Practical Mashups

Mashups don’t live alone. They connect to a dizzying array of information sources, both public and private, and deliver just-in-time answers to all sorts of destinations. Last quarter I took part in a cool project to integrate Jackbe’s Presto into WSRP-compliant portals like Oracle Portal/WebCenter, IBM WebSphere, BEA WebLogic, and the like. It was, in my opinion, a practical example of where mashups can add meaningful value in an enterprise. And I am proud to say that I’ve just completed yet another cool-yet-practical integration. JackBe’s Presto is now integrated with HP SOA Systinet through the HP Governance Interoperability Framework (GIF).

Imagine this: you use your friendly BEA/Oracle/Sonic/IBM ESB to create a new service endpoint against a database. But then what? Sure you can share it with the folks in the nearby cubicles in a direct ‘lemme email ya the WSDL’ kind of way. But any good enterprise architect knows this doesn’t work in any real enterprise. How would your 500 or 5,000 enterprise mashers (or your 50 developers for that matter) find this new service endpoint? And how do they know what it represents? And how can you ensure that only permitted users mash with it? And what happens when you make version 2, 3 or 4 of that service?

At JackBe we get asked these practical questions all the time and the industry experts talk about them quite a bit too. Our friend Dion Hinchcliffe, in his post 'The top10 challenges facing enterprise mashups', discusses governance, security and version control as some of the most important issues you need to tackle before your mashup effort is enterprise-ready. And Clint Boulton at eWeek wrote a well-titled article, ‘Mashups Show Promise but Require IT Governance’, on this topic just last week.

And now JackBe mashups can be created from secure and governed SOA services. Good and practical. We’re thrilled that HP has helped us address a common concern about mashups in the enterprise. HP and JackBe will be holding a webcast to discuss and demonstrate our new integration on March 12; you can register here.

I’ve got other projects in the works. Just wait until you hear what’s next!

Read More...

Summary only...

Welcome to the Party IBM!

Wow, two big weeks in a row for enterprise software. Last week was all about the now-passe Oracle/BEA/Sun/MySQL acquisitions. This week is another big week for enterprise software: IBM announced ‘Lotus Mashups’ at LotusSphere this week.

This is a major milestone for ‘Web 2.0 for the Enterprise’. Sure upstarts like JackBe talk about this stuff. But when leading-edge technologies like enterprise mashups are productized by a tried-and-true software provider like IBM, you know that your conservative, non-early-adopter type of CIO has gotta ask ‘what’s that and do I need it?’. I think Ian White at ComputerWeekly summed it up best: “Using browser based technology Mashup will enable internal and external business objects to be deployed and connected by end users. This will create a new generation of self service applications defined by end users and connecting processes and data at the glass in a way that suit the business not just IT. Potentially this is a very exciting announcement.”

And this should also be a wakeup call for the rest of the big enterprise software providers who don’t have a concise enterprise mashup offering (I’m talking about companies like SAP, Salesforce, Oracle, CA and HP). Soon these guys will realize what IBM already knows: enterprise mashups are the face of a SOA platform for the business user. How can you not have that as part of your enterprise software portfolio?

IBM’s Rod Smith and David Boloker have been spearheading mashups at IBM for well over a year and we grudgingly admit they’ve done justice to the concept (we like their alphaWorks QEDWiki site). And we think JackBe and IBM are not just aligned in its marketing-speak, but also aligned in our overall architectural view of the enterprise mashup space. Check out the architectural similarities between Lotus Mashups and JackBe’s Presto. It's qualities like this that [we think] make mashups enterprise-ready and enterprise-grade:

• Lightweight and server-based;
• Built around security and governance;
• Dynamically driven;
• Consumes multiple data sources;
• Gets data to the user quickly;
• Let’s the user tag, search and share mashups.

You probably know that IBM has five major software brands: DB2, Lotus, Tivoli, Rational, WebSphere. If you know what these brands encompass then you’d probably agree it makes sense they’re putting mashups under the Lotus brand. Lotus is the most ‘user-centric’ of the 5. And I think it is also a testament to the fact that enterprise mashups can actually be about the user, not the developer or some back-office middleware software. This is, of course, exactly what we mean with our now semi-infamous tagline, ‘The User is the Killer App’.

