Yesterday, the GeoMetrick Enterprises’ monthly company newsletter was issued. August’s topic was about data sharing (to subscribe: (This newsletter no longer exists so this link has been deactivated, here http://www.geometrick.com/newsltr_signup.htm ). Just as with the last few months, the percentage of people reading this month’s newsletter was relatively high. I commented on this statistic in the blog, last month (http://outonalims.com/2010/07/02/comments-on-statistics/), but I’m still pondering the reason for the suddenly high reading rates. However, if you read this blog post about statistics, you’ll know it’s because my newsletter is just so totally awesome. 😉
This month, however, I’ve received enough e-mails to comment on the topic of sharing data and what a chore it is that I’m convinced this topic is of particular interest. The article mentions that, despite lots of new tools and people hawking them claiming them to be solutions to our data integration problems, we still have problems with integrating systems, thus making the topic of “data sharing” one that continues to be among the top problems I hear about in our industry.
I was talking to someone about Google searching and the outcome of the conversation was that is sounds exciting to use the Google apps to search our data but it still requires a lot of work to prepare the data so that it can be properly searched. However, we do like the Google interface and I think we agreed that having a standard interface that people are familiar with can be a positive factor.
Not long after that, someone asked me if data warehouses weren’t a great solution. My response was that they can be a good solution but are still a lot of work to create. So far, I haven’t heard of anything that replaces the hard work of data mapping that continues to be required to share data between systems. Even when it’s a drag-and-drop tool, there’s got to be someone that knows where to drop the things that are being dragged and it’s hardly ever the case that one person really knows the fields to map between ALL the systems.
In any case, I’m sure we’ll continue to see interface projects that are as large or larger than the actual implementations.