BEA & Plumtree: The M&A Q&A

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

A few minutes before 3 PM MT yesterday, I got an email asking me to attend a breaking BEA conference call at 3. Fortunately, I had the slot open (increasingly rare these days), and dialed in. Generally these calls mean that the firm has made an acquisition (if BEA had been acquired, I would likely have gotten an email from the acquiring party rather than BEA so that wasn’t it), have a major product announcement, or less frequently that they’ve signed a big partnership. While the conferencing service had us on hold for a couple of minutes playing some truly awful music, I was speculating on what, in fact, the breaking news might be. Something open source? A spin off of one their product lines? Well, I was wrong on all counts. Instead, as you’ve all probably seen by now BEA used the call to announce their intention to acquire Plumtree, one of the last remaining standalone commercial portal products. Before I get to the Q&A, the standard disclaimer applies here: BEA’s a customer of ours, while Plumtree is not. So read into that what you will. on to the first question.

Q: I haven’t been hearing much about portals for the last year or so, why does this deal make sense in that context?
A: Well, despite the fact that portals might not be top of mind in the blogging world, they’re still an important cog in many an enterprise infrastructure. I’ve been fairly unambiguous in my belief that thin client based infrastructures are becoming far more compelling for previously unheard of functions, from email to CRM to – gasp – ERP. Portals are the enterprise aggregation point of this trend, where big firms can surface a variety of applications and content sources within the context of a browser – thus eliminating the expense and management headaches of traditional client/server architectures. Another way to think of enterprise portals is as the enterprise equivalent of Google Maps mashups, hybrid blends of separate services.[1] So while they don’t get the pub they once did, portals play a vital role to many medium and large enterprises.

Q: Well, if portals are such a hot commodity, why did Plumtree agree to sell out?
A: Well, I don’t know that I’d refer to them as a hot commodity, but the reason to me is simple: it’s difficult for any 415 person company, however talented, to compete with industry giants like BEA and IBM on a level playing field. There are ways of tilting the field in your favor, as JBoss has shown by leveraging open source, but the surprising thing to many was that Plumtree remained independent as long as it did. This really is a credit to their focus on the business side of portals, which we’ll get to in a moment. In the end, I have to believe that a deal like this was simply a matter of time. Not only would a standalone vendor like Plumtree have to continue staving off competitors who’s engineering resources number more than their total headcount, they would undoubtedly increasingly face competition from open source projects like Liferay and Exo which are receiving due consideration from open minded Fortune 500 architects. And that’s just the enterprise category, we haven’t even gotten into the simpler, lighter weight alternatives like Mambo from the PHP world, or Sharepoint from Microsoft. Plumtree did exceedingly well, all things considered, in fighting a one front battle, but the opening of a second front – the low end – likely made going it alone a difficult proposition for the portal vendor.

Q: Got it. So what’s in it for BEA?
A: Well, as my colleague mentioned to me this morning BEA now has the ability it lacked to sell to a new audience of .NET developers which, while introducing some complications, does open up a new channel. That’s sort of a good news/bad news deal, as he put it. According to Forbes, the win for BEA is the broadening of its customer base. This was a point stressed by both BEA’s CEO Alfred Chuang and CTO Mark Carges – that BEA and Plumtree were really different markets with minimal overlap, so the combination is a 1+1=2 rather than 1.5 situation. I’m not sure I buy that, myself, given that I know many of Plumtree’s customers are big businesses that would otherwise be in the market for BEA-like solutions. As different as the strengths of the two products are, I simply can’t believe there won’t be situations – a fair number of them – where they’ll be competitive with each other.

Q: Well, ok, what do you think BEA actually gets out of the deal then?
A: Well, I think what’s actually critical is that BEA just acquired a company that knows how to sell portals to business people. That point, to me, cannot be stressed enough. As I told one reporter this morning, BEA and many of the other J2EE portal players could sell portals to CIOs and enterprise architects in their sleep. But like their respective pitches for SOA, selling to the actual consumers of portals – regular, line of business folks – BEA, along with most of its J2EE brethren, were less than stellar. That, however, is something Plumtree excelled at. I had the opportunity a while back to attend one of their customer sessions in Massachusetts entitled “No Empty Portals,” and there was little discussion of JSR 168, EJBs and the like – it was instead a talk about how various Plumtree customers had built portals that business people actually liked, and enjoyed working with. They talked about portlets for employee classified ads, comic strips, and other seemingly frivolous things that served an unexpectedly important purpose – they brought people back to the portal. In keeping users involved and interacting, they kept the product from the fate of many a portal product – shelfware. When helping to design part of a portal initiative for a customer in my SI days years ago, the customers’ architects I was working with had actually forgotten they even owned an Epicentric license (long since acquired by Vignette) – that’s how little they got out of that buy. So the ability to sell business customers on and engage them with portals, to me, is the best asset that Plumtree brings to the table, and a likely driver behind the acquisition.

