Pentaho Hosted Demo went well.
First and foremost, every time I tried to login, the first attempt failed. I always had to backtrack and login again because the site would just hang at “Just a few moments please.” This is extremely frustrating. I was only able to get in every once in a while. Not cool. Also, the timeout is fairly low – I think I even got booted once while watching a tutorial video!
The demo, naturally, is targeted at Business Analysts. That means it’s more functional and less technical. As a Data and Business Intelligence specialist, I had lots of questions initially. However, I got the gist of what was going on from the demo. Clearly this demo site wouldn’t help me install and configure a Pentaho server, but it would help me connect to data sources and build some reports.
One thing to note here is that there are some terms used that aren’t defined well on the demo site. Terms such dimensions and measures. These are specialized jargon used in the database industries. However, I’ve had to explain the terms to every Business Analyst I’ve ever met. I don’t think they are common terms on the Business side of the house. Let me quickly explain these terms using analogies.
- A dimension (which has nothing to do with dimensions as in Star Trek or The Twilight Zone) is like the horizontal axis on a chart. A dimension is a “thing” so to speak, like time, people, houses and so on.
- A measure is like the values that are inside the chart. These are typically numbers – anything that can be counted, really. A measure is a “value”, like $52, 30.525. If you’re just counting “things” (like dimensions, say, fruit) the value can just be 1 – which perhaps indicates “it exists.” Measures can be additive (meaning they can be summed across any dimension), semi-additive (they can be summed across only some dimensions), or non-additive (they cannot be summed at all). Please consult the information superhighway for more information.
For those of you who aren’t familiar with those terms, I’ll break them down.
- An interactive report is a fairly standard report (a grid, a matrix, a chart) that can be filtered using – as Pentaho calls them – prompts. Prompts in other systems are known as parameters or run-time options.
- An analyzer report is much more interactive report that allows drill-downs. In fact, let’s say you click on a country: USA. If you have your data hierarchies set up correctly, the report will drill down to states. Pick Colorado and it will further drill down to cities. This type of report is dynamic and allows data to be "sliced and diced" as the saying goes.
- A dashboard is a collection of charts and graphs. Each graph or chart is shown in a panel. The panels can be moved around, made larger, smaller, or reshaped. Colors and indicators are integral to dashboards. The whole point of a dashboard is to give basic, high-level information. A red color is often used to indicate to the viewer, “hey, pay attention to this!”
He couldn’t answer the question, unfortunately. It sounds like publication is not a major part of Pentaho’s business model. If it’s not, then the product will unfortunately not be useful to me. I have plenty of tools for data exploration that do the exact same thing (Power View jumps to mind). What I need is a way to get these interactive reports and dashboards on a website! Anyway, he said he’ll get back to me with that information.
UPDATE: Embedding Pentaho in a webpagesounds way more complicated than I was hoping.