Portal Quiz: What is WSRP?

February 5, 2010 at 4:23 pm | Posted in Fun, portals, Standards | 4 Comments

OK, it’s time for a portal quiz: What is WSRP?

Possible answers:

A. A Spanish-language AM radio station in Jacksonville, North Carolina

B. The Republican party state branch covering Microsoft headquarters

C. Web Services for Remote Portlets

D. A research project related to the biblical times of West Semitic peoples

It’s a trick question of course – they’re all true.  If you don’t believe me, see here, here, here, and here (for a fascinating treasure map from biblical times called the “Copper Scroll”!).

Well, now that I lay it out that way, my chosen topic for today – the Web Services for Remote Portlets standard – seems pretty boring.  But I’ll try to carry on.  A rating category called “standards support” has found its way into most portal evaluations I’ve ever helped with.  And WSRP support is usually on there.  Things get murkier when I ask what they really plan on doing with it:  “Well, we’re very into standards here … architectural guidance to use web services … there are a few different portal products in house that we may need to talk together …”

Unfortunately, WSRP shouldn’t just be a quick checkbox item on evaluations. WSRP can be useful, but in a limited set of use cases where it applies. And even if one of those use cases applies, you’ll also need guidance on how security, trust, UI frameworks, and optional services should be used. Usually if you’re just trying to solve portal proliferation problems, developing RESTful services for applications (or RSS/Atom for content) and then writing “last mile” portlets for each portal works better.

So what are those use cases where WSRP makes sense?  I did some digging and found three that hold water:

1. Syndicating a Branded Portlet to Users on Platforms Outside The Syndicator’s Control: This is where it’s not enough to just make the information easily usable in multiple portals, but the branding with it (the exact style, layout, colors, etc) is important too. But keep in mind that there are other choices (like Flash) if branding or formatting integrity—not the formalism associated with a portlet—is all you need.

2. Developing a “Portal of Portals”: If you want to create a new portal from portlets of existing portals, you could use WSRP wrappers to do this.

3. Exposing Portlets from Another Platform the Development Team Doesn’t Know: Are you Java based and your .NET programmers won’t talk to you?  Have them give you WSRP portlets instead of Web Parts and you’ll get along much better.

Now that I’m done with WSRP, you can get on to finding that ancient treasure. Let’s see … “In the Second Enclosure, in the underground passage that looks east…”

Advertisements

Microsoft Loses Open XML Vote

September 5, 2007 at 7:56 am | Posted in Standards | Leave a comment

Allegations of stacking the deck by recruiting a slew of small countries with one full vote but little real power. Buying votes. Intense lobbying efforts. Possible voter fraud. Is this another round of government diplomacy aimed at toppling a thuggish regime? Nope – it’s the vote on Open XML, Microsoft’s proposed open document standard.

Yesterday Microsoft lost its first bid to get Open XML approved as a standard (despite a New York Times prediction that it would win). A lot of digital ink has been spilled on this topic. See representative stories in Computerworld, InfoWorld, and eWeek. All of these mostly talk about the diplomacy and procedural issues, not technical issues. The British Standards Institution wouldn’t say why it opposed Open XML for some reason. In New Zealand, “Stakeholders raised several philosophical concerns, and identified technical omissions, errors and inconsistencies within the draft Standard” according to GeekZone.

Philosophical concerns? C’mon, what’s the real issue? Bob Sutor of IBM has a list of blog entries about why Open XML is a bad idea here. Many of them are still of a political nature. Rob Weir, also a strong opponent, has some charts and graphs of voting oddities here.

To check out the other side, Jason Matusow gave the Microsoft side of things in his blog in July. He addressed more of the back room diplomacy accusations last week.

There are some technical accusations that the ODF standard is superior (or at least already exists so there’s no need for another) and that Open XML includes features that companies other than Microsoft or on non-Windows operating systems would struggle to implement. But BetaNews says the real purpose of trying to get Open XML approved “is to enable the company to be perceived worldwide as cooperating with businesses and with nations in the drive for interoperability, especially for the benefit of the European Commission which maintains that Microsoft is an unfair competitor.” It goes on to say “Microsoft needs to have the opportunity to be seen as going through the motions, as playing by the rules, in order to escape the wrath of a lawmaking body that needs the presence of an evil enemy to justify its own authority.”

If Microsoft’s real goal was to get press coverage of their effort to get a standard approved, they succeeded admirably.

Is SecondLife the Next AOL?

August 9, 2007 at 8:34 am | Posted in Standards, virtual worlds | Leave a comment

AOL flourished as an early provider of a proprietary network, viewer, and content at a time when equivalent open standards did not exist. Once HTTP, HTTPS, HTML, ECMAScript-based scripting, and standardized browsers became prevalent AOL was unceremoniously pushed down an icy slope.

There are many similarities with SecondLife’s lock on virtual worlds, but the differences should be laid out first. Linden Lab is not trying to create a proprietary lock on Second Life. Its only revenue source is real-estate. It is not trying to lock in the virtual browser and, in fact, has placed it in open source with a GPL license. The server will still be proprietary, although that may change over time too. Glyn Moody’s blog reports that Second Life embraces open source products itself, running on servers with GNU/Linux, Apache, Squid and MySQL. SecondLife also allows third party systems to access SecondLife (after a successful reverse-engineering of the system revealed the possibility).

Still, there is pressure for even more openness. The Economist wrote “Lots of companies are setting up shop in Second Life, but some might prefer to have their own worlds, not just islands in someone else’s world, just as they have their own websites.” They mention Multiverse, a competitor that aims to open source the entire experience, including the servers, for a 10% cut of revenue. And they are off to a decent start. The Wall St. Journal reports that Multiverse “raised $4.2 million in Series A backing in May [2007] from investment firm Sterling Stamos Capital Management and a group of 10 angel investors.”

I too believe there is room for a fully open source model. In addition to all the virtual world equivalents of browser standards (browser, web server, HTTP, HTTPS, HTML, scripting) there are many more standards to consider. To name just a few:

  • People: Identity stores including appearance, names and psuedonames
  • Objects and content: Virtual object definitions, inventories
  • Security: Authentication and authorization mechanisms, virtual certificates
  • Communication: IM, gestures (tokenized emoticons)
  • Hyperlinking: Virtual URLs (SLURLS)
  • Economics: Currency, transaction processing handshakes to enable two-phase commit

Perhaps the key will be to follow the internet model and standardize as little as possible to avoid stunting growth. But I believe the virtual world standards will have to go much deeper than the internet did to provide a useful platform given the richness of the interaction model it needs to offer.

All this doesn’t mean Second Life goes away. I believe its value proposition continues for quite a long time as it has a critical mass of virtual users, a mostly open model, and has shown willingness to adapt.

Create a free website or blog at WordPress.com.
Entries and comments feeds.