Say Goodbye to Portal Servers?
Got a change to go through Forrester’s article "Say Goodbye To Portal Servers." Here is the summary of report -
Do you think that counting open source portals in pure-play category is correct?
Vendors like Plumtree Software and Epicentric created the portal server market in the late 1990s by offering servers with technical features that just didn't exist within firms' existing IT infrastructure. Now those features — like UI abstraction, integration, workflow, and delegated administration — have been co-opted, improved, and embedded in general-purpose infrastructure platforms from vendors like IBM, BEA Systems, Oracle, and Microsoft. The standalone portal server market is gone, absorbed into infrastructure vendors' app server platforms and emerging interaction platforms.Conclusion looks OK that infrastructure vendors are adding portal features like UI abstraction, integration and workflow in their suite. But I don't think that pure play portals are out. The major missing point in this report is the exclusion of open source portals which are pureplay portals. As always, Forrester guys forgot to include Open Source portals and hence the final conclusion they have drawn it not IMHO a valid one. I personally want to classify open source portals like Liferay and eXo as Pure Play portals than infrastructure portal. Though the pure play portals like Plumtree and Epicentric are out, but Open Source portals continue to have its market share. A survey in my portlets community suggests that Pure Play portals are still having more market share than the infrastructure vendors (again note that I am counting Open Source Portals as pure play portals). Keeping IBM, BEA, Oracle, SAP, Microsoft & Plumtree (as it is acquired by BEA) votes in Infrastructure category and eXo, Liferay, Jetspeed, uPortal and Vignette votes in Pure Play category here is the picture –
Do you think that counting open source portals in pure-play category is correct?
Forrester always skips the open source projects- because the open source guys don't buy expensive research papers or sponsor conferences.
ReplyDeleteThey've become notorious for this in other fields- look at their coverage of web agencies for a real eye-opener. The firm I used to work for was black-balled by them because we didn't co-sponsor an event.
I think the idea behind pure plays being dead is that Portal features are becoming commoditized. There are becoming part of collaboration (Oracle's 10 G, for example calls itself as collaboration platform) or something else. I think players who can provide something more than just portal features will be leaders. So either infrastructure vendors like IBM, Microsoft etc or vendors like Vignette, EMC etc who provide portal features along with content management, document management and so on will be the ones at the top. In that sense, I donot consider Vignette as pure play.
ReplyDeleteI'd posted similar thoughts here: http://www.apoorv.info/index.php/2005/08/25/another-stand-alone-portal-vendor-dead/ and http://www.apoorv.info/index.php/2005/06/14/portal-servers-a-dying-breed/
BR,
/a
http://apoorv.info
I bet if you were to ask an enterprise architect that works for a Fortune 100 enterprise his perspective on whether pure play portals are dead and the response may be very different :-)
ReplyDeleteI think the answer lies somewhere in the middle in that pure play commercial is dead but plays such as Liferay will continue to provide the value we all seek.
FYI. I agree with previous comment that industry analysts need to do a better job of covering open source along with commercial product offerings...
The Vignette Application Portal, (aka VAP) portal is still a pure-play portal product (originally known as Epicentric). The original and continued top product of the Vignette company is their CMS engine. They are doing a good job integrating it with Vignette but a winning feature for them is that because Vignette has stayed as a pure-play portal product, integration with it also completes the majority of integration with most other standards-based portal products (adhering to JSR 168 is the most important for this integration).
ReplyDeleteAnyway, there's something to be said for the value added features that wwwbot refers to, but even in some of the larger companies (where I have been consulting), there is a lot of interest in simply using a non-cms portal to wrapper and integrate to their extensive back-end legacy systems, standardize their security (authorization and authentication), and provide a consistent look and feel. This is certainly what the open-source pure-play portal servers offer, and I'm seeing a decent amount of traction, EVEN when the infrastructure products are owned and licensed already. This is due primairly to the complexity of some of those products. And when it comes down to it, changing will probably save these companies tons of man-hours.
If you want or need the features of the infrastructure vendors, and their products can optimize or automate your processes, then it makes business sense, otherwise, it holds you up and costs your more in both licensing and man-hours for implementation and maintenance. But marketing is a magical thing and sometimes that just doesn't matter.
Which way are we going? I don't know, but the possibility sure is there for both platforms to succeed.