When Red Hat revealed an updated roadmap for JBoss back in February, there was little mention of the enterprise portal, except a sentence about "an expansion to the Java portal engine" coming over the next 12 months. Evidently the JBoss Application Server, SOA Platform, and the Enterprise Data Services Platform are marked as higher priority at Red Hat.

As Enterprise Portals Report subscribers know, Red Hat has taken JBoss Portal 2.6 and changed name and version number to JBoss Enterprise Portal Platform 4.3. This has brought it in line with the Red Hat versioning scheme, but also creates the (mis-)impression that more things have happened with the portal than is actually the case.

To illustrate the low standing of JBoss Portal, I find it very interesting that JBoss.org has chosen open source Web CMS Magnolia for hosting more than 40 project sites with 20 million page views per month. If Red Hat wanted to eat its own dog food, they could have chosen to deliver content via their own portal platform instead.

Have you seen JBoss Enterprise Portal recently? If so, we'd like to hear from you...

In these changing times, I find it partcularly interesting to listen to honest presentations from practitioners that share lessons learned and perhaps even some unsolved challenges. In case you are interested, please contact me directly at jb@jboye.dk.

In any case, hope to see you there...

Lively was one of many services that came from the Google Labs environment. The service was released in July 2008 and introduced an innovative, 3-D social platform, where users could build their own interactive rooms and embed them into web pages. But only 4 months later Lively was discontinued, leaving the active user community in a sad state.

As a response to the sudden lock-down, the community has created an online petition, which many have signed, adding comments like "Lively has given me so much - please don't shut it down" and "Lively is more than a chat in 3D. [...] it is a world of feelings." All a little sad really, in so many ways...

Yet while the shut-down is indeed sad for the many users, it is also a timely reminder that

1. You need to be careful with any beta release.
2. Vendors -- including big vendors more often than not -- sometimes discontinue offerings

The case with Lively illustrates that Google is a business just like any other and that businesses need to make money. As they themselves write on their blog: "[...] we've also always accepted that when you take these kinds of risks not every bet is going to pay off. That's why [...] we've decided to shut Lively down at the end of the year."

In a future Google in the Enterprise Report, we will take a close look at Google based on our many on-going conversations with customers about their experiences. Thanks go to my colleague Peter Sejersen for his eagled eyed analysis of the Lively situation. And If you would like to contribute and share your thoughts, we would both still like to hear from you. Please contact us directly at jb@jboye.dk. I look forward to hearing from you!

To be more accurate, Google Gadgets are actually not a standard, but simply a Google specification. It is interesting though to note that portal vendors have added support for gadgets with some alacrity, including IBM in a royal wedding-like partnership in early 2007, JBoss in much hyped integration a few months later, and this week eXo as they released eXo Portal 2.5 with support of Google Gadgets.

Programming gadgets may turn into a very useful skill for 2009, but as Durga writes there are also non trivial issues to address if you want to roll them out for an enterprise solution. While researching for a new evaluation report called Google in the Enterprise, we've talked to quite a few enterprises that have experimented with gadgets. Not surprisingly, some of them were more gloomy than the vendor marketing, but interestingly gadgets have indeed been deployed for difficult problems. Few enterprises, though, seem to have considered the impact of using a proprietary alternative to standards, as most seem just excited to use something new and hot.

Here at the turn of the year, it feels a bit like the browser wars, where developers are busy writing for their own preferred platform, but not really considering the longer term impact. My advice to practitioners is that experimentation is good, and so is organizational learning. But if you want to avoid surprises, make sure to test and plan carefully.

According to the 10.3 release notes, Oracle WebLogic Portal 10.3 is really a very minor release. Support for WebLogic Server 10.3 and Java version 6 has been added. Oracle product naming and branding has been implemented and developers can now work with a new version of the Workshop IDE.

