about which network to use. Even if they can't get their friends to switch, they'll be forced to run multiple clients anyway to stay on all of their services, so why not toss in a Google Talk client, or better yet a third-party Jabber client with the nice features they want, and use it to talk to anybody they can convince to switch?
In any case, it'll be interesting to see if this pans out as expected. I personally use iChat for AIM at the exclusion of almost everything else, and the few people I talk to with Yahoo Massenger are going to get very strong recommendations to switch if I can't continue using my third-party client.
Wow, that was a long post! I must be procrastinating on something more important....
Posted by: Alex Reid at August 29, 2005 05:19 AM
I'm not sure how I'd feel about this if it all goes down. On the one hand, I see that yes, the companies own their networks and aren't doing anything wrong *legally*, it'll be a mighty pain for us Mac users whose friends are all on MSN, AIM or Yahoo!—whichever ones decide to filter out all other clients on the Mac side, too. The official clients all suck somewhat (although with MSN 5, Microsoft have thje nicest, I think). Oh well. I don't *need* IM…
Posted by: Iljitsch van Beijnum at August 29, 2005 05:20 AM
It's interesting to note that on the one hand, Google is talking about a "federation" of interconnected IM networks (the term alone will score them major geek points). See /talk/developer.html on the site of a well-known search engine that I'm not allowed to link to...
On the other hand, they're not being a good Jabber citizen: you can't talk to non-Google Jabber users, but you can "invite" Jabber users that try to contact you to the Google service. (I think an important reason for the walled garden approach is to keep out spam, though.)
I'm looking forward to see how Apple is going to react to this. Apple's iChat already works extremely well with Google Talk (except that it warns you that the login is unsafe even though it happens over SSL). Since Google is in the service business and Apple mostly in the client software business, it makes sense for the two to work together. Too bad there isn't any voice chat interoperability between the Google Talk Windows client and iChat at this point. (You can voice and video chat with other iChat users on Gtalk, though, just like on AIM or regular Jabber.)
Posted by: Anonymous Coward at August 29, 2005 05:38 AM
and will stay that way even if someone tries to pull an Apple
Interesting post if true, but was that really necessary?
Posted by: craftyc at August 29, 2005 06:22 AM
I don't think that this sunsetting of older clients will be as big a deal as you say it is. All of my friends currently use MSN. they are more than capable of updating their clients. It just takes a few clicks these days. Also none of them use different services. They are all content with MSN. As a result I am also content with using MSN. I don't have any friends on other networks. I know people with Yahoo or AOL addresses along side a Hotmail account, but they still primarily use Hotmail.
Google entering the market will mean nothing for them, or me. Microsoft, Yahoo and AOL will secure their services. But most people will still be good to go. They use one network. The network that most of their friends use. All they will see is an updated client with a new feature or two.
Of course, I could be totally wrong but I am speaking from my experience here in the UK. Maybe elsewhere the situation is different and people use different services. I don't know. But I know that in terms of the people I know, it is a non issue.
Me
Posted by: Colm at August 29, 2005 07:37 AM
I can only wonder if this is somehow related to old TOC (Open AIM) protocol plugins breaking recently.
Posted by: Nugget at August 29, 2005 08:16 AM
One one big unknown in my mind is what Google's ultimate stance on the other half of Jabber/XMPP will be. They've embraced an open protocol for client access but they haven't even built the other, more critical, half of the Jabber spec which would enable server-to-server message passing. Their public statements on the subject are cagey and don't withstand much scrutiny.
I disagree that Google's weird "Federation" concept wins them geek favor. I think that's just about the worst approach they could take and I'd almost rather seem them stay closed if that's what they prefer. "Federation" doesn't help us at all.
http://nugget.livejournal.com/97081.html contains my thoughts on the matter, which serves as a decent primer and companion piece to your entry here.
Posted by: w at August 29, 2005 10:26 AM
"...but the key isn't about making it full-proof..."
You mean "foolproof," I believe.
