CMS Watch Sun Feed

Cloud Computing and Content Management

aps@cmswatch.com(Alan Pelz-Sharpe) — Fri, 30 May 2008 14:43:00 -0400

If there is a buzz around Web 2.0 in the Content Technology community, then there is a roar in the wider IT community around Cloud Computing. It's a great term, "Cloud Computing," as it conjures up visions of an invisible Internet -- an ether-like zone in the sky where computing power and storage is unfettered by the petty restrictions of boxes, cables, and technicians. Cloud computing sounds fluffy, it sounds cool, it sounds limitless, it sounds like the future.

In fact Cloud Computing simply means moving things to big and bigger Data Centers. Data Centers are anything but fluffy. They are huge, energy-sucking giants -- many the size of small towns. They are environmental disasters and the only thing fluffy about them is the C0₂ emissions they belch out. Data Centers will in time according to The Uptime Institute become bigger polluters than the aviation industry. Data Centers require massive amounts of energy to operate -- often as much energy is used to cool the centers as to power them. All that heat has to go somewhere. If you think your air conditioning unit is an ecological no-no, then consider the AC demands on a data center the size of 5 football fields, then consider further that according to market research firm IDC, there are over 7,000 major data centers worldwide, and many more in the process of being built. By the way, just because they are big does not make them efficient; it is estimated that around 1/3rd of Data Center servers continually sit idle.

Yet there is very vocal group that thinks the Cloud is the future of ECM and Archiving -- so, why bother to dispose of data? Why not just send it to the "Cloud" with its limitless processing and storage capacity? The people who run the Data Centers, and those that sell equipment to run them such as Sun Microsystems and EMC, would all think that a good idea.

What puzzles me though is the contradiction between good Archiving and ECM practices versus a demand for ever greater processing and storage capacity. Aside from the fact that Cloud computing by definition means that your data has to move to and from a distant location, and latency issues (and simple physics) will always dictate that this will be slower than on-site deployments, there are much more important issues to consider.

For if you only stop for a moment you will see that there is something fundamentally wrong with the Archiving/ECM equation. Take E-mail Archiving for example. Almost all EAM systems promise (and typically deliver) a reduction in storage capacity needs of about 80%. Good e-mail management can reduce that still further, till ultimately your e-mail mountain has dropped to less than 10% of its original size. In other words the efficient management of content can reduce server farms to single servers, with active archiving and disposal of redundant data keeping the volumes from growing exponentially. Now that is not only Green Computing, it's smart computing, because in reducing the data volumes, you increase the processing speeds by as much more than an order of magnitude and you leave a tiny carbon footprint. So much better than the SUV -- sorry, "Cloud" -- approach to computing.

Sun Portal Server rides into the sunset in favor of Liferay

info@jboye.dk(Janus Boye) — Sun, 25 May 2008 04:52:00 -0400

In a bold move Sun used the JavaOne conference earlier this month to announce that it will begin to work closely together with Liferay on next-generation web technologies.

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.

Sun to pursue Java-less Java?

kthomas@cmswatch.com(Kas Thomas) — Fri, 29 Feb 2008 08:03:00 -0500

Sun Microsystems CEO Jonathan Schwartz, speaking at the SugarCRM Customer and Developer Conference earlier this month, made a pronouncement that went largely unnoticed by industry pundits. "I think what you'll see from Sun," Schwartz remarked, "is that we're just going to take the 'J' off the 'JVM' and just make it a 'VM'."

Wait. Roll the tape back. Did Schwartz just say that he wants to factor Java out of the JVM?

If so, it could have important ramifications for the Web CMS and Portal marketplaces in particular.

At first blush, it sounds like a radical notion. But it's not. Microsoft pioneered the idea of a language-neutral virtual machine years ago with its Common Language Infrastructure specification, also known as ECMA-335. The CLI spec provides the basis for the .NET virtual machine (which does indeed support multiple languages).

The Java community's comeback is something called the Da Vinci Machine project, which has an explicit goal of "extending the JVM with first-class architectural support for languages other than Java, especially dynamic languages."

