Username:

Password:

Fargot Password? / Help

web services glue

Sticky platforms

This blog post about Leo Laporte's strained relationship with Twitter has me thinking this morning. The post outlines the "dangers" of twitter (using the front men of Laporte and Dave Winer) as a "closed" or "centralized" system. Now, as a twitter user (and fan), I completely get the closed and centralized argument, but I think it's a bit off. Since time immemorial, techies have talked about "lock-in," specifically the dangers of lock-in when it comes to platforms. Microsoft's past predatory behavior probably didn't help much, but I do think a lot of things have changed. These days platforms are expected to have APIs and to allow for some level of data portability. Those that don't are quickly ridiculed and often shamed into changing (at least somewhat). We can argue about whether Twitter is "too closed," but what we can't argue is that Twitter is a modern day version of the "walled garden" (circa AOL). We can't even argue that about Facebook. The old walled gardens were just that -- walled. As in closed. As in, you ain't getting an app in or info out until you have 7 meetings with our biz dev guy. [Later: a friend reminds me that the "7 meetings with a biz dev guy" practice is currently how LinkedIn operates, thus making LinkedIn the new AOL.] That's just not the case anymore. And the debate should change accordingly. It's no longer a matter of open or closed. It's a matter of how open (closed is off the table as an option). And once you enter the realm of "how open," you've now firmly stepped into the topical topography of Glue. Gluecon is really all about how open. Platforms can "get sticky." So can architectures (WOA vs. SOA) and clouds. The question is to what extent and through what mechanisms.The way we talk about these problems matters. Getting stuck in the "open vs. closed" platform debate just keeps us going in circles in 1998. To advance things we need to be hammering away at the "hows and whys" of making platforms, clouds and architectures sticky -- where "sticky" may mean interoperable, integrated, federated, whatever. Returning to the Laporte post, can we put the Twitter pandora back in the box? I doubt it (sorry laconica). But we can begin talking through the glue of platforms.All of which begs the logical question: what platform (and underlying cloud) discussions will we be having at Gluecon? Short answer: plenty (and note to vendors: we will discuss your platforms whether you're there or not - sorry). Long answer: Facebook, Salesforce.com, Azure (and the Geneva toolset), Twitter, Google, and even just the good ole interweb itself. Come help us move the discussion forward.

A down day leads to an agenda refresh

I'm a bit under the weather today, so I ended canceling my Monday phone calls (which makes Tueday through Thursday hellish, of course). That left me with some "easy time" to work on the Glue agenda - something which I get more excited about every day. Here are some updates: 1. In case you missed it earlier David Heinemeier Hansson and Mitch Kapor were announced as Keynote speakers. 2. To that I've now added Josh Elman (of the Facebook Platform group) and Bob Frankston (co-creator of VisiCalc). Facebook's recent movements around "openness" should provide a pretty fertile foundation for Josh talking about what "gluing together the web" looks like from Facebook's perspective, and Bob (who is just flat out, scary smart) will be addressing how we need to re-think some of our assumptions around bindings, platforms and achieving the structures of simplicity needed for innovation. 3. Keynotes are always fun to ooh and ahhh over, but when you really start to dig into Glue's agenda, things get truly glorious (hat tip to Will Ferrell). Sidenote: I'm now writing about stuff that, for the most part, hasn't even been updated on the agenda page yet. 4. Pam Dingle, of Nulli Secundus (and one of the smartest "identity people" I know) is going to be leading a workshop session on the "domain-less" enterprise. That is, "how far could a company get today, with the tools and protocols available to them, towards creating a business where employees could access exactly the same work environment from any computer in any location, without any concept of being inside or outside a network perimeter. " The session will take a look at what's available in the way of distributed identity tools, and specifically show how to do this with Microsoft's beta Geneva set of tools, as well as a "how to" with open source tools. [sidenote: those of you that have seen me write about optimism in the past know that I love Pam's blog's name.] 5. A session around Data Portability will feature Daniela Barbosa and Chris Saad from the Data Portability Working Group and Ben Metcalfe with a perspective of what he thinks is right and wrong about data portability. I think this discussion is *vital* to the overall Glue picture, and I'm hoping this talk (in particular) draws a bunch of the cloud computing crowd. 6. There's a session coming together around Cloud database standards. The whole idea actually comes from a post that Albert Wenger wrote, and the session is meant to tease out some of the threads around what I'm finding to be a REALLY important topic to this whole "glue mess" (databases, that is). We've enlisted Alex Iskold of Adaptive Blue so far (Alex made a key design decision around SimpleDB), and I've got invites out to some Google folks, etc. [sidenote: I applaud Alex's product naming choice. ] 7. There are a whole bunch of sessions coming together around web app description languages (needed?), cloud interoperability on a grand scale (possible?), the open social stack, and how RESful APIs play into rich internet applications....and so much more. 8. I'm also adding some key moderators/discussion leaders -- notably, Jeff Nolan and Stewart Alsop (of Alsop Louie Partners). 9. Don't forget all of the already announced stuff around Web Oriented Archiecture (Aaron Fulkerson), Harnessing the Cloud (Mike West of Saugatech Research), Leveraging API infrastructure (Kevin Matheny from Best Buy and Oren Michels from Mashery), Complex event processing across web apps (Mike Clymer), etc etc. 10. And the best part of all? The agenda isn't even CLOSE to finished, so the goodies are just gonna keep on rolling in. Topics yet to come: data integration and mashups, web app integration, glue metrics, gluing together devices and data, social networks and glue, the evolution of the client-server model, etc. I really hope you're gonna choose to join us for Gluecon. If you wanna get lost in the crowds and listen to some non-impactful stuff, go somewhere else. If you want intimacy, connections, and sessions and interactions with impact, come to Glue. Seriously. (last sidenote: don't just take my word for it, ask Pete Warden or Sameer Patel, or Zoli Erdos about the kinds of interactions you can expect.)

Financial crises and the leading edge of Glue

Everyone's hearing about the mess on Wall Street. In that context, this bit in Computerworld caught my eye:
 Paul Polishuk, president of the research group at Information Gatekeepers Inc., said that the increasing number of mergers and acquisitions will boost the market for IT firms that specialize in integrating networks. "Because Merrill Lynch is going to be bought by Bank of America, their assets are going to have to be upgraded," he said. "And since Bank of America and Merrill Lynch are two very different businesses, a good deal of work will have to be done to get them integrated."
It made me think about some of the new "web oriented architectures" and how all of that will get effected. Now, ideally, the "web as platform" erases those problems, right? Yea, right (not)! It seems to me that "glue" companies might just be born in the fires of various cutting edge integration projects that this crisis will bring. After all, nearly every great software company was born in a "recession." 2008-2009: the years of crisis that give birth to Glue software companies. Mark the dates.

Gnip - a great early example of G-lue

We're ramping things up around here -- revamping the website, beginning to recruit some sponsors, and just generally turning up the heat. In the meantime, I thought it would be good to start highlighting some "Glue" companies. My feeling is that folks are going to be bombarded by "cloud" or "SaaS" or something conferences in the next year - none of which will properly hook into the Glue aspect that is crucial to it all. As such, I wanted to starting getting a flavor of what we mean when we say "glue" (from so many different angles). Accordingly, here's a great ten minute video interview with Eric Marcoullier of Gnip (pronounced, Gah-nip) - it's a great early example of Glue (gah-lue - sorry, couldn't help myself) stuff.