Q: Assuming the deal goes through, what does this do to the competitive landscape? What’s the outlook for the combined entity?
A: Well, as far as the competitive landscape goes, it’s consolidation pure and simple, no different than we’ve seen in the past with the application server market and other enterprise software segments. As far as what the outlook is for the BEA/Plumtree offerings, I think it’s interesting. Tactically, it should be a boost to Plumtree’s sales because it removes at least some of the questions potential buyers might have about longer term viability, support, etc. The strategic outlook is less clear, for roadmap issues. The road ahead for BEA’s own portal offering is likewise hazy; clearly its reputation for scaling outstrips Plumtree’s, but apart from that Plumtree’s offering would seem to be more usable to a wider audience – and perhaps the most important one – the business person, as Carges alluded to on yesterday’s call. I don’t know about you, but if I have one product that business people can use and one that they can’t, or at least are less inclined to, I’m probably going with the former unless I have a really compelling reason not to. So the product roadmap has me a bit confused.

Q: In what way?
A: Well, all the participants in yesterday’s call promised that the product lines would be maintained as separate virtually indefinitely. In the short to medium term, this obviously makes sense, as all three codebases (BEA, and presumably a Plumtree Java and a Plumtree .NET) are undoubtedly fairly complex. But I would have expected more discussion of at least the possibility of a consolidation somewhere down the line – maybe just for the Java versions, incorporating the best features from each product, etc, etc. But there was little to none of that. That seems odd.

Q: What does this mean for other portal vendors like IBM, Oracle, or Sun?
A: Well, from an IBM perspective, probably not a great deal. Oh, one of their major competitors just acquired a popular and competitive offering, and undoubtedly that will be a point of concern for some of the Portal folks over at IBM, but IBM’s walking a different path these days. For customers that want a portal and nothing else, they have that option, but the bigger picture strategy from them is Workplace – which is about trying to blend the best of thick/rich client and thin/portal client worlds. For a very positive view on Workplace, you might check out frequent RedMonk guest Gary Edwards’ comment over on one the ZDNet blogs here.

Oracle’s probably likewise not thrilled that one of their middleware competitors has acquired some new capabilities and – in theory – a new channel, but the conventional wisdom regarding Oracle’s software position these days has them more focused on being an all-in-one supplier of business applications. A standalone portal solution then would seem to be less of a concern for them.

As for the folks from Sun, they likewise are not really competing on a single product basis in this space (though they will certainly sell the component standalone, it’s not their focus), pushing instead the per employee pricing integrated Java Enterprise System offerings. So this announcement, while relevant, is not likely to affect their field sales force as much as might otherwise be expected.

Q: Any other important considerations?
A: Well, the really interesting one from where I sit is whether or not Plumtree’s business-savvy DNA will be overwritten by its acuirer’s strongly technical culture, or whether or not BEA can practice some recombination on-the-fly and apply some of the Plumtree lessons to other areas of its business, like the aforementined SOA product set. If BEA twelve months from now is marketing successfully to your Joe Line-of-business, the technical assets from this deal might become an afterthought. For now, however, BEA has to hope that it can successfully manage two separate product lines (and three separate codebases), while maintaining complimentary but not overlapping marketing messages amidst competitors that we will be FUDing them at every opportunity.

Q: Anything I should have asked but didn’t?
A: Well, I do find it interesting that most of the coverage of the announcement that I’ve read has turned a blind eye towards potential open source or lightweight threats to BEA/Plumtree, as well as other portal players. If you talk to James McGovern, he’ll tell you that every enterprise architect should be actively considering some of the available open source packages, and that’s with a relatively immature open source portal landscape (as compared to, say, application servers). Likewise, I think the proliferation of web services may represent a long term portal-as-central-aggregation point threat, because services like Greasemonkey – kluge as they are – empower ordinary users to build their own portals with little or no help from IT. This may all be more of a strategic than tactical threat, but it’s something that portals will ultimately have to contend with, like it or not.

