Will ORM mass adoption kill SQL Skills
Will OR/M products succeed in delivering a Generation of developers with no or little understanding of SQL and does it matter? Recently I decided to take a close look at another OR/M (Active Record) and it got me thinking……..
After some initial suspicion, I eventually became a HUGE fan of using OR/M technology. I also appreciate my many years of having to write intricate T/SQL, however I don’t like having to write masses of data access code any more, it’s tedious and I have come to the point where I don’t ever want to have to revisit an anaemic domain model either. I want to continue using POCO’s and couldn’t be happier that REAL object oriented thinking is slowly coming back into vogue with DotNet developers (as a community) and shows up in how we model our domain objects. There is more to say on this, but ultimately it moves me farther from my question, actually its more a nightmare scenario.
The Scenario
Lets assume that in five years from now, most data access code is being managed 
by OR/M’s whether it be LINQ To SQL, NHibernate, Entity Framework, IBatis.Net, etc. A fraction of the SQL / TSQL & stored procedures we used to write is now deemed unnecessary, in fact it’s become almost entirely negligible. Universities are now (it’s 2013 remember) spending a bare minimum amount of time teaching database technology in preference to teaching OR/M technologies and almost no developers are seeking to sit 3rd party database development exams or take up an Oracle or Microsoft database certifications. Finding a burgeoning DBA in the making is like the search for new fossil fuel. Microsoft Developers are now whipping out Entity Framework 4.0 to quickly dial up an ad-hoc query cause they are not sure about how to write the SQL.
I know I have presented an extreme view and implicitly retired a lot of us SQL dinosaurs off, but is this scenario possible to some degree? What do you think? Will demand for the DBA increase as developers become less skilled with SQL.
2 Comments so far
Leave a reply
I tend to think it will be like programming languages in general. Most people don’t program in 1st, or 2nd generation languages because there are more intuative, easier to learn options such as C# and Java available. There is obviously still a niche for people with the skills to hack into C or Assembly but (device driver, embedded stuff aside) most people can remain blissfully ignorant of whats going on under the compilation covers of .NET. I’m all for this layered approach to development and suspect we will see a day when we’re just dragging a dropping high level blocks around in an IDE. Theres just one small problem with what i’ve said…. what happens when it breaks and you have to go digging into the internals to find errors. Its at that point a deeper understanding of whats going on is critical. SQL hacks will live on I suspect, for at least a little while longer…
[Reply]
I certainly hope your right. I cant imagine that every time someone is required to get some data that developers will be cracking open an OR/M when all that was asked for was an ad-hoc report of some description. By the same token I believe that my last DAL is behind me (thank heavens).
[Reply]