This may not sound like much and I doubt that it is enough evidence to comfort those worried buyers that have invested deeply in WebLogic Portal. To be fair, some other large vendors -- e.g., Microsoft and SAP -- have also done very little to their portal products during 2008.

While Oracle may have its own reasons for slowing down the pace on product improvements to WebLogic Portal, change is certainly still happening in the marketplace. Both from the open source vendors in the space, but also from vendors that don't call themselves portal vendors, e.g., Google with iGoogle and gadgets, Netvibes with widgets, Facebook and LinkedIn apps, and those hyped enterprise social software vendors. 2009 promises to be another interesting year in the portal market!

Percussion also has a CTO blog, where the most recent entry is a slight bit fresher, from July.

Perhaps in these times, no news is good news. However, I say that 6 months feels too long to be quiet. A big part of surviving in tough times is saving money, so it might make sense to cut down on press releases, updating the website, blogging -- but in this industry, you the buyer should carefully investigate any radio silence.

[Update, Nov 13: After reading this blog Percussion contacted us to provide an update on their activities over the last 5 months. Expect to see more from their side shortly]

Tracking the number of contributors can be a relevant indicator of the health of a product development community, but looking behind the numbers reveals a slightly  less positive impression.

With 85 employees, most of which are developers, the majority of the contributors happens to be on the eXo payroll.As I mentioned back in July eXo is in rapid growth mode and has many new hires. I would argue that if you deduct the eXo employees from the numbers, it would seem like there are very few contributors left from "the community," and as a buyer that is not necessarily a good thing.

With many government customers and a global footprint, I suspect that eXo, unlike other commercial open source vendors, has not been as badly hit by the current financial crisis.  But despite their enthusiam, their community has some way to go.

SharePoint 2007 SP2 is expected to be released between February and April 2009 and seems more focused on end-user functionality rather than administration and security, which were the prime domains for previous updates. Among the main areas for the upgrade are performance and manageability improvements to variations in including STSADM commands for repairing links between source and target pages.

The news was first announced on the Office Sustained Engineering blog and later a few more details were released on the Microsoft SharePoint Team Blog. What strikes me as missing based on conversations with the large SharePoint eco-system, including several MVPs, is no mention of improvements in the areas of accessibility and standards or enterprise search.

As our readers may remember, Microsoft has historically been very secretive about the SharePoint roadmap, and I am certain that it has taken many customer requests for Microsoft to become more open and transparent.

Still, I continue to recommend that all buyers keep asking questions directly of Microsoft, as they may be willing to share further details, potentially under a non-disclosure agreement. It's possible that a new feature or bug fix is set for release a few months out, and that could save some time and money on your implementation budget.

I would be really interested in hearing from any readers that might have some experience as customers.

Today Google offers many relevant enterprise tools, and I'm not even sure that Google has the full grasp on what they offer themselves.

I plan to focus on:

• Google as an e-mail vendor
• Google as a web development helper
• Google as an enterprise search vendor
• Google as a desktop suite vendor
• Google as an app. dev platform

If you're interested in sharing, our discussion would be entirely "on background," which means I would apply your experiences to my knowledgebase, but never indicate your name nor mention your institution in any publication.

Please contact me directly at jb@jboye.dk. Look forward to hearing from you.

As readers of the Enterprise Portals Report know, WebCenter is an impressive offering for developers, but unfortunately offers very few out-of-the-box services for business users. According to Oracle, WebCenter adoption is growing, and many of the new projects include integration to the vendor's "UCM" (formerly Stellent) offering for content management. System integrators are now beginning to build practice areas around WebCenter, as they go through initial project cycles.

If any existing BEA customers thought that Oracle would rest on the portal laurels, they clearly were wrong. Since the updated portal strategy was announced a few months ago, ALUI became a part of WebCenter Suite. As of this week, WebLogic Portal is now also a part of WebCenter Suite, so that new customers can mix and match among the four portals in Oracle's offering. Oracle claims that most BEA engineers have been retained at Oracle, which now has more than 25,000 developers on the payroll.