Posted by: Jan at August 29, 2005 10:35 AM
Two points:
1:
I couldn't care less about the 'established' networks. All people I want to IM, connect to a Jabber server, and I run my own. I have a backup ID on another server, so in a pinch I can switch easily. Call it decentralisation or freedom if you like.
2:
Most people consider IM to be private. I don't want anyone to poke around in my conversations and certainly not to deliver 'added value' aka ads. (For the same reasons, I do not have a email account at hotmail, gmail or any of the webmail providers.)
I know most internet users don't seem to care about these points, but I do.
Posted by: Peter da Silva at August 29, 2005 12:03 PM
Personally, the only reason I use AOL's service is that I can use it from Adium. The people who I chat with only use AOL's service because they can use it from iChat or Adium or whatever other application they use, or they just run multiple clients. If AOL doesn't work with Adium or iChat or whatever, I'll just use Jabber (Gmail and our local Jabber server) and we'll adjust.
So I just hope that AOL and MSN and Yahoo work real hard on locking out other clients, because I'd really love to have an excuse to tell the people I'm chatting with "I'm using one client, it's a multiprotocol client, I can use any service that's open to multiprotocol clients, but AOL isn't one any more, get an account on Google or use our Jabber server".
Posted by: allgood2 at August 29, 2005 12:15 PM
Another thought provoking article. I do love reading your site.
While I was reading your article, three comments made my mind instantly start connecting the links between Google and Apple (whether indirectly or directly established):
1) Apple IDs tie into iTMS, Apple Store, .Mac, Apple Support, iChatAV and other services;
2) Sunsetting older Windows clients and the immediate and dramatic effect this would have on users of each system (AOL, Yahoo, MSN), but the even wider or cumulative effects this will have on Microsoft Windows users; and
3) How one provider (possible others) will only be filtering Windows traffic, allowing Mac/Apple traffic access relatively hinderance free
It seems to me, that there are a growing number of battles for dominance and or ownership of markets that perhaps shouldn't be owned. But what I've slowly been noticing is that Apple is a growing influence in these battles, often just by its very nature to stay slightly on the outside of them.
For example, when you discuss network or service convergence/tie in, Apple has been doing this strongly and effectively since the release of Mac OS X. Before OS X, I remember having different accounts for support and the Apple Store, vs my Apple Developer membership. But since OS X, a single account allows me access to iTunes Music Store (iTMS), my .Mac services, the Apple Store, and iChat.
Yes, I'm most tied to my iTMS account, but Apple somehow managed to create a single user login for multiple services without announcements or even a lot of hardship. I remember, mostly in hindsight, once attempting to login to Apple Support, and being notified that my account username didn't work. I just assumed I had entered the wrong username, and went down the mental list of usernames associated with Apple service ( my .Mac account was my second guess), and it worked with no problems. It wasn't until later that I realized, that it (.Mac account) worked on ALL Apple services. Sneaky, subtle, and very well done (at least for people who had .Mac accounts).
So while Windows user are often actively resistant to a single sign-on as representative by .Net, both Apple, Google, and even Yahoo have made progress in this arena.
Next we have the protocol wars, since tons of vendors are competing for the Windows market; Apple has been the 800lbs gorilla for the Mac market. This has provided benefits and headaches for Apple developers. But one of the benefits that many have overlooked is NOT having to wade through a see of protocols to hack. Typically, many Mac developers can start with Apple's software, protocols, and or standards. Hook into Address Book, iCal or iChat if you can make your software work with these applications, you have access to XML, LDAP, iCal, Jabber, etc.
Apple is the outside runner, against a pack jockeying for position against each other. Look at the iPod, iTunes, and the iTunes Music Store. You have 60 odd some portal MP3 players that are Windows Media compatible trying to grab hold of 20% of the market; because they bet on Microsoft's DRM. But Apple bet that users didn't care a damn about DRM if they could do mostly everything they wanted with the file.