[1] Interestingly, the consumer side services are generally more adept at actually blending the services, while enterprise portals have typically been more centered around simple aggregation than service recombination.


  1. Should BEA use Mono to run Plumtree portal on Linux servers ?

  2. i think you know what my answer to that would be 😉

    but i should have thought to mention it; maybe in a follow up, and certainly will be sure to mention it to them when i get the chance to connect with them directly.

  3. Actually, I can’t really understand this buy. Plumbtree was a dead company. Technologically, they brough the portlets concept to the market but, since JSR168 they became irrelevant.
    On the sales level, they had some very focused sales team but, 200 Million USD would pay all their salaries (doubled) for a long time.
    Has for buying the competition, Plumbtree decided to stick in the Windows market (they had a solaris version that was to horrible to use) so, it would be just a matter of time before Microsoft finishing them off.
    IMO, BEA just bought a few new customers before sharepoint being a credible offer in the market.

  4. guess we’ll have to agree to disagree here, Jaime. i wouldn’t contend that they’re the technological leaders from a scalability perspective, but from the standpoint of making portal functionality available to everyday people, i certainly don’t classify them as “irrelevant.” on the Solaris front, i’ve spoken with customers that were quite happy with the Java version in the past, so neither would i classify them as a Windows only ISV.

    we’ll see what comes of this, and i’m not convinced this is a good move for BEA – much depends on who they’re able to retain from Plumtree, as someone else commented to me privately – but i don’t see the move being as universally poor as you do.

  5. jaime I bet if if it was Sun you would applaud the move, bar a few gripes about overlaps in functionality and a lurch to towards the .NET darkside. 😉

  6. Maybe that’s the whole point James. Perhaps BEA purchased Plumtree to keep it out of the hands of Sun? With media frothing expectations of SharePoint finally reaching parity, perhaps the market for Plumtree has been capped. Like everything else in the Windows universe, when Microsoft sets their sights on a Windows market category, it’s a done deal. Game, set, match – without need for a single instance of real world validation. But that’s the way it goes when you’re a monopolist.

    If Plumtree’s customers weren’t concerned about the survivability of the company, perhaps they should have been. If BEA can create a mixed environment model that assures the Plumtree user base that their investment in Plumtree will continue to be competitive, then that’s a big win for BEA. Customers who are locked into the Windows desktop still need alternatives on the server side of things. Although i doubt very much that BEA has what it takes to make a credible argument for mixed environment solutions. They have to back the Plumtree purchase up with a lot more than the usual stockholders – analyst mix of grit, determination and chutzpah.

    The real winner though is IBM. Why? Because they have outflanked Microsoft, avoiding entirely the fate of being Plumtree’d. With WorkPlace IBM is able to attack Microsoft at the point of extreme MS advantage, the desktop – OS.

    Think about all those future bound customers who, because they feel locked into the Windows desktop, think that there is only one place to go – SharePoint. The monopoly base knows full well that there is an enormous cost of moving to SharePoint. But they also know that for them, there really isn’t any other choice. Because they control the API’s, Microsoft holds an indisputable integration and information flow advantage. The monopoly base knows this. So they buckle in and pay the incredible price of moving to SharePoint. That means upgrading desktop OS (XP – Vista), desktop hardware, and MS desktop productivity environment (MS Office 12, WinForms, VB.NET and the .NET framework), and embracing the rest of the MS server suite (Exchange, Collaboration Server, Server 2003 – Active Directory, and SharePoint).

    Along comes IBM WorkPlace, making a credible and demonstrative pitch to the entire monopoly base that they can have their cake and eat it too. With WorkPlace, the monopoly base can freeze the desktop, and still have access to advanced collaborative computing services. Meaning no hardware or OS upgrades. No license renewals. No Advantage Microsoft concerns. And no need to upgrade the desktop productivity environment. Just install a WorkPlace client on your Windows desktops, and achieve surprisingly superior integration with robust and proven “non Windows” server systems.