So now Sun wants to out-dot-NET dot-NET. But that's something Sun could have attempted years ago. Why do it now? No doubt the groundswell of interest in dynamic languages (Ruby, Perl, Python, Groovy, and so on), driven by Web 2.0, has awakened Sun to the realization that programmers now want the best of both worlds: They want the rapid development that dynamic languages afford, yet they also want the very real benefits (in terms of thread management, garbage collection, security sandboxing, memory management, platform independence) of running inside a VM.

But still, why would Sun want to go in the direction of a Java-less JVM if it didn't have to?

It turns out there's an elephant in the room, and its name is Adobe. It's no secret that Flex (Adobe's answer to Web 2.0 development) is gaining traction every day, and Flex outputs bytecode for the Adobe VM. Close observers of the Mozilla ecosystem have known for a while that Adobe's VM (a.k.a. Tamarin) will soon be an integral part of Firefox. And lest there be any doubt about what's going on, Adobe CTO Kevin Lynch admitted at an industry event last Monday that Adobe is working on a project to allow C, C++, and other languages to run atop the AVM. Let's recap: Adobe is in the "Java-less" VM business. Big time.

As it turns out, the AVM is actually fairly mature (it has powered several generations of Flash), not to mention fast (it supports just-in-time compilation), and it already powers a sizable percentage of the world's (ahem) flashiest Web 2.0 applications. It will soon be built into Firefox's scripting engine, and if Adobe has its way, it will be on every PC user's AIR-conditioned desktop faster than you can say "Silverlight."

Sun clearly has its work cut out for it, if it wants to remain relevant in the VM wars.

Regardless, unshackling Java from the JVM (or vice-versa) can only be A Good Thing in the long run. It means programmers will (at last) be free to take a best-of-breed approach in choosing the right language for the job. (Bytecode, after all, is bytecode. Who cares how you produce it?) Most of all, it will mean rapid application development, finally, for server apps that will run on the JVM, something most of us have been wanting since 1995.

Why does all this matter for the tools we follow? Well, we've been seeing a long-term trend among packaged applications away from lighter languages towards more "enterprise-ready" Java (c.f., Nuxeo, Interwoven, Vignette, et. al.). But this transition has been a difficult one for vendors and customers alike. Perhaps you'll see a renewed investment in dynamic languages across various content technologies. We'll be watching.

Enterprise Portal Marketplace: 2008 Vendor Risk Profile

jb@boyeit.dk(Janus Boye) — Thu, 17 Jan 2008 00:01:00 -0500

It is all too easy to identify vendors for your shortlist based on their supposed "leadership" status in the market. But CMS Watch contributing analyst Janus Boye argues that CIOs, procurement officers, and other technology leaders considering investments in enterprise portals should carefully examine the risk profile of prospective vendors to help identify the right "fit" for their needs.

Uncle Sam cracks down on vendor selection abuse

aps@cmswatch.com(Alan Pelz-Sharpe) — Wed, 5 Sep 2007 09:21:00 -0400

While nattering to a U.S. attorney while traveling back from Kuala Lumpur last week I was directed to this ongoing story. You can read all about it here, but in short the U.S. Department of Justice is suing Accenture for allegedly receiving kickback-like payments from technology suppliers it recommended and/or implemented at DOJ. The alleged fraud was a collusion with big-name IT suppliers (e.g., HP, Sun) and smaller vendors (e.g., Vignette) to defraud the Government.

The relationships between consultancies (especially big systems integrators), vendors, and buyers has always been murky. In this industry it is common to see an expensive product selection phase of a project (undertaken by a systems integrator) ending up with a recommendation of the system integrator's favorite vendor partner. To be sure, sometimes this is because the consultancy knows that tool the best, and in the end, a good partnership between software supplier and implementer can be critical to project success.

But as Matthew Clapp pointed out several years ago in these pages, very often money changes hands (from vendor to consultant), and that will affect selection decisions for sure. Says Clapp, "In fact, my experience found that more times than not, the Big 5 partner already knew exactly which CMS solution she was going to pitch before she walked through the client's door."