On the roadmap is the delayed 11g release, which is now scheduled for H1, 2009. 11g is currently in an on-going customer beta program. Among the themes for 11g are social computing, collaboration enhancements, and better standards support. Later in 2009, Oracle expects to introduce support for IBM WebSphere as an application platform, perhaps in a move to help its sales force sell WebCenter into traditional Big Blue accounts.

Finally, while SharePoint is indeed changing the portal landscape, many customers here in San Francisco and the very large Oracle ecosystem reminded me that there is still a thriving enterprise portal market outside (above?) SharePoint. I even met some who decided not to go through the difficult upgrade from SharePoint 2003 to SharePoint 2007, but instead decided to shift to Oracle WebCenter.

Other customers, meanwhile, are headed in the other dirction; see for example my commentary from earlier this week: Questioning Oracle's Portal Leadership.

With Oracle's acquisition of BEA, I was in particularly interested to hear about their latest enterprise portal developments. In the opening keynote by Charles Phillips, President and Chuck Rozwat, Executive Vice President, Product Development, they did make some relevant announcements:

• A new beta release of the much-discussed but only marginally-deployed WebCenter Suite (which has seen several delays), as the product gets moved to Oracle WebLogic as the default application server
• Fusion Middleware for Cloud Computing, based on Amazon Web Services
• Oracle Beehive, a new product for enterprise collaboration, that Oracle has built from scratch with integrated security (via the SealedMedia acquisition). According to Oracle, Beehive has already been adopted by a few customers, including the European Space Agency

Enthusiastic Oracle managers also claimed market leadership in many areas, including enterprise portals, something that IBM has traditionally claimed. I don't know how they come up with this, but I do urge buyers not to put too much emphasis on any vendor claiming such a mantle.

While this week is indeed exciting for everybody in the large Oracle community, it still seems like very few customers have adopted WebCenter Suite, Oracle's strategic portal platform for the past year at least. It certainly can't be WebCenter Suite that's "leading the market."

Also, in my interviews with users of Oracle Portal (the "old" portal), it seems like quite a few are upgrading to Microsoft SharePoint, instead of going for Oracle. Interesting times indeed!

According to Sun, new customers with community or social networking requirements will get steered towards WebSynergy rather than Sun Portal. An additional Stable Build 3 is slated for October and then Sun expects a v1.0 commercial release to follow in January 2009, with complete documentation and training.

The Sun initiative is worth watching, as Sun Portal is slowly fading into the background. In what appears like a step in the wrong direction, support for IBM WebSphere or Oracle WebLogic as application servers has been removed from WebSynergy (they were supported in Sun Portal), in lieu of GlassFish is now the only supported application platform. This only adds to the quite unnecessary Portal - SOA divide.

With our recent update to the Enterprise Portals Report we decided to drop regular coverage of Sun Portal as our research indicates that adoption has stagnated since the announcement back in May 2008. At some point in the future we might decide to cover WebSynergy (or whatever Sun calls the final product name), but for now we'll point you to our Liferay review as the nearest approximation of where Sun may go.

I'm very pleased to be talking on this exact topic in London on September 15 at a Henry Stewart conference on The Latest Thinking on Optimising Enterprise Search. In my talk I'll cover some of the usual pitfalls and also share some recent project experiences (both positive and less positive) with SharePoint's search functionality.

At the conference I'll be joined by a global team of noted enterprise search gurus including James Robertson, Adriaan Bloem and the Martin White. I'm also looking forward to two interesting case studies from the BBC. I'm sure I'll learn much during the day and hope to see you there!