    Although there is a WorkPlace SDK available, it’s not clear yet just how far IBM will go towards allowing competitors like BEA and Sun to take advantage of this shot through the heart of the beast. I’m thinking that if Plumtree had access to WorkPlace, they could have gotten a much higher price. The reason? When you remove or neutralize the monopolist advantage, you remove Microsoft’s ability to cap a competitors marketplace with mere announcements.

    Of course, since WorkPlace is comprised of open source components (primarily OpenOffice.org and Mozilla), there is no reason why BEA, Sun, or Oracle couldn’t perfect the same thing. They wouldn’t be able to achieve the hard acceleration WorkPlace has with the IBM server stack (WebSphere, Apache, Lotus Notes, DB2 and Oracle). But why would they need to if the purpose is to achieve integrated acceleration with their own server stack models? It’s interesting to note that once you neutralize the monopolist’s desktop advantage, the market decision quickly shifts to who has the best server systems.

    With WorkPlace, IBM has eliminated “integration with the desktop” as the pivotal advantage of Microsoft servers. If BEA can’t do the same, buying up marketshare isn’t going to get them very far.


  7. Gary, instead of “IBM has” I would replace for “IBM would”. Workplace still has some proves to give and I have some doubts that IBM’s sales people will be able to push it’s Desktop offer a long way (and believe me, I hope I’m wrong).
    James, my company sells Plumtree. If Sun had bought it, I would have mixed feelings. For a side, it would give me some satisfaction but, for another, Sun would have made a wrong move.
    Stephen (and James), I’ve bloged about how I make decisions about with products I recommend to customers. Basically, what I said was Start looking at the architecture. Features can be ported to the competitors products.
    The way I see it, Plumtree’s only advantage (technological) was to have an enourmous amount of portlets already developed but, that could be easlly solved by BEA’s resources.
    If BEA just wanted to have an expirient team of Sales and Marketing to target business people, they could simply hired the 300 people relevant at Plumtree (300, of course, is an extimated number of the 500 people staff).
    But, I don’t exactly think this buy was an error. It makes sence IF (and only if) BEA is planning on targetting the windows market.
    That’s fine by me, if it’s BEA to do it. If Sun had tried to move into the .Net market, it would be a failure (we all know that). Sun may have some sucess with technologies integrating with Microsoft with the purpose of easing migrations. Never to fight Microsoft inside Windows based technologies.

  8. Yes Jaime. You are right that IBM has yet to move WorkPlace into the market. But since WorkPlace is comprised primarily of OpenOffice.org and Mozilla components mounted on a Cloudscape database, it's not a “new” product. Rather, it's a new way of wrapping and putting to use open source components that are road tested and proven. In particular, WorkPlace does a wonderful job with OpenOffice.org multi file format and XSLT transformation capabilities to create a shared space rich with Open XML technologies. The collaborative space WorkPlace open up is an Open XML space.

    I don't expect to see IBM sales people selling WorkPlace as a product – although they could do that. Rather, i see IBM's Global Services group promoting WorkPlace as a central piece of every SOA – ESB – Web Services effort. Wherever you find WebSphere, IBM Apache, Lotus Notes, DB2, or Oracle on the back end of an IBM Global Services contract, you'll see WorkPlace being installed on the desktop.

    There are many thoughts about why IBM “wouldn't” productize WorkPlace and directly challenge Microsoft's desktop dominance. To me the best rational is the “Divide and Conquer” strategy. This is a complicated concept that covers everything from WorkPlace, to the Power PC deals, to the Power PC 6 power, to the sale of Lenovo, to the cozy relationship with AMD, to the fascinating Xbox – Sony deals, to the stubborn hesitancy of IBM to license future PPC cores to Apple (resulting in Apple having to find succor with Intel). But the basic idea is simple enough. It's revenge time for IBM, and WinTel is the target. IBM is not stupid though. Once they got all their ducks in line, they moved quickly to the strategy of divide and conquer.

    The idea is simple enough, separate Intel from Microsoft. Then crush them both.

    Damn, i love this. Who wouldn't?

    Now is not the time to go into the glorious details of how IBM has prepared to pull this off, much as i would like to, but i would ask you Jaime to reconsider WorkPlace as part of this much larger strategy. A war with which you have no part, but could greatly benefit by if you pay attention and manage yourself into a favorable position.