And we hope that continued focus on the business user can remove some of the FUD (that ‘fear, uncertainty, and doubt’ for you non-warrior types out there) concerning user-facing/user-driven Web 2.0 technologies like enterprise mashups, wikis, and blogs. I think Ross Mayfield expressed these concerns best: “The new [Lotus Mashups] tool gives users an easy way to build composite applications that they can share with others and publish to their own or a shared workspace. One analyst said he wondered if IT administrators would be concerned by the possible security and management implications that may arise.”

While we at JackBe agree there are issues like security and governance to consider (and we’d like to think we have a pretty good handle on them), the real impediment to ‘user-driven’ enterprise solutions seems to simply be our 25+ years of inwardly-focused IT efforts. This makes it easy to forget that the average business user is more-and-more technically-inclined and self-sufficient every day. You can thank the constant flow of consumer-type sites like FaceBook, Digg, and NetVibes for that. ‘Born Digital’ has an entirely new meaning now.

So, if you are a Lotus customer, congratulations! You have something to look forward to. But, if you’re like many companies who don’t have Lotus, come give JackBe’s Presto a look.

I can't wait to see what next week brings...

Read More...

Summary only...

Interesting Times

‘May you live in interesting times.’ It may be real a curse from ancient China or merely something invented by an imaginative book editor. Either way, it fits the world of enterprise software. In case you missed it yesterday, approximately nine BILLION in cash and stock went from Oracle and Sun to BEA and MySQL, respectively.

Here at JackBe we love Oracle and Sun. And not just because they give me a couple of gazillion dollar corporate acquisitions to talk about during my guest appearance on Fox Business News last night, but because we’re close partners with both companies and we have over a dozen Oracle/Sun alumni. So it’s no surprise that we follow both companies closely. In my opinion, both acquisitions are biggies for the acquirers, the acquirees, and the enterprise software industry in general.

In case you didn’t know, Sun already has two database products. JavaDB is based on Derby and is all Java. Maybe it’s not an enterprise-grade database but it’s there. And back in 2002, Sun acquired Clustra, a high-performance clustered database. Many of us at Sun at the time thought this was the Oracle competitor but it didn’t seem to go anywhere. So, I bet you’re thinking what I’m thinking: what the hell is Sun going to do with MySQL? Is Sun onto something here? Are they the one to bring some fresh perspective into the database space? Boy, I hope so. This is certainly their chance. Here are a couple of possible outcomes:

1. Sun offers a shrink-wrapped ‘DB-in-a-Box’. For a few million you get a high-performing Sun E25K server optimized and bundled with MySQL.
2. Sun creates the first database on a chip. If anyone can do it, Sun can. They own the hardware and the operating system. It’s right up their alley.
3. (My Favorite) They service-enable MySQL and make it their entry into the Web 2.0 space. Quoting directly from Sun’s press release: “This broad penetration coupled with MySQL's strength in Web 2.0, Software as a Service (SaaS), enterprise, telecom and the OEM embedded market make it an important fit for Sun.’ Sun, if you’re reading this, we’re here to help.

As for Oracle, a quick Google search on the phrase ‘Oracle acquisition’ will confirm what you probably already know: Oracle has done a lot of acquiring and BEA isn’t likely to be the last one. So, ho-hum, another part of the enterprise-solution pie is now owned by Oracle. Interestingly, in one odd way the Oracle acquisition parallels the Sun acquisition: Oracle will now have 4 distinct portal solutions (Oracle Portal, Oracle Webcenter, BEA WebLogic Portal, and the Plumtree products that are now under BEA AquaLogic brand). The Oracle Fusion and BEA AquaLogic feature sets also have a great deal of overlap. It kinda makes Oracle the 800-pound gorilla of portals and SOA.

Are you a Portal/Fusion/WebLogic/AquaLogic user who is now a little concerned about the future? This screams for consolidation of the portal and SOA product lines. And as we’ve written in the recent past, this becomes yet another Oracle-owned property that has only marginal tangible integration to other Oracle products for the immediate future. Sure, Fusion will fix it all but that’s not a journey that will be completed anytime soon. Oracle, if you’re reading this, we’re here to help.

