Pentaho, quo vadis?
Today Pentaho announced the acquisition of parts of the LucidEra BI solution (formerly called 'ClearView') to replace the current OLAP client JPivot. For the past couple of years JPivot has been the ugly duck in the Pentaho BI solution but although there are better/prettier open source OLAP clients available (e.g. JPalo) it was never replaced by something better, nor has there been done much work to improve JPivot. But don't get me wrong: JPivot has a lot of nice features and is fine for some basic OLAP stuff. To do any complex stuff however (or even quite simple things like adding a calculated member), you'd have to master the MDX language. Oh, and did I already mention that it's ugly? So there was some pressure in getting a better piece of software in place, especially since Jaspersoft has had a nice drag & drop OLAP gui for some time already. So from a competitive point of view, acquiring the LucidEra software was a very good idea. On the other hand, adding it as a closed piece of software might raise some eyebrows here and there...
Enterprise Edition only
At first sight it looks like Pentaho is following Jaspersofts lead in becoming an 'open core', instead of an 'open source' vendor, meaning that the basic stuff is open source, but the advanced components like the nice OLAP client and the Dashboard Builder are only available in the paid Enterprise Edition. That's not necessarily a bad thing; the Enterprise Edition was always intended for companies who needed a) full support, b) ease of use and c) additional features, especially in the area of deployment & managability. Adding a better and more user friendly OLAP client fits in the strategy of adding value to the Enterprise Edition, as would adding a new web based ad-hoc reporting component. After all, without customers willing to pay for your software or services any company would be out of business pretty soon. Still, I think acquiring this product and releasing it as an EE only component shows a couple of things:
- Customers are apparently not interested in paying for a supported full open source product plus services without getting additional features
- Pentaho apparently wasn't able to either enhance Jpivot or build a better solution itself, internally or with the help of the community, within a reasonable amount of time.
It was not only Pentaho the company that was looking for a better OLAP client; the community has been working on this as well. During the last Pentaho Community Meeting in Barcelona Tom Barber and Paul Stoellberger presented the current state and future development plans of PAT (Pentaho Analysis Tool). Since then they've been working feverishly to have the 0.5 version ready for presenting it at the upcoming Community Webex meeting on October 7. To give you an idea of what PAT will look like, here's a screenshot of the tool and the first chart produced. Please note that this is a flash chart (unlike in Jpivot and ClearView) which will be drillable in the 1.0 release (currently there are no drillable charts in Jpivot nor ClearView)
It'll be interesting to see how the Pentaho BI Suite will evolve and keep both the community and the paying audience happy (and thus, the company financially healthy). Pentaho is one of the rare open source/open core companies that actively seek cooperation with their community, and many new developments have been initiated by the community and were later adopted by Pentaho. And in contrast with other BI suites (whether open or closed source) you can build a complete BI solution with the community edition of the Pentaho suite (as Roland Bouman and I showed in our book 'Pentaho Solutions'), including ETL, OLAP, Data Mining, (interactive) Reporting and Dashboards. And maybe the acquired OLAP tools will trigger the PAT team to build an even better solution, but than as a Community version. Long time community member Pedro Alves already showed them how to do that: the Community Dashboard Framework (CDF) and the more recent CDF Editor may require a bit more work in setting things up, but are capable of creating far more powerful solutions than currently achievable with the Enterprise Edition Dashboard Builder. So for now, kudos to Pentaho for acquiring a nice piece of software to be added to the Enterprise Edition (and for offering it free of charge to both existing and new customers!), and hopefully these developments don't stop Tom and Paul from delivering an even better solution with PAT.
ps. it seems that James Dixon (CTO of Pentaho) was writing a blog simultaneously which makes for a great complimentary read. He clarifies some of the doubts I expressed in my post and he's right on the mark. The part I like best is quoted here (underlining by me):
"The open core model only works if the open source software is full-featured and valuable. If it is not there will be no community, and no community contributions. Without this the development costs will be the same as with a proprietary development model and the company will fail."