Changes at Real Software Part Deux

I stand by my previous post, but I’ve also had a chance to think about it a bit more and I have an area that could have been cut or shelved until further notice to help cut costs instead of laying developers off:  REAL SQL Server

Geoff Perlman showed off some nifty new features of an upcoming version of REAL SQL Server at Real World 2008.  He also talked about how usage of RSServer increased something like 20% while usage of MySQL decreased roughly 20%.

My first question is this:  Do you know anyone using REAL SQL Server?  I can say I know of one developer using it in a commercial application.  I come in contact with a lot of RB developers and only knowing one seems a bit suspicious.

However, in the summer of 2008, the Association of REALbasic Professionals asked the RB community some questions on how they use RB.  We asked specifically what databases they were actively developing for.  Less than 7% of the community responded with REAL SQL Server while nearly 20% actively developed for MySQL and 9% for PostgreSQL database servers.

I understand the reasoning for developing REAL SQL Server – it’s the next logical step up from using the single-user SQLite database and there are no other multi-user databases that use SQLite.  It’s incredibly easy to setup and use which is a big step from most db servers but in the last year or so that argument is becoming less and less of an issue as the Open Source databases have simplified their use and setup considerably.  REAL SQL Server lacks many of the features that the FREE database servers (MySQL and PostgreSQL) have had for years.  Among the features off the top of my head:
•    Stored Procedures written in a SQL-like language.  Sure, REAL SQL Server is introducing plugins that are sort of like that but they’re written in C (unless something has changed since Real World).
•    Strong Datatypes:  REAL SQL Server, like it’s little brother SQLite, is typeless.  Can you think of another database server in the world that is typeless?
•    REAL SQL Server, like SQLite doesn’t appear to enforce foreign key constraints.
•    You can’t drop a column once it’s been created.

Let’s face it, REAL SQL Server is trying to hit a moving target with REAL SQL Server.  If it was five years old today it would still be a hard sell because MySQL and PostgreSQL database servers are free (you can argue that MySQL isn’t but that’s a different topic) and even at the very reasonable price of $500 it is still too high when competing against free.  You could also argue that both of those free database servers have hundreds (perhaps thousands) of developers poking around in the internals making important changes to the code on a regular basis.  REAL Software just doesn’t have those types of resources available.

And really, what does REAL SQL Server get you that you can’t get on other database servers?  Um…easy setup and use?  Other than that, working with REAL SQL Server is exactly the same as using any other database server.  Well, okay, you get LESS features than other database servers because you can’t drop columns, don’t have strong datatypes, or get stored procedures.  I just don’t see how REAL SQL Server can help sell more copies of REALbasic which is where RS’ volume sales are.

Why not develop a complementary product that developers will pay extra for?  One thought is an integrated Report Designer and Generator.  It’s useful for those 80% of RB users developing database applications and they’d even pay something extra for it.  Have a lite version that ships with RB and for an extra $200 add very useful functionality.

The counter argument is that they’ve already spent a boatload of money in development of REAL SQL Server.  Just because you’ve started down the rabbit hole doesn’t mean you need to keep going.  A company that is experiencing financial difficulties should take a look at their entire product line and determine if each and every product makes sense financially now and in 6 months.

My two cents worth.  Your thoughts?