Last August Gizmodo wrote the obituary for AIM, it turns out they were only 7 months early. With the majority of my old AIM team let go last Friday, the team is down to a handful of engineers, testers and others to maintain the client. We have seen this story before with the AOL dial up client. When AOL was put in maintenance mode it kept a few key employees around in case something went wrong. So here is my obituary for AIM:
AOL Instant Messenger passed away on Friday March 9th after 16 years of complications related to poor management. As the inventor of many key social features and the inspiration for a few important features in Facebook, AIM was the web’s first social network. AIM will be remembered by the many who worked on it and it leaves behind many patents that AOL may choose to sell or license.
So AIM’s servers are not getting turned off, there are still a few million using the service but I doubt we see too many new features. And while Mashable does not declare AIM dead, Christine Warren does touch on where AIM went wrong. Going back over in my mind we lost our way when we could never convince AOL’s corporate bosses to open up AIM to 3rd party networks until it was too late.
It was a great run for AIM, and all of us who worked on it should be proud of the work we did to change the way we communicate online.
I was struck by an interview that Mark Zuckerberg gave at a Y Combinator event this weekend in which he said that if he were starting Facebook today he would have stayed in Boston. Zuck talked about how Silicon Valley is “a little short-term focused” and that there’s “a culture in the Valley where people don’t commit to things.” He brought up an example where in Seattle employees stay at the same company twice as long as the Valley.
Zuck makes some good points, though the stat he quotes from Jeff Bezos about Seattle workers is a misleading number. Seattle has strong companies that treat their employees well with Amazon and Microsoft. When companies like Google and Facebook opened up Seattle offices, they were able to recruit top talent from Microsoft and Amazon to those new offices as easily as they did in the Valley.
I do agree with the comments about the Valley being short-term focused. The company I work for today, Clearspring, which is based in Washington, DC, is a perfect example Valley focus versus non-Valley focus. Clearspring has had a couple of different acts in its life as a company. If Clearspring were based in Silicon Valley, odds are the company would have been sold once we shut down our widget business. Instead our investors remained patient and management focused on the long term, we actually MADE an acquisition, and now are the largest sharing platform on the Internet.
When Zuck talked about Boston it made me nostalgic for the days of AIM and AOL. Before I started at AOL in 1999, we bought a company based in Boston called Booklink. Booklink was a browser that eventually became the foundation for the way we would compose and display instant messages in AIM for many years. The best part of that acquisition were the guys that came with the technology. A few of them were the original authors of the Windows AIM client.
The point Zuckerberg makes, and the thing I am trying to reinforce is that talent and technology can succeed at any time and any place. Silicon Valley may be the heart of consumer technologies but life does exist outside that bubble. Whether Mark really meant what he said about not moving the Valley does not matter, companies like Foursquare, TripAdvisor and Clearspring we thrive no matter where we are. Success comes down to ideas, people and execution, it does not matter where you are located, and I think that was Mark’s point and I know it is how I feel.
Finally, I have rescued all my old blog posts from 2007-2010 and imported them here. If you are looking for old information on AIM, ICQ, AOL or AIM MusicLink, the posts are all here. It was a trip going back to read all the things I worked on back then. By now, Google should have re-crawled my blog and indexed the posts.
Years ago I wrote a post on AOL Journals about September 11th, 2001 and the memories of the day. While I can’t link to that post anymore since AOL Journals was retired at the end of 2007, I still have the text from the post.
The day started like any other for me. I got to AOL very early, around 8AM, thanks to living 10 minutes away from the office. We were preparing for a release, and at that time the AIM build machine was in the QA Lab on the 3rd floor. To make sure all the builds were kicking off properly I went up to the lab at 830A. The TV in the lab had The Today Show on and in the next 18 minutes my view of the world was about to change. The build was long done by the time the second plane hit the World Trade Center, but I was not about to stop watching. Already shaken up by watching a plane crash into WTC, at about 940A, the fear hit much closer to home as images appeared of an explosion at the Pentagon. Being directly under the flight path for Dulles where planes landing come over the building at about 500 feet, AOL decided to evacuate the buildings. Life changed that day for me, as it did for countless others I am sure, but for that hour I was in the QA lab will sit with me forever.
One thing I remember most about that day is how AIM kept everyone connected. Phone lines were tied up and email did not always work, but people’s online presence on AIM was reassuring to see and to know they were OK. We heard from users for weeks after the attacks that AIM helped them through that day.
On September 11th, 2011 we need to never forget what happen 10 years ago.
Turning off the multihead unwinds one of the most over engineered solutions in the history of messaging. When AOL and Google did the original deal around messaging in 2006, we were not ready to do federation, so we proposed a multihead. At the time the negotiations started Google only had a desktop client, but soon after we started talking Google unleashed something called Caribou on us. Caribou was the code name of Gtalk in GMail, and needless to say it threw a wrench into our plans. In the end Google wrote a server side multiheaded experience using a client side library that had to handle 1000s of threads. It was a miracle it worked, and a lot of great engineers made it happen. Thank goodness we can turn it off now.
With federation comes the hope that we can break down the walls of the IM networks that have existed for 12+ years. For users they get the ability to talk to their friends regardless of networks and use the application they like best. For the networks, federation means an application battle royale. Personally, I will use GTalk when I am at my laptop at work since we use Google Apps here at Clearspring. For my iPhone and iPad I use AIM because Google’s mobile experience on iOS is lacking.