    Under normal circumstances, selling Plumtree is a an absolute certain dead end. Sorry, but that's the truth. Microsoft owns that space. In fact, they own all the opportunities related to the Windows platform. It was only a matter of time for the market category to reach profitability worthy or Microsoft's attentions, and, for Microsoft to get rid of some other nagging details like the anti trust trial. But were now at the point where Microsoft is ready to claim what is theirs. It really shouldn't surprise anyone that Plumtree was only a temporary placeholder for Microsoft, prepping the marketplace at great cost and growing the customer recognition of how important portal technologies are.

    What makes today different from any time in the past is “not” that the DOJ might act to reign in Microsoft. That's a done deal. The DOJ has legally sanctioned Microsoft's ownership assertions over all Windows based opportunities. No, what's changed is that IBM's WorkPlace is a blueprint of how to use open source to break Microsoft's monopoly grip at a most important moment in time. A moment when all software and all software systems are moving to the Open Internet.

    It's the perfect moment for IBM to strike.

    It's also the perfect moment for Plumtree providers to make new claims regarding the vast herd of Windows users. Just follow the WorkPlace blueprint, and you too can cash in on an opportunity at Microsoft's expense. Because WorkPlace is based on open source components, anyone can come in and do the same thing. Instead of promoting an IBM server stack though, you might be promoting some kind of BEA – Plumtree – Oracle – SAP configuration. The one thing you shouldn't do though is dismiss WorkPlace. And don't make the mistake of seeing WorkPlace as an alternative to MS Office! The game is much bigger than that!

    WorkPlace is a freight train ready to leave the station. It's also a blue print for how anyone willing to work with highly interoperable open source components can wreck havoc with Microsoft's monopoly base.

    Understand also that you would not be alone in this effort to climb onto the WorkPlace train. Mepis and Ubunto are two Debian based Linux distros similarly aiming to ease the migration path by jumping onto Windows with their own open source versions of WorkPlace. Maybe someday even Sun will wake up and realize that WorkPlace was the play they should have made long ago. It is after all their StarOffice engineers, and their component model magic, that have made WorkPlace the wonder that it is. A classic triumph of engineering and failure of management if ever there was one.

    Let me also point out to you Jaime something you know all too well. Neither BEA or Plumtree, or Real, or Sun, or Apple or anyone else is ever going to tell the marketplace the truth about Microsoft's dominance. They can't tell the truth until it's to late. Any hint of Microsoft's dominance and underlying control over the fortunes of others would wreck disaster on investors and stock market pricing. So they remain hush until it's too late. What i'm begging you to understand is that WorkPlace will change that scenario completely. It doesn't mean you have to adopt WorkPlace to survive. Indeed, i don't think you can. What it does mean though is that once IBM freezes that MS OS-Desktop monopoly, it's open season for anyone who gets it. As for IBM? Well, they have a date with destiny quite different from anything the rest of us will ever face. I for one think they are ready.


  9. Gary. I agree with most of what you're saying but, is it possible that you never heard about Java Desktop System? I mean, Sun has a Linux Desktop in version 3 already.

  10. The Java Desktop is only available on Linux and Solaris. There isn't a Java Desktop for Windows, although for a while there i thought we would see one. At LinuxWorld 2004 i had a chance to speak with Sun's Tim Bray, Simon Phipps and Curtis Sasaki about a portable desktop productivity environment – a Java Desktop for Windows.

    With near 70% of OpenOffice.org and Mozilla users running on the many versions of Windows, and everyone of them concerned about a possible migration to Linux in the future, this seemed to be something that Sun might want to pay attention to. A month later, in a Jonathan Schwartz blog, there did appear a brief blip of sunlight, when he mentioned that the Java Desktop would be available for Linux, Solaris and Windows. The world is still waiting though.

    Well, let me rephrase that. The world no longer needs to wait for Sun to come up with a bridge from Windows to Linux. Some of us will have WorkPlace.

    One of the more brilliant aspects of WorkPlace is that the productivity environment is wrapped in an Activity Manager – Project Manager purpose. While you could replace MS Office 12, IE, OutLook, VB.NET, Access, and InfoPath with WorkPlace, the unique activity manager bent of WorkPlace pushes the collaborative computing envelope out so far as to make it a necessary MS Office accoutrement. Especially if your dealing with an IBM stack of servers. Even though the core of the Java Desktop productivity environment is the same as WorkPlace, the difference between the two is night and day.