It is also a common lament of vendors who do not wield the clout of the major systems integrators (SIs) that SIs will not even meet with them to review their product offerings. At a time when customers are clamoring for simpler solutions across the board, non-enterprisey vendors often cannot get so much as a toe in the door.

What has this to do with ECM? Simple, be careful where you obtain your product selection advice. Make sure it is independent, and think twice about paying firms (SI, Consultancy, or Analyst) with relationships to particular vendors to provide you with "independent advice."

OpenPortal: A new open source project from Sun

info@jboye.dk(Janus Boye) — Wed, 30 May 2007 00:01:00 -0400

Almost exactly a year after open- sourcing its previously commercial portal product, Sun has decided to make a renewed push with its launch of "OpenPortal." OpenPortal is a new portal community where you can find all the source code to Sun Portal. OpenPortal remains a young project, but the portal package is indeed evolving. According to a Sun blog entry a new CMS/DM infrastructure will be introduced soon, based on the Apache Jackrabbit repository. As readers of the Enterprise Portal Report know, Sun Portal has thusfar included FatWire Spark as its default Web CMS. Perhaps this will change over time as Sun beefs up its own repository and workflow capabilities. This screencast from Sun offers some nice details about how it all works. More interesting for us is its emphasis on developer power -- so common to open source portal projects -- at a time when many commercial competitors (e.g., BEA AquaLogic, Vignette, and Oracle OWS) are pushing business-user configurability. [Update, June 3: Just to clarify: A Sun employee blog, interpreted this posting as complimenting the forthcoming CMS features. New customers should be aware that the planned CMS offering will not only be untested, but also likely offer fewer features than FatWire Spark.]

Using AJAX with portlets

info@jboye.dk(Janus Boye) — Fri, 2 Feb 2007 03:35:00 -0500

An interesting technical article on the Sun Developer Network provides helpful tips on applying AJAX to JSR 168 portlets. The article concludes that, "Indisputably, limitations and caveats exist in programming AJAX with respect to JSR 168 portlets," and offers a sample that bundles the Dojo Toolkit. Together with most other vendors, Sun is also behind the new and delayed portlet standard, a.k.a. JSR 286, which according to the article "is evaluating solutions to better support portlets that make asynchronous calls with AJAX." As readers of the recently updated Enterprise Portals Report know, the presentation layer should be one of the first things you test -- and implement -- in any portal project.

New IE7 Shakes Up CMS and Portal Implementations

tbyrne@cmswatch.com; jboye@cmswatch.com(Tony Byrne and Janus Boye) — Tue, 31 Oct 2006 16:19:00 -0500

Is your CMS or Portal vendor ready for IE7? As your read this, Microsoft is updating PCs around the world, but Tony Byrne and Janus Boye point out that support for the new browser varies across the content technology marketplace. Whatever your application, you likely have some important testing to do...

A new standard for portlet repositories

info@jboye.dk(Janus Boye) — Wed, 18 Oct 2006 08:49:00 -0400

The enterprise portal teams at JBoss and Sun have jointly announced the start of a new protocol for communicating with portlet repositories. According to the JBoss blog posting, "the idea for a standard repository protocol came after discussions with Sun over the interoperability of disparate portlet repositories with many portal vendors (as you know, they also have a portlet repository), and how we could offer a standard medium of communication between all players involved." As readers of the Enterprise Portals Report know, most vendors have their very own repositories to store portlets. These can be quite a mixed bag. The Portlet Repository Protocol standard intends to simplify the often underestimated and very time-consuming task of portal administration. If successful, the standard should enable administrators to make use of portlets from a myriad of different repositories from within one portal.

Most commercial portal vendors behind new portlet standard

info@jboye.dk(Janus Boye) — Fri, 11 Aug 2006 11:34:00 -0400

The new Portlet Specification 2.0, aka JSR 286 was just released as first public draft. The standard is currently supported by major portal vendors BEA, IBM, Oracle, SAP, Sun and Vignette. The main goal of this new version is to align the Java Portlet Specification with J2EE 1.4 and other JSRs relevant for portlet programming, and enhance caching support. The current specification, JSR 168, has been very important for portal developers. Unlike the CMS marketplace, the portal marketplace has benefited from some important standards, even if some vendors have developed proprietary "extensions." For buyers this one means that portlets developed to the spec work in all the portal products that support the standard, potentially reducing implementation time and making it easier to migrate between products. Portlet specs also enable you to potentially use re-use code across multiple different portal product installations -- a common scenario in large enterprises.