Apple choose Jabber, and selected AOL as an IM partner (I'm assuming for tons of reasons, but I'm betting the lure of building relations with AOL Music Channel & it's AOL Sessions, Disney, etc. contributed to the selection somewhat). Now Google has also chosen Jabber, and while you can argue whether or not the current lack of S2S support makes this selection more or less groundbreaking, it is yet another tie in for Apple into Google-based services; and Google-based services into Apple.
Sure we don't have Google Desktop, Picasa, or Google Talk, but we have Sherlock, Spotlight, iPhoto, and iChat which provide services in the same arenas and that hook into Google in various ways. Now we have GMail, and better yet a Google identity, all we need is a single method to tie our Apple identity to our Google identity (at our own initiative, of course); and the mounting methods of avoiding Windows while participating in Windows-based community keeps growing and growing.
As someone already mention, your Google ID (for GMail and Google Talk) can be used in iChat AV. I can't state how well, because I haven't played with sending anyone using Google Talk messages (yet). So now, while on the face of things Google has yet again snubbed the Apple community, but in actuality has provided fairly easy access to its systems and services (see GoogleMaps, and GMail initial offerings...no Mac support, but Safari users had fairly immediate access with some limitations on specific tasks, until subsequent updates).
Anyway, this is long, I should have done a trackback, but I try not to talk about technology too much on my blog. I see the services being offered by Apple and Google as joint alternatives to the current Microsoft' hegemony. Combine this with the beating Microsoft is taking from virus, hacking, and spyware markers and IT's reluctance to upgrade; I think we have not a direct changing of the guard but a more indirect change of what's being guarded. Services are the new desktop and Google is offering them, and Apple is offering easy ways to tie into them without leaving the comfort of your desktop.
Posted by: Pascale Soleil at August 29, 2005 12:20 PM
I think you're missing a coupla words here:
"but Jabber and until Google Talk was gaining little traction in the grand scheme of things."
I know, it was late. I was up too.
Posted by: Nabil at August 29, 2005 12:35 PM
Interesting post. GTalk has a lot of potential to be worthwhile and cool, but at the moment it's just another app to have open, because the features they're offering at the moment aren't enough to entice people from their current IM choices. If they actually manage to fulfill goals such as multi-service, multi-protocol, multi-platform, server-to-server interoperability, et cetera, then they've really got something.
I'm a Mac user, so the actual application is kind of moot for me... I just have my gmail account set up through iChat's Jabber support, and mostly I ignore it because I don't have anyone on it yet.
Posted by: RexBinary at August 29, 2005 12:38 PM
Jabber 'should' be the future of IM, it's secure and open. I just don't know how long it will take people to wake up on this one. I think UNIX/Linux users and Apple see it, but who knows if/when the M$ drones will see it.
Google though, they are using Jabber, but it's a closed server. It seems to me they want to look 'open' by letting you use any client and let you use an 'open' protocol, as long as your on their 'closed' server. Which I think is worse then the other guys.
The other guys started out closed and are fighting to stay closed. Google is taking an open protocol and bastardizing it using a closed server.
I have a Jabber.org account that uses Jabber protocol, yet I cant contact people on the Google Jabber server, nor can they contact me.
Posted by: mindflayer at August 29, 2005 12:39 PM
Posted by: Colm at August 29, 2005 07:37 AM I can only wonder if this is somehow related to old TOC (Open AIM) protocol plugins breaking recently.
That was probably to reduce SPIM.
Posted by: pat at August 29, 2005 01:01 PM
One way Google could implement a multi-protocol client is to use a Jabber gateway. This way everything (possibly even the gmail users aol/msn/yahoo accounts) could be handled server side.
You tell gtalk that you want to add a new msn contact, it shows up as contact@msn.gmail.com and you are good to go. This way if Google has to play "chase the ball" with IM protocols they only have to do it on their servers. This would cost them a buttload in bandwidth, but would make the user experience so much better.