As we mentioned in the Socialtext review in the Enterprise Social Software Report 2008, customers were concerned earlier this year over quite a few upgrade problems with the hosted Socialtext service. Recent reports from customers suggest that the popular service remains quite a bumpy ride. Here's a few of the issues that Socialtext customers have reported over the past few month:

• The one and only reporting feature in Socialtext -- weekly usage reports -- stopped working back in May
• The search feature was suddenly substantially changed. Now search results no longer sort by default according to last edited date, but by relevance. This may seem an improvement, but try a search on a public Socialtext wiki and you'll find that it still takes some time getting used to.
• In another incremental upgrade, international characters were transformed to garbage, e.g., the Danish "æ" became an "Ål" and "å" became "Å¥."

It is also worth adding that while Socialtext formally announced version 3 back in April, the hosted service is today still running version 2.22, almost 5 months later.

The larger issue here is that as a SaaS customer you remain very much in the hands of the vendor. Often that's a good thing ("our upgrades happen automatically"); sometimes it's a bad thing ("we can't chose to postpone an upgrade"). How well your SaaS vendor prepares you and performs proper regression testing may seriously impact your project. Enterprises don't like surprises, but it appears like Socialtext still has some way to go here

It may come as little surprise that an employee of Unicon, the dominant uPortal systems integrator, has good things to say about the product. By contrast, our research for the The Enterprise Portals Report 2008 finds uPortal is an unlikely fit for e-business and self-service portals. Let's briefly explore why.

As prelude, I'll note that the products we cover are among the most widely-used products in the portal industry. Consequently, almost all vendors can claim references across all scenarios.

Despite that, I'd first point out that just because uPortal has a few universities that use the product for self-service doesn't mean that the product compared to other enterprise portals, such as Vignette, is a good fit for this scenario.

Second, we find that uPortal isn't particularly rich in out-of-the-box functionality for self-service portals (below, I mention some of the features you should look for and test for this scenario). Together with Vignette, which has strengths in this area, Oracle also provides you more to get you up and running with self-service. Among the open source projects, even JBoss Portal and Plone has more to offer than uPortal with regards to self-service.

Third, self-service portals come in many flavors. The way you define self-service may not be exactly the same as I define it, or Unicon does, but there are likely a few common requirements we would all have. Good reporting, for example, is typically a key requirement, as is integration to existing repositories, such as those used for customer data. In our research we found users struggling with uPortal integration, and citing very little beyond technical and basic reporting out-of-the-box. Neither integration nor reporting was among the major improvements in the recently released uPortal 3.

Always remember, when evaluating which product to use, be sure to rely on matching your key requirements and use case scenarios to what the different products have to offer, and be sure to test with your own content. That's always more important than impressive references, however relevant they may be to your case.

In a recent blog posting by Sitecore's VP of Technical Marketing, Lars Nielsen, he discusses their upgrade strategy and explains the company's choice between delaying the release of Sitecore 6 or let the database conversion tool follow afterwards. Similar to many other vendors in this situation, e.g., Microsoft, Sitecore decided to get the new product out the door and worry about upgrades later.

The definition of immediately afterwards may extend beyond the 2 months that have transpired since V6 came out, but I see that Sitecore themselves have still not upgraded their very own website. According to Sitecore, an alpha release of the upgrade tool is expected this week, but there is no news on when customers can expect a final release.

Regardless of vendor, upgrades are never straightforward, and you typically want to wait until the vendor has gone through the pain itself before teaching them the ropes. In this case, though, it is telling that Sitecore -- a vendor with a support model that we have previously questioned -- has focused more on pleasing new prospective customers and less critical analysts alike with exciting new demos rather than supporting its faithful customers. If the past is any guide, do remember to budget and plan any upgrade carefully.

Then last week eXo announced a new branch in Tunisia covering the emerging (but also largely underresearched) marketplace in Northern Africa.

When we started covering eXo back in 2006 the firm had a 27 employees and had just opened a US office. Today they have 70 developers have joined the growing ranks of open source projects trying to offer an alternative to Microsoft SharePoint.

eXo has also been an early adopter of Adobe Flex, which is actively used in several of the eXo products, including the beta version of the new "Liveroom" video conferencing component.