Portal Kudos and Shortcomings -- Summer, 2006

jb@boyeit.dk(Janus Boye) — Thu, 27 Jul 2006 15:24:00 -0400

Segmenting the portal software marketplace by putting products into boxes on charts is a popular exercise among pundits. But CMS Watch contributing analyst Janus Boye argues that for buyers, a meaningful vendor breakdown must describe how well the various offerings fit actual requirements across specific business scenarios. See how Janus compares the major portal products in the marketplace today...

A closer look at Sun Portal 7.0

info@jboye.dk(Janus Boye) — Tue, 4 Jul 2006 17:48:00 -0400

In the latest issue of Intelligent Enterprise magazine, I put Sun Portal 7.0 to the test. As previously mentioned Sun Portal is available for free. The article is an excerpt from the Sun Portal review in the Enterprise Portals Report and contains a few words on first-mover hiccups and also recommendations for which business scenarios the product might be a good fit and when it might be less usable.

Sun Portal goes open source

info@jboye.dk(Janus Boye) — Wed, 17 May 2006 09:34:00 -0400

According to a Sun employee blog the company has decided to put Sun Portal into open source. A new open source portlet repository will be developed and seeded with JSR-168 compliant portlets. The new project has its own website, which should be interesting to follow. At JavaOne, the annual Java conference, Sun Microsystems CEO Jonathan Schwartz promised to make Java open-source code. "It's not a question of whether, but a question of how." Other commentators have pointed out that key questions remain unanswered. As readers of the Enterprise Portals Report know, Sun Portal (which is already free of charge) has seen comparatively limited adoption worldwide and is somewhat feature-poor beyond collaboration services. Will a fresh brush of open source win over the masses?

Sun blog discloses issue with major portal release

info@jboye.dk(Janus Boye) — Mon, 10 Apr 2006 21:09:00 -0400

In late 2005, Sun announced version 7.0 of its portal platform. As with all new major software releases a few problems have been identified. Keeping inline with the "Web 2.0" focus of the release, a Sun employee recently blogged about a reasonably significant bug regarding portlet errors. The blog posting nicely documents some work-arounds to the problem. This communications approach is a welcome departure from the usual norm, where vendors only acknowledge issues late in the process and rarely document them well.

A more user-centered Sun?

tbyrne@cmswatch.com(Tony Byrne) — Thu, 23 Feb 2006 22:52:00 -0500

"I'm sure there must be a big rule written somewhere, that I can't seem to locate, that states that all enterprise software must have a less-than-optimal user experience, especially the interface." So concedes John Loicano, Sun's EVP of Software, who adds, "Features and functions don't have to be ugly." [Via Tim Bray.] Our forthcoming Enterprise Portals Report points out that Sun is doing some interesting things with its portal product, but portal usability across the industry still leaves much to be desired...

Web 2.0 focus for new free Sun Portal Server

info@jboye.dk(Janus Boye) — Fri, 16 Dec 2005 16:53:00 -0500

Sun has just released version 7.0 of its portal platform, "Sun Java System Portal Server." Sun is trying to embrace the so-called "Web 2.0" by in their own words offering the "first platform to integrate wikis and other next-generation platforms." And of course, AJAX is supported. Even though Sun is giving the portal away for free to customers of its other enterprise products, the package has heretofore seen limited adoption in a crowded market for portal solutions. It seems like the traditional Sun customer has so far either opted for a portal from commercial vendors like BEA or IBM, or popular open source offerings from Liferay and JBoss. Still, Sun's new, community-oriented positioning is a welcome departure from others in the portal market. But it will take more than a fresh coat of Web 2.0 to win over the masses.