Now, the DMCA stuff, I don't know how to get around other than using their billions to get that fucktarded law struck down, but the technical side would be relatively easy to do better than anyone else has done, just more expensive.
Posted by: Emil at August 29, 2005 03:58 PM
I think that it will not be possible, legally or technically, to lock out 3rd party clients from yahoo/aol/msn.
Legally:
* Reverse engineering is permissible even in America for the purposes of interopability. Courts have now ruled that the DMCA does not grant more "protections" than already provided for under copyright act. (See StorageTek etc. Outside the US this is a non-issue anyway, for the moment.)
* Hashes on copyrighted executable (See below.)
Technically:
* Actively trying to lock-out GAIM etc users will add a massive incentive to hackers to figure out how to thwart this. So far the hackers have won every time: DVDs, itunes DRM etc...
* SSL. Not a problem, just makes sniffing the packets harder. Makes things more fun for the hackers.
* Hashes of executable: Just store a full copy of the executable on a server somewhere and have a protocol for remote clients to obtain a hash of it. (DMCA allows for interopability, using hashes is fair use.)
Anyway good article even if I *totally disagree* that it is technically feasible to lock-out 3rd party clients.
Emil
Posted by: Samuel Sidler at August 29, 2005 04:26 PM
Another form of backlash which you didn't mention -- though it may have been mentioned in the comments (I didn't read them all) -- is that of those using third party clients that will be locked out and where they will go.
As a user, if MSN/AIM/Y!M kick me and several of my friends off their network (by disallowing Adium or another Mac client), there's almost no doubt that we'd switch. This may require occasional ventures back to the other clients for video conference or associating with another user, but the vast majority of us would switch so that these ventures would be less frequent.
And... as seems to be mentioned whenever someone locks out a good majority of Mac users; lose the Mac and you lose the war. (Though no one's really sure how true that is.)
Posted by: Jon R. at August 29, 2005 04:28 PM
no one really know what they had planned."knew", I'm guessing.
typo:
surprised were allotted for
word missing from here:
someone Google trying
Personally, I used to use Trillian, almost exclusively to talk to friends on ICQ, until a friend of mine got a Mac. Somehow there was constant incompatibility between the two third-party clients when typing in Hebrew (either I saw only gibberish or she did), so I switched to the official client.
You're right about ICQ's substantial international following - none of my friends have ever even heard of AIM, a few use MSN. None that I know of use Yahoo's.
I think AOL might have a serious problem "sunsetting" current ICQ clients. There was such a public disturbance with their latest client (everyone hated it, performance issues, etc.) that they still provide a link to the previous one on the download page.
Posted by: Colin Barrett at August 29, 2005 05:10 PM
As an IM developer, the idea of being completely locked out is a very, very scary idea.
For once, I hope you are dead wrong, DB.
Posted by: Mindflayer at August 29, 2005 06:01 PM
Legally:* Reverse engineering is permissible even in America for the purposes of interopability. Courts have now ruled that the DMCA does not grant more "protections" than already provided for under copyright act. (See StorageTek etc. Outside the US this is a non-issue anyway, for the moment.)
It may be permissible to attempt to make a client interoperable. It is also permissible to allow only certain clients to attach. The backend of any instant messaging system can be quite expensive. It's not the code that is the issue, but riding the system.
Posted by: Emil at August 29, 2005 06:18 PM
Reply to Mindflayer:
ah yes, perhaps i was looking at it from the wrong angle.
But surely if the 3rd party program is used by someone who has permission already, then that is ok. Companies put stuff in EULA's all the time that is not enforceable, (e.g. that you can't sell it, but of course you can (usually) because of the First Sale Doctrine etc.) What I mean is that even though there may be a contact, is the court going to enforce it? e.g. could websites require certain browsers, or email servers say that only specifc programs can connect to them?
My point is that contracts of use always contain bulls** that a court would not uphold. And what a court would think about a 3rd party IM client is not at all clear.
Emil
Posted by: Christopher Forsythe at August 29, 2005 08:20 PM
I've been looking for an excuse to push jabber on everyone, this would be it.