I don't pick favorites and am not saying that eXo is the "best" or "leading" enterprise portal, but in these times where the portal market is increasingly dominated by large vendors (IBM, Microsoft, Oracle) it is important to remember that the open source portal market may indeed offer you viable alternatives. Beyond eXo we also cover Apache Jetspeed, JBoss, Liferay, and Plone.

Combined under a new Oracle-specific definition of Enterprise 2.0, I'll highlight these two statements:

• Oracle WebCenter Suite and Oracle WebCenter Services are Oracle's strategic solutions for developing Enterprise 2.0 enabled portals, composite, and web applications.
  
• Oracle plans to continue to develop and support Oracle WebLogic Portal and Oracle Portal, and expects to converge these products with Oracle's strategic solutions over time
  

The former enterprise portals from BEA have now been renamed as Oracle WebLogic Portal and Oracle WebCenter Interaction (previously known as ALUI). For now Oracle Portal, WebLogic Portal and AquaLogic Portal are officially categorized as "continue and converge," which means that new customers are not steered in this direction, but support is maintained for 9 years.

Despite an apparent flip-flop from Oracle, the news comes as little surprise. Back in May 2007 I cautioned buyers that Oracle was shifting portals and placing WebCenter at the forefront with Oracle Portal edging retirement. Now that Oracle is being more open about the future of their portal products, new prospective Oracle customers should consider WebCenter, while existing licensees of the other three portals should already now build migration -- in some cases really outright replacement -- into their project plans and budgets.

[Update, July 9th: Clarified positioning of portal products, including ALUI)

As a follow-up to the widely-adopted JSR 168, this portlet specification 2.0 moves to make portals more like integrated apps and less like collections of disconnected windows. Specifically it adds support for events, public render parameters, resource serving, and a portlet filter.

Some vendors like eXo, IBM, JBoss and Liferay have already been supporting earlier iterations of the standard and two years ago, I commented that most commercial portal vendors are behind this new portlet standard. While this is still the case, many significant changes have happened in the marketplace since the initial draft of JSR 286 in August 2006.

Jason E. Shao from the CampusEAI Consortium asks in a blog whether the next generation portlet specification really matters and over at the TheServerSide.COM you can find a healthy discussion on the final spec release.

Standards generally go missing in this marketplace, but judging from the very limited attention this new version of the portlet spec has received, it makes me wonder whether the marketplace has already left the need for it in the dust. As a buyer the new industry standard might seem the preferred option over the many proprietary implementations that build on the shortcomings of JSR 168, but make sure to study the emerging implementations of the new standard carefully to avoid an early mover disadvantage.

The story is, a Denmark-based Oracle partner had posted something "not unambiguously positive about Oracle WebCenter" and was "immediately flamed by an Oracle product manager, and any trace of negativity edited out."

Of course, wiki owners always have a choice about how strongly they want to moderate, and that choice naturally will affect the culture of participation. Oracle clearly decided in favor of strong moderation. The Rules of Conduct on the wiki say nothing about excising mentions of product weaknesses, but WebCenter is certainly a hot topic at the moment with customers waiting for roadmap details after the BEA acquisition.

So, official wikis present a dilemma for vendors. Clearly Oracle made the edits in order to keep damaging material out of competitors' hands. Other vendors, like Microsoft, also limit who can post to their public wikis based on an application process. That's their prerogative. But as a customer, you should understand the rules of engagement are rather different than, say, Wikipedia, and evaluate the content accordingly.

(Thanks to Sten Vesterli from Oracle partner Scott/Tiger for the heads up.)

This is an interesting remark:

• Many customers find themselves with a combination of BEA and Oracle portals. Which one should they migrate to?
• While the 4 products overlap to a large extent, there are also significant differences between them, e.g., WebLogic Portal versus BEA AquaLogic User Interaction. Should a new or existing WebLogic Portal customer really adopt the product for all their portal requirements?

