Customer, Advocate, Expert, Analyst

I don’t know why you’re here.  No, really, I don’t.  You may come here for opinions and advice.  Maybe you like the content or the style.  Maybe you like the soothing color scheme — I like the green.

After looking over my posts lately, I can see that many of them span a wide range of roles — sometimes I sound like a customer, sometimes an advocate, sometimes an expert, and sometimes an analyst.  The point is that unless I’m heavily quoting something, everything here is coming from a particular point of view — and not always the same one, even though there’s only one author.

I’ll be honest — if it’s not already clear — that I’m a big fan of the Oracle database.  Heck, whenever I come up against another database, I run into something I can’t do or see as easily as I can in Oracle.  The market supports Oracle’s leadership position (from the recent 11g press release: “Oracle is the #1 Database: Gartner 2006 Worldwide RDBMS Market Share Reports 47.1 Percent Share for Oracle”).

One other bit of my personality that I’ll share is that I’m a huge baseball fan — following the Boston Red Sox.  For those of you who understand the psyche of the average Red Sox fan, you know that they’re always waiting for something bad to happen (2004 lessened this feeling immensely). 

It’s with that sense that I sometimes sound less than positive about Oracle — it’s clearly the leader, but I worry.

Like anybody, I also like things to be simple — I think that the level of complexity in the world has increased immeasurably over the past 15 years, and I often wonder why.  I used to poke fun at my Microsoft friends that the Oracle 6 kernel code was smaller than Excel.  That’s no longer the case, and I can understand and appreciate that because Oracle 11g is easily 10-20x more usefully feature-rich than Oracle 6 was.  But part of me yearns for that simplicity.

Those twin feelings of waiting for something bad to happen and a skepticism of complexity lead to my recent postings on free data and mass market Oracle.  I’m concerned about Oracle getting hit at from the low and medium end.  (This recent announcement from Oracle gives me hope that they recognize it too).  The fact that I can procure a reasonably full-featured MySQL development environment online within minutes without a comparable Oracle counterpart worries me.  Posts like this and this by Nuno Suoto worry me.  And they lead me to express concern about things like Enterprise Manager Grid Control, Diagnostic and Tuning Pack and Change Management Pack.  I erroneously said that information and documentation on EMGC D&T Pack was hard to find — it’s not.  Even though Change Management Pack seems like a poor stepchild, the rest of EM is pretty easy to find.  I had a hard time installing and configuring EMGC for evaluation, but I’m sure much of that was due to the archaic equipment I was working on.  And I’m just not used to the navigation paradigm.  But once I get to the screens I’m looking for, I like what I see.

I can’t imagine how hard it must be to keep all of the various products up-to-date on the Oracle web site, and to make navigating to them easy via the various ways people like to browse (via searching, via menus, via links, etc).  With that being said, I’d like to point you at something interesting.

You may not be as familiar with Oracle Fusion Middleware as you are with the base database technologies, but the Master Data Management stuff is hot right now.  And the Oracle offerings here look cool, in particular the Customer Data Hub and Product Information Management Data Hub.  Take a look and tell me what you think.

Free Data Now!

So this post is only tangentially related to the current online effort to get access to Oracle 10g’s AWR/ASH data.  It’s actually more about data mashups.

When building an internal corporate application, most of the time you’re familiar with the internal sources of data and how to get at them — although, you can often have lots of “fun” doing the detective work to uncover hidden sources of internal data, fighting the data guardians and mastering the necessary API incantations to pry loose the secrets. :-)

By basing your application on only internal data you may be missing out on interesting opportunities to leverage public and freely available data and feeds.  Although, even there you need to be wary of usage restrictions and odd APIs.  Usage restrictions can be odd — unlimited access for “non-profit use”, or “public use”, but heavy costs and fines for “private” use.  An interesting example is Google Maps — here’s the quote from the Google Maps API page:

The Maps API is a free beta service, available for any web site that is free to consumers. Please see the terms of use for more information.

To use the Maps API on an intranet or in a non-publicly accessible application, please check out Google Maps for Enterprise.

Following the links shows you that to use Google Maps on an internal application will set you back $10,000 per year.

The question of usage rights for data has been a huge source of problems in professional sports leagues.  In 1997, the NBA sued Motorola and STATS, Inc. regarding the real-time dissemination of scores and game information and lost.  However, recently the NCAA clamped down on live-blogging at one of its events.  Even today, you’ll have a hard time finding real-time RSS feeds for MLB sports scores.  Although, those of you who love data analysis may enjoy poring over the MLB Enhanced GameCast data — latest example here and initial research here.

So, clearly you’ll need to be selective when working with external data — there’s also the questions of data reliability and validity.  However, the sheer plethora of public RSS feeds should be a new and interesting source of data considered when building data analysis applications.

I was working on this post last week when the 10g AWR/ASH petition came online here.  I’m of two minds about the issue.  In the end, I think limiting access to this data isn’t going to achieve what I believe to be Oracle’s goal of encouraging sales of Enterprise Manager Diagnostic Pack.  I’ve tried to evaluate the Diagnostic Pack — it’s not easy to install, and not easy to navigate.  Go ahead, try and find the links on to the download and/or documentation.  (I tried this for the Change Management pack, and it was just about impossible to find).  Compare that with how easy it is to find and download the database, client tools and/or SQL Developer.  In the end, I think you’ll see people trying to re-create STATSPACK for 10g in a “public” fashion rather than running out and installing Enterprise Manager.  So, you’ll have a lot of time and effort spent in lobbying Oracle for access to this data, and/or effort spent on re-creating the data capture and storage in a public way — instead of inventing useful analysis tools based on the data that’s already there.

Ah well…