    My discussion with Simon, Tim, and Curtis had to do with migration strategies. Of providing end users with a desktop interface that could quietly go about the business of transitioning information and information processes into portable file formats (OpenDoc), and portable run time engines (Java, Python, MONO, JavaScript & XUL, etc.). With the interface, the information, and key information process in a portable condition, when it comes time to purchase a new computer businesses will find that they've finally broken the Windows platform lock and can purchase equipment based on price and performance. Something like the Linux Java Desktop.

    IBM has a different approach to this. They're playing for stakes much higher than that of breaking the Windows platform lock (OS & Windows API) to assist in migrating users to some IBM version of Linux. IBM doesn't even want to be in the OS distro business. Nor do they seek with WorkPlace to push MS Office off the desktop. Although that might in fact happen given the fact that the advanced collaborative computing aspects of WorkPlace will run on most of the 450 million Windows desktops now in use – and the MS Office 12 collaborative computing features demand a near universal upgrade of all desktops and servers. Peaceful coexistence, or at least a non threatening appearance, is essential for the divide and conquer strategy to succeed.

    End game for IBM is breaking the WinTel lock and clearing the way for the next generation of Power PC chips. Of course there is still the need to migrate information and information processes, and that means first getting these needs into a portable state without disrupting the user interface to the productivity environment and the many information systems beyond. WorkPlace can do this. Importantly though, WorkPlace accomplishes this in a way that dramatically increases the value of IBM's server side stack, while at the same time providing IBM's Global Services Group with the kind of glue and stitch magic every other solutions provider and integration specialist would kill to have in their quiver.

    In this context, even if there was a Windows based Java Desktop, Sun's migration opportunities, while important and very beneficial to them, would still be small potatoes compared to what IBM has cooking with WorkPlace.

    Sadly, Sun has other problems more pressing.

    In the aftermath of the disastrous anti trust settlement, the Solaris guys at Sun stepped up and forced out all the Open Internet and Linux supporters. They are now driving the company, and Solarian hubris is about as dense as it gets. But who knows? Maybe the Solarians will drive the company through the fog and come out on the other side with an OpenSolaris that runs on the Power PC, uses ZEN to seamlessly host and run all those Linux applications, even the many portable run time engines sprouting up all over the place, and then takes NoMachine to the next level. Maybe when all that happens we might finally see a Java Desktop for Windows?


  11. Gary. my former comment was made because it looked that you didn’t knew about Sun’s Linux offer.
    Of course I don’t agree with much of what you said this time (specially the last paragraph, of course) but, since this is Stephen’s Blog and the discussion is about BEA, not client desktops, I think we should stop 🙂

  12. Hi Jaime,

    Agreed. Although i might add that you and i have much more in common than you think. We both want Sun to succeed. And not just to succeed, but to profit mightily from their open source and open standards efforts.

    We would also both agree that there is a problem.

    I want Sun to succeed because i'm a long time member of the OpenOffice.org community, and Sun has been a magnificent benefactor. The OOo community owes everything the incredible engineering, patience, and genuinely cooperative attitude Sun's engineers bring to the daily grind of building and maintaining a thriving open source community. I know first hand how difficult and resource intensive it was for the engineers in Hamburg to convert the megalithic StarOffice application suite into the highly interoperable UNO component model we see today. While the Sun engineers are doing this incredibly innovative but yeoman work, IBM was busy weaving the discreet components into an entirely new next generation collaborative computing interface that bridges local information resources and server side resources into a magical middle space canvas where users aggregate and collaboratively work the information and information processes critical to their digital lives. (Sorry, but I write god awful long non sentences when i'm pissed, and Sun continues to disappoint me).

    So Sun does the work of both engineering and community building, and IBM benefits. If you ask the Sun OOo engineers about this they will say that WorkPlace is a perfect example of exactly why Sun chose to open source StarOffice under the LGPL. And it is true that IBM deserves a tremendous amount of credit for the imaginative reconfiguration of OOo – Mozilla components. The fact that IBM has reinvented the Open Internet based collaboration model using Open Source components and Open Standards should excite Sun to jump on board and up the ante. Certainly Sun has the expertise to blow the lid off the WorkPlace model, and i for one want to see them do this because they have been a such great benefactor. Besides, where would OOo be without them?