Let’s hope that the other networks can figure out how to federate with each other and we can all stop using clunky multiheaded IM applications.
It definitely brought a smile to my face to see Gizmodo praise AIM 5.x. My favorite AIM version I worked on was AIM 5.5 on Windows, back when I still used a Windows machine. That particular client really was close to perfect, and achieved perfection thanks to James Dennis’ DeadAIM addon that removed the Ad and added a few other key features. CNet calls it plain, but by then we had built a nice Expression engine and had audio and video chat.
What made AIM so important back then was the away message which were as irrelevant as a lot of Tweets are today, like “In the Shower” or “I am watching Survivor.” We used to watch our users in usability tests view each of their friends Away Messages. It even drove us to create a simple web page where you could view all of your buddies current Away Messages and their profiles.
Ah, profiles, 1024 characters (HTML markup included) of personalized goodness. This was MySpace years before Tom became everyone’s friend. Profiles were so important, that when we launched the ill fated AIM Triton without profiles, I added an AIM plugin to put them back. (Props to Justin for keeping his old Running Man blog alive on Blogspot after AOL Journals was shut down.)
So while Gizmodo declares AIM dead, I think they missed the AIM Desktop death by a few years. AIM is making a big push on being a web destination and they still have a respected mobile experience. Hopefully AIM 5.5 can rest in peace, as the real-time communication platform moves on to what’s next.
I am proud that the last thing I worked on during my 11 years at AOL finally was released yesterday. AIM AV was something that was a long time coming. It addressed two of the bigger weaknesses AIM suffered. It was getting lapped by messaging clients on the web by Meebo and GTalk among others. The other area AV addresses is a new set of audio and video technology.
AV finally starts to get AIM away from the original audio video stack I wrote with others back in 2003 and refreshed in 2006. I am proud of the work we did 5+ years ago when we had the foresight to move away from Microsoft RTC and integrate ON2’s VP7 and Global IP Solutions. Later both those companies were acquired by Google and now make up a major part of WebM.
The new experience lowers the barrier to entry, where anybody can arrive at aim.com/av and immediately start a video conversation with a friend. With no download necessary, getting non-technical people to use the product is super easy, no Geek Squad required. Simply just copy and send the link. The team did a great job getting AV out the door and making it the easiest audio and video experience on the internet, congrats guys.
I have been meaning to write about Twitter’s handling of UberTwitter and how hard it is to run an API program when balancing that program against revenue targets. I certainly felt this when I was leading the Open AIM program. We were constantly debating internally what was more important, owned and operated AIM clients, or the “power of the network.” Twitter, I think is having some of these same internal discussions now.
For services and products that have open APIs, it is an awesome opportunity to grow your network and delight developers. APIs work well if they can directly contribute to your monetization through licensing, or indirectly by contributing to the ecosystem. Where it gets tricky is if your APIs do neither. Companies start putting in controls and terms that prevent developers from innovating, and it is a slippery slope.
I cannot speak for what is happening at Twitter. I love Twitter’s API, I have used it for years, and know that without that API, Twitter would not be here today. I can share what happened with OpenAIM, that early on we were able to monetize through licensing, but never took it seriously enough. Since we could not monetize indirectly off of OpenAIM, because monetizing AIM was always a difficult task, we started putting limits on the access and rules for developers. That was a death knell for OpenAIM, especially when Google had opened up GTalk. Hindsight being what it is, I wish we had pushed harder on the licensing front and basically completely opening up the rest of the program with no strings attached.
For Twitter, I hope they can find that balance that allows 3rd party developers to continue to have the access that they have enjoyed. Innovation is a great thing, and Twitter has to thank a lot of the developers who have used the platform to build things like TwitPic.
At peak AIM MusicLink had nearly 10M sessions per month on the AIM Windows client. I am confident to say it was the most successful AIM plugin of all time. Looking at the latest data, there are still a lot of people who are still using the plugin, for which I am thankful. It was the community that drove me to make this plugin, and it was the feedback from my old blog that helped drive features.
Looking at the analytics from this blog, it is clear that people are still looking for info on AIM MusicLink. So what happened to the plugin? In the middle of 2009, when I was working on AIM, it was decided that we stop supporting plugins in the Windows client. There were a variety of reasons for removing the support, but as soon as we did I stopped working on the plugin.
So what started out as a test for OpenAIM 1.0 in the JAMS plugin package, became a real user favorite. I even got a designer to give me a fancy icon for the plugin see at the top of the post. Before there was last.fm, Pandora, Ping and music scrobbling, there was MusicLink. Hopefully it made the experience of sharing music with friends a little bit more interesting.
We pushed live a new AIM build last night that sets the stage for some good things in the near future. The first change that I am glad to share is that the AIM Running Man is back. When I started on AIM almost 10 years ago, the running man was the icon for our messaging app and remained so until 2004, now he’s back. You will see the new running man icon all over the place including OpenAIM, aim.com and AOL properties.
AIM 6.9 also takes your AIM buddy feed to the next level by replacing the orange feed indicator icon with the “favicon” of the most recent update. For example, if I were to upload a picture to Flickr, instead of seeing the old orange icon next to my screen name, you will now see the Flickr favicon.
New Feed Indicator
AIM 6.9 also supports better ways of finding friends and adding new connections in your buddy list via Add Buddy and our first time user experience.
For OpenAIM developers wanting to build plugins for the latest release, we will be doing a new Open AIM SDK in the very near future with the latest API additions that coresponds to the AIM 6.9 release.