"This is worse than reading an RL Stine book. This is worse than reading YouTube comments! This thing sucks donkey ****!"
I've been in the office with a great many very smart, non-IT people using a BI tool for the first time. I see their faces twist and grimace, watch them turn shades of pink, see the sweat bead on their foreheads and drip on their keyboards. I know exactly what they think about the developers at that moment: "They were too busy watching Magnolia and cupping each other's balls."
I review and use a lot of enterprise software and "usability" is not a word I associate with most of it. I provide occasional toned-down feedback to vendors when I really want to track down and throttle the "usability experts" they hired to give them advice, usually people with zero domain expertise. I refrain because I'm afraid of choking off business.
I thought it would be an interesting idea to take this shared pain, go back to things that work and are engaging or infuriating in game design, and bring some of the design lessons to the application audience. Working with software, particularly BI tools, ought to be enjoyable, or at the least satisfying. The one-off mashups done by web 2.0 developers are sometimes a better model for BI than the approach BI vendors are taking today. At least the mashups make looking at data fun.
The talk is an abbreviated version of of the session I'll be running at the Boston TDWI in mid-May. A portion of the talk is basic technology theory, designed to help you convince managers that open source in the BI market is inevitable. The rest is a fast run-through of open source projects at the different architectural layers and then information about adoption and risk. 120 slides in 120 minutes, which means I'll be moving fast to cover everything.
Sadly, no demos in the session. They had to be cut to make the two hour deadline.
My first reaction to the scary Hillary picture was that this is the terrible secret of space. If you have no idea what I'm talking about, well, that's pop culture for you. Posted by Mark Saturday, April 07, 2007 9:27:00 PM |
Graphs as Entertainment
Love this video making the rounds of housing prices as a roller coaster. You really notice the big rises and falls in the market when you experience them "firsthand".
I'll still read the court documents, but I'm finding that Vincent McBurney read it all ahead of me, commented, and linked to some of the underlying information. No need for me to repeat any of that. You can read about it in chronological order in post 1, post 2, post 3 and post 4.
I'm not sure I agree that this is the ETL "trial of the century" since I'm thinking the focus is narrower and the patent in its broadest interpretation is not enforceable. If he's right then more lawsuits will likely follow since you could read it as a patent for "modular code" and sue the crap out of everyone, rather than more narrow and specific implementation.
Patent and copyright lawsuits are the deep-vein thrombosis of the software industry. Sooner or later, one of these is going to be big enough that it causes a tech market stroke. It's bad enough that patent searches and IP lawyers are part of starting a tech company, without worrying about overly-broad patentsstiflingentire industries. Posted by Mark Monday, April 02, 2007 5:05:00 PM |
Informatica Wins Round 1 in Patent Lawsuit Against Business Objects
The latest case in the ETL space is an old lawsuit between Informatica and Acta that carried forward when Business Objects bought Acta. It's hard to tell if this is a slow news day or is actually meaningful to the ETL business. My guess: slow news day.
INFA went for the expected "we want an injunction to stop them from selling the product" and BOBJ responded with the usual "we'll swap out the code if the court upholds this." Depending on breadth of the infringement, that could be tough.
I read the two patents in question - 6014670 and 6339775 - and they're pretty dry reading. It's hard to work out the exact breadth of the patents since companies try to word patents to cover the maximum breadth, sort of like stretching the last of the butter on too much toast. In this case the patent could be read specifically, but certain parts are an attempt to go wide, e.g. take data from point A, transform it, and drop it in point B, using metadata and an intermediate language. If you read it loosely, you get the impression that SQL and scripting languages infringe.
I haven't seen the court documents so I can't see how and where the patent applicability was narrowed by Business Objects in this case, or what specific features were determined to be infringing. More dry reading awaits.
I'm not a fan of software patents. Most of the time they're complete crap based on so much prior art that the person submitting should be put in thumbscrews. My first scan of the patents took me down this path. An actual reading made me think that there was more substance, but I was so sleepy by the end that I might have been dreaming. Until I see more documentation it's hard to say whether this has any import for any other ETL vendors out there. At the moment, it seems unlikely given that the lawsuit is over four years old.