    WorkPlace is not “just” a desktop solution. It's an Open Internet collaborative computing solution. It just so happens that most Open Internet users come into the collaborative space with information and information processing capabilities based on a desktop interface. I fully expect the “WorkPlace work space” to be device independent, able to arbitrate connectivity, synchronize information, and maintain state.

    The reason the collaborative space is relevant to BEA, Plumtree and anyone else in the portal – messaging – application server arena is that this represents a serious shift in the browser/server model towards empowering the end user. C-Space is a nebulous area where users can aggregate local, server realm, and server side services according to “their” activities and projects. Portals on the other hand put the aggregation and available information tools under the control and vision of server side programmers. An end users view of needed activities could include the services of many portals and server side information systems (realms).

    Web services and streaming web services (XMLHttpRequest) are only part of the story of how server side information objects can be be summoned and aggregated as activity components. Dynamic URL aggregations from Google, Yahoo, Amazon, and eBay services can similarly be joined to the activities canvas.

    In some ways the C-Space phenomenon is a variation of what Clay Shirky, in his essay “PC's Are The Dark Matter of the Internet”, has called the integration of the PC's latent computing power into the fabric of the Internet. What makes the C-Space phenom truly extraordinary is that the model is emerging just as the methodology for the break up and broadcast of info chunks on the Portal – Web Server – information realm side of the equation is accelerating. Both sides of the Web equation, local and server, are able to feed the C-Space.

    Some early portal efforts pioneered this effect. Both Aetna and ING were offering extraordinary portal based aggregation services, where users could gather personal financial information from any banking, insurance, and trading accounts where users had access rights. The aggregation of accounts gave users a first time ever holistic view of their net worth. There just wasn't a whole lot users could do with that view other than to weep or shout and refresh for another emotional round. C-Space initiatives change all of that. Collaborative computing is interacting with people and information with the latent but also interactive computing power of both your local resources and the global info grid.

    Another important aspect of the C-Space phenom is what Google is doing. As Sam Ruby points out in his essay, “REST + SOAP”, “Google is a service. One which permits parameters to be encoded on the URL.“ For a C-Space, they not only get to aggregate information from information realms, they also can embrace and interact with services from Open Internet aggregators. A dynamic Google aggregation of either information or interactions can be joined into an activity management view that includes privileged information from carefully controlled server realms.

    I believe that Sun could out WorkPlace IBM, and do so with a loosely coupled C-Space that ups the participation ante far beyond the IBM server stack. What JotSpot, Zambra, and Tachometry are doing with C-Space is very interesting. What they all lack at this point though is that ability to engage the local productivity environment and perfect the synchronization – XML transformations that Sun perfected with OpenOffice.org, and IBM put to use with WorkPlace.

    As for my comments about the Solarian coup at Sun? Yes, i have lots to say about that. I maintain to this day that Sun would have been better off embracing Linux at the low end of the market, commoditizing the hell out of everyone else. And reserving Solaris for a high end deployment where reliability is at a premium and failure simply isn't an option. The biggest problem Sun faces is that Oracle on Red Hat Linux is seriously out performing Oracle on Solaris. How open sourcing Solaris changes that is beyond me. And unless or until Sun can change that particular performance differential, the open sourcing of Solaris looks as much or more like a public escrow of the code base to offset customer concerns of default as anything else. Sadly Jonathan Schwartz's OSBC keynote attack on the GPL did more to harden this perception than anyone at Sun can imagine.

    I sincerely hope Jaime that you will continue to defend Sun where they deserve to be defended, and jump their bones when they waiver, mis step, miss the opportunity, or just plain step in it. For my part i will continue to insist that Sun is the single most reliable corporate defender of Open Standards, the Open Internet, and Open XML technologies. I will continue to point out that Sun has been a magnificent benefactor, provider, and supporter of both OpenOffice.org and Mozilla. Now, if they can only get their act together and get a handle on the leadership horns of the Web 2.0, horns that rightfully belongs to them, instead of being gored. Haplessly gored again and again.


Leave a Reply

Your email address will not be published. Required fields are marked *