Posted by: Ben at August 29, 2005 09:02 PM
In The Future, all chat clients will use something OpenId-like for identification and FOAF-like for buddy lists.
Until that day, due to network effects yada yada, all the chat services are just poisoning the pool. Chat is less useful for everyone. It's offensive. Unless Google goes all the way and allows users authenticated by other services based on a published standard, it hardly matters. Just like I can right-click and "Also show Gmail contacts" I want to be able to right click and "Also show LiveJournal friends". Why not?
Posted by: Baylink at August 29, 2005 10:37 PM
I concur with Mikeash's comments early on: if they sunset all of *their own* old clients, what they're doing is effectively *requiring* all their current subs to re-evaluate what they do and how and where they do it.
Which is exactly what Microsoft's been doing with operating systems. If you're not familiar with the uptake numbers on 2003 and XP vs 2000 for 98 and NT upgrades, ask someone who is. Or look at MSFT's current quote.
And they think the sea-change which Longhorn/Vista is supposed to be will play in Peoria. Heh. I may have been wrong about 2005, but MS *is* going down.
Posted by: Skatch at August 30, 2005 04:54 AM
No one's answered this question posed by mikeash early on, which is the first thing that popped into my head with regards to 3rd-party clients:
One thing I don't understand is how a provider could only lock out third-party Windows clients while still allowing them elsewhere. The server doesn't know what OS you're running, and any non-stupid third-party client is going to lie if asked.
I also don't understand. If it is indeed the case that they are only going to be screening windows 3rd-party clients, then these clients will just masquarade as Mac clients. Is there something I'm missing?
Posted by: drunkenbatman at August 30, 2005 02:07 PM
I also don't understand. If it is indeed the case that they are only going to be screening windows 3rd-party clients, then these clients will just masquarade as Mac clients.
They're going to have a hard time glossing over small differences in the TCP/IP stacks.
Posted by: Stripes at August 30, 2005 02:54 PM
Most people consider IM to be private. I don't want anyone to poke around in my conversations and certainly not to deliver 'added value' aka ads. (For the same reasons, I do not have a email account at hotmail, gmail or any of the webmail providers.)
Oddly enough while gmail would be percieved to be the most invasave with it's targeted via your email ads, gtalk is technically the only major im client that is hard for peopel to easedrop on.
It is the only one that is encrypted (and they use a fairly well proven encryption too). So when you send your gtalk messages you know what you typed, nobody between you and google's data center knows, then (unfortunitly) google knows, and then nobody between them and the person you are chatting with knows. If you look at AIM when I type a message to my wife it goes from CA to VA across a ton of routers and many many miles of fiber (and a few miles of copper), and anyone along the way can read it. Probbably not by tapping the fiber, but definitly by breaking into the router, or putting something else on the gigabit ethernet in an ISP's POP...oh, and then AOL can read it, and then anyone on the return path back to CA can read it.
If you don't beleve it, open a terminal window (I'm assuming you have a mac...if you have Linux, FreeBSD, OpenBSD, NetBSD, or some other unix you should be able to translate this) and type: sudo tcpdump -s 2048 -X -i en1 -v tcp port aol
If you have ethernet rather then airport use en0 rather then en1, for a modem use ppp0, anything else you are on your own.
Type something in your AIM client. Then look in your terminal window for it. MSN and Y! were both the same last I tried (diffrent port for the traffic, not "aol"). If you want to try with Google try 5222 or 5223 rather then aol, but you won't get anything you can decode (at best spying on that traffic can give you ideas about who is talking to who, and how frequently, and how much they say...which if you are the CIA you can do something with, not a whole lot, but something).
Posted by: Nugget at August 30, 2005 03:43 PM
If you connect to Google Talk using your own Jabber client then the traffic is encrypted. However, Google's own client does not encrypt the traffic. They say they plan to do this eventually.