Finally, here’s a result of these acquisitions that hasn’t gotten as much attention as it should: Sun and Oracle become real software competitors. MySQL has nibbled a $50M hole in Oracle’s database pie already. And once you remember that Sun sells a lot of hardware that runs Oracle databases (especially the big boxes that sell for up to $4M), you’ve got to wonder how these moves might impact the lucrative Sun-Oracle server-database relationship. On the portal front, Oracle’s portal domination could also impact Sun’s portal, Sun Java System Portal Server. Do you turn to a low-cost Sun alternative to avoid the confusion of the Oracle portals? Perhaps it’s too soon to tell.

Interesting times indeed. And it’s only January 17. It’s shaping up to be quite a year!

Read More...

Summary only...

Web 2.0 in 2008: What's Out, What's In

Some of us are ready for the Web 2.0 wave that is now breaking over us, and some of us are not. The McKinsey Quarterly just put out an insightful article, “Eight Business Technology Trends to Watch” (registration required), that outlines eight unique business trends that will be enabled by the Web 2.0 technology wave. With apologies to the guys at McKinsey for oversimplifying their detailed work, here’s my brief synopsis of their ‘Big 8’:

McKinsey’s Eight Business Technology Trends to Watch

• ‘Distributing co-creation’: Let everyone in your supply chain create.
  
• ‘Using consumers as innovators’: Let the people who consume your product also customize and extend your product.
• ‘Tapping into the world of talent’: It’s not plausible to hire all experts as full-time employees. It’s becoming more effective to hire experts for specific jobs as needed.
  
• ‘Extracting more value from interactions’: Tacit interactions are becoming more valuable and will be core to the workforce by 2015.
• ‘Expanding the frontiers of automation’: Continue to give more and more visibility to customers and supply chain.
• ‘Unbundling production from delivery’: Treat your IT systems as a platform and not only a destination so other companies can “plug-in”.
• ‘Putting more science into management’: Syndicate analytics to partners and customers so they can analyze and provide better information back into the system.
• ‘Making business from information’: Innovate above raw innovation to create higher value information systems.
  

After reading this I was inspired to create my own ‘What’s In and What’s Out for the Knowledge Worker in 2008’. I expect some of these will resonate with you no matter what your role is in your enterprise.

JackBe’s What’s Out and What’s In for Knowledge Workers in 2008

• Out: Big IT, In: Big User - Don’t grow IT to help the knowledge workers. Grow the ability to help the users create (mashups).
• Out: Integration, In: Collaboration - Don’t integrate systems, put collaborative tools in the hands of users so they can connect information in their own unique ways.
• Out: Consumer Mashups, In: Enterprise Mashups - Tired of Google Maps as the premier mashup? Enterprise mashups based upon real ERP/CRM/SFA/homegrown data help drive dramatic competitive advantage.
• Out: Big SAP, In: Little SAP - Help users have “smaller” access to their SAP data. No single user needs it all; most want it in bite-size chunks.
• Out: Integrate Everything, In: Let the User Mashup - Knowledge workers need ad-hoc and situational data which is dynamically integrated in small amounts. Provide the tools to let the users integrate (mashup).
• Out: LinkedIn the Website, In: LinkedIn in your Enterprise - Professional networks are growing quickly. Enterprises need this capability within the enterprise, as users are coming to expect this kind of empowerment.
• Out: All-in-One Web Destinations, In: Embeddable Mashup Applications - Users will increasingly resist going to a new one-size-fits-all Web destination every-time they need information. Instead, provide capability for users to embed mashup applications in their own portal, application and even their iPhone.
• Out: B2B, In: U2B - When you think B2B, you think EDI. When you think EDI, you're thinking about lots of time and cost. Instead of trying to integrate disparate businesses using B2B, promote User-to-Business access to information. Don’t push information to the user, notify them and let them have access to the amount of information they need.
• Out: Emailing Excel Spreadsheets, In: Mashing in Excel - Copy and paste into Excel and emailing the spreadsheet is Mashup 1.0. Let the user mashup in Excel and collaborate by pushing out parts of Excel information as shared data services.
• Out: Portals, In: Portals - Portals will certainly be around for a while. 2008 is a good time to re-assess your portal 1.0 technology and start moving to an agile and lightweight portal 2.0 technology like Netvibes and PageFlakes.

This is a lot for any enterprise to embrace but the paybacks can be massive. So, is your enterprise ready?

Read More...

Summary only...