I had the pleasure of sitting on the panel and used the opportunity to share my open question that WebLogic Portal might have had its final major release. The panel took an unexpected turn when the IBM representative took the microphone to defend Oracle, despite the fact the IBM has WebSphere Portal as their only portal product.

There are different ways to segment the portal market, but I still favor scenario-based analysis. While Oracle may have a track record for sustaining products they acquire, at the end of the day, license money from customers will decide which products survive. For those considering buying an enterprise portal from Oracle, I would recommend taking a closer look at WebCenter, which seems to be positioned as the product of the future.

It took me a couple of weeks to digest the news and distill what the press release did not spell out. Really what will happen is that Sun will take a snapshot of the Liferay Portal code and use this to create a Sun-branded portal with added functionality. An initial version is expected in late 2008 or early 2009.

This is pretty big news because Sun already has a portal offering, which now will go away. The current release of Sun Portal is 7.2 and customers should not expect a Sun Portal Server 8. Sun says it will provide some level of migration tool for existing Sun Portal customers with the initial release and more will come down the road (e.g., by a 1.1 or 1.2 release).

For Liferay this is clearly a great commitment to their open source platform, which has risen in popularity in recent years. Those presently considering Sun Portal should already today take a closer look at Liferay, to avoid future migration costs. As readers of the Enterprise Portals Report know, the two products overlap substantially and share many similar strengths and weaknesses. Liferay Portal recently released version 5 and promises version 5.1 in late June, with several improvements to enterprise features, e.g. workflow and global distribution of portlets. I would certainly expect an even stronger enterprise focus in the Liferay Portal roadmap after the Sun announcement.

While the announcement may not impact Oracle in any way, it is interesting to note that while Sun effectively is taking their product out of the market Oracle still maintains its position with 4 enterprise portals.

As an analyst I'm frequently asked about license prices. A recent interesting discussion among peers challenged my views and provided helpful feedback that might assist you in arriving at the right numbers in today's marketplace:

List prices aside, buyers can presently obtain significant discounts on enterprise portals and on Web CMS tools. This may be caused by the increased SharePoint infiltration. A commentary in February on big software discounts and recent numbers from Vignette seems to confirm this trend. SharePoint licensing for websites is the exception that proves the rule. In general if the Web CMS comes from an ECM vendor, it will be more expensive -- potentially way more expensive

With enterprise search at the high end, the reverse is true. The marketplace is seeing strong demand at the moment. Many enterprise-tier search offerings come only as a bundled offering, so there is little list pricing to benchmark against. Deals quickly run into the millions of Euros in large, global, and complex enterprises.

Among the huge array of mid-market vendors across different content technologies -- many them local/regional in footprint -- you can typically find solutions that meet the needs of even organization-wide deployments in most enterprises, but at a factor of five (or more) cheaper than the higher-end solutions

If you are willing to serve as a reference client or appear on the customer list -- or better within a press release -- this is very valuable for the vendor and should help you to get significant discounts. (And of course as you look to evaluate vendors and they provide such testimonials, you should also understand how this game is played.)

Remember that enterprise deals entail complex negotiation and pricing models that ultimately boil down to what the salesperson thinks you can afford. Perhaps needless to say, but still: Implementation costs are higher than licensing costs and open source projects are not necessarily cheaper just because you might save licensing costs.

Thanks to Alan Pelz-Sharpe, Apoorv Durga, James Robertson and Martin White.

Version 3.0 is mainly a technology release, but also ships with a fresher user interface and updated default content for better demonstrations. On the technology side the product now has improved portlet support (ready for JSR 286), a new unified caching framework as well as it has migrated to using the Spring development framework.

As readers of the Enterprise Portals Report know, uPortal is comparatively feature-thin, and its platform-like complexity sometimes comes as a surprise to developers expecting a simpler product. To facilitate the upgrade for existing adopters, uPortal ships with a wide set of import/export scripts, but as always make sure to test carefully before taking the plunge...