From their FAQ:
"Google Talk currently does not encrypt chats or calls. But we are working hard to make many improvements to Google Talk while it is in beta, and we plan to fully support encryption of chats and calls before our official release."
Posted by: blue at August 30, 2005 07:07 PM
it seems to me, the big prize here
is in creating a video/audio/text communications tool.
somewhat like the iChat AV client.
it should be Open, for maximum network effect
but the company that makes the First Move will
want to grab a monopoly quickly (like iTunesMusicStore)
apple already made the first move with iChatAV.
What is the second move? and does google have
anything to do with it? In order for Apple to successfully
leverage the video portion of the iChatAV.app they need
there to be cameras in the hands of their customers.
you can't expect the kids to all go buy an iBot thing.
the cameras need to be able to hook to the macs
via firewire/usb][ and function as 'computer cams'
a digital camera built around the alphamosaic chip
should do the trick, connection is via the poddock
to synch iphoto and the camera.
1,000 photos in your pocket
-][
Posted by: Anonymous at August 30, 2005 09:06 PM
I don't understand how you know your prediction to be true.
Posted by: P. Huesken at August 31, 2005 04:10 AM
Interesting article. Some comments:
- Gmail isn't coming out of Beta 'now'. It just opened up the beta some more for users in the US willing to share their mobile phone nr.
- The article doesn't mention Skype; Skype focusses primarily on voice, but then again, IM will shift more and more to voice anyway. And with 155 million downloads (and +3 million users online) they'll cause some rounding errors in your calculations.
They'll keep dragging chatters to their chat clients because they offer (almost) free SkypeOut and (more and more) SkypeIn opportunities.
Ánd they've just opened up their SkypeNet API's
- Google will get a lot of the *normal* IM users as soon as they start to offer their Google Talk via a webservice (like MSN is offering), ánd after they start to deliver presence info in Gmail (which i'm sure will happen pretty soon)
Personally i doubt whether Yahoo, Microsoft and AOL will join to fence of (others and) Google together; and i seriousely doubt it'll be a succesful effort if they choose to do so.
Furthermore, an interesting part of the chat market will be mobile pretty soon, and *a lot* of those mobile will not be running Windows Mobile. And the driver to get a non-Windows Moble phone, or a new chat client on your mobilephone, could as well be 'cheaper telephone calls' form Google, Skype etc.
Posted by: sits at August 31, 2005 06:48 AM
They're going to have a hard time glossing over small differences in the TCP/IP stacks.
TCP/IP fingerprinting is fraught because you don't know what's in between you and the Windows machine. Some firewall/router products/operating systems can do normalization on the packets thereby scrubbing identifying marks away (or replacing them with something other marks). I'd imagine large corporations are likely to have such products deployed and MS is smart enough to not cause them undue problems.
In short, you are unlikely to see MS using TCP/IP fingerprinting to refuse connections (that doesn't mean they won't try and find something else).
Posted by: Naib at August 31, 2005 07:13 PM
Quoting P. Huesken,
"IM will shift more and more to voice anyway."
I disagreee. Voice/video is fun - I have an iSight, and pull out iChat now and again to use it - but for the most part I do text. Voice limits you to one thread, not only for your chats (you can only have one voice conversation going) but, say, if someone walks into the room and you want to talk to them. With text, all conversations are queued up and waiting for your attention; anything realtime is, despite being more fun, much less convenient.
Posted by: negrul at September 1, 2005 07:59 AM
Great article, DB.
There is a general impression that Google is just what the market needed, a "kick in but", sort of speak and the step forward that comes with it. Mostly Google is playing the "rogue" act, trying to push into the market what most of the users "imagined" was needed. It does not take a genius to realise a new "flagrance" was required to push some more doe in some pockets.
However and this is just a pea in the pot, how sure are we Google is not a product of the "Triumvirate", just to get a "license to kill", to change the net by carefully avoiding any charge in "attempted full control", like Microsoft was accused some years back. This is just a wild theory but does any of you spot a pattern in the development of "free for all" enterprises? I guess this popped up after reading DB's article, never knowing that "the three" were actually chumms.
I incline to the ideea that users see in Google a new era and associate all its products with the "last and best resort". Isn't that dangerous?
Don't get me wrong, I only use the Google mail account and I salute their development efforts, especially some very well crafted "signals" that denote balanced and focused "programming" (I am referring to Google's mail features). A keen pair of eyes watched the creation and disposition of the features making sure all things we missed in the others services are now in one place. I call that good market evaluation and very good bussiness.
Posted by: Paolo at September 1, 2005 08:44 PM
Before I read this and about a week after trying GTalk, I made a pretty big decision. My decision was to bail on Trillian and all associated IMs, even at the risk of having an empty contact list, for two reasons. First, I was tired of the bloated chat programs, and having to have so many different e-mails and logins (in spite of being a Trillian user), etc., just to keep in touch with friends and family. Second, I could see the future of what GTalk offered and I liked it. So I made an announcement to my contacts that I was closing up shop with Trillian and moving to GTalk. Since then, many have switched with me. Having read this post, I'm now convinced it was a good idea.
As an aside (I'm surprised no one has mentioned this), it is possible that Google knew how the "Triumvirate" would react and is counting on them doing exactly this for some alternate reason. In short, while they're all forcing users to upgrade and mounting a united front, Google slips some new, and unexpected feature into the game and profits from their distraction. Classic Sun Tzu.
Posted by: Julian Bond at September 2, 2005 02:11 AM
As someone else said, why no mention of Skype? And note to the linux expert above, Skype is the other IM client that is all encrypted.
Google Talk is v0.1 I'm waiting for v0.3 at least before passing judgement as v0.1 is little more than a proof of concept. But there's enough there to see more or less where it's going. I think a Google, Jabber, AOL, Apple, SIP link up could change the landscape considerably and make life very hard for MSN, Y! and Skype.
But what a week. Google sidebar, Google Talk, MSN v7.5, Skype 1.4, Skypenet and Skypeweb announced, MS buy Telio. Really good to see this area getting busy again.
Posted by: Richard Jones at September 3, 2005 12:27 PM
From a technical standpoint, locking out third parties forever is pretty simple. Force the client to provide a hash from a server-specified byte range of the executable, and you're done. Third-party software would either have to distribute a copy the official client (very much illegal), or require the user to download his own copy and piggyback on it (extremely inconvenient, possibly illegal).
Actually there's a way for 3rd party clients
to bypass this mechanism. The 3rd party itself
runs a service which answers these requests.
The 3rd party client asks this service. The
service itself needs a copy of the official
binary - perhaps even a purchased one - but because it's only answering hashes
there is no (C) problem.
Rich.
Posted by: at September 7, 2005 05:58 PM
I disagree with the following:
>At the very least, they can make it so difficult >for third parties to keep up -- and reverse >engineer what they're doing -- that its such a >nightmare few if any will bother
Few will bother? Well, it really only takes one. It's been done before and it will be done again. It will be impossible to keep third party clients, Windows or otherwise, off of their network.
Posted by: Xenex at October 12, 2005 02:58 AM
I didn't believe it would happen... but it's beginning.
And I just modded this up too!
Posted by: ctsc at February 11, 2006 11:49 PM
Something that seems to be missing is the classic response from the triumvirate. An open protocol (or seemingly open as some have noted) comes along and what does the triumvirate do? They close up shop, shutting users out. If any of the big three would wise up and smell the roses they would see the obvious strategic decision would be to trump google and offer multi-protocol multi-network, multi whatever to the masses of users they already have, before google gets around to it. Think of the headlines on msn, aol, and yahoo. Now you can use your MSN/AOL/Yahoo IM account to chat with users on Yahoo/AOL/MSN!!!!! This would immediately eliminate incentive for users to switch to google outside of the early adopter set.
Interesting stuff. I wondered what you were putting together about this. Good take, and might be worth watching from the sidelines.