THE SQL Server Blog Spot on the Web

Welcome to - The SQL Server blog spot on the web Sign in | |
in Search

Greg Low (The Bit Bucket: IDisposable)

Ramblings of Greg Low (SQL Server MVP, MCM and Microsoft RD) - SQL Down Under

LINQ Terminology 101 for DBAs and Others Puzzled

This blog has moved! You can find this content at the following new location:

Published Friday, February 22, 2008 1:49 PM by Greg Low

Comment Notification

If you would like to receive an email when updates are made to this post, please register here

Subscribe to this post's comments using RSS



steve dassin said:

In the movie 'Dr. StrangeLove' when the Russian ambassador announced they had the utlimate weapon, a doomsday device, the good Dr. replied, "what good is a doomsday device if the world doesn't know about it". One can now ask the same question of MS in regards to linq/ef. MS finds itself between a rock and a hard place, between a blank stare at the very least and mis-understanding at the very most.  They are set to launch a new set of ideas in much the same fashion as one would throw an anvil to a drowning man. The sad fact is way too few have a good understanding of what the hell all this fuss is really about:) MS has done a terrible job in communicating what linq is, what issues in fact it is an answer too.  The sql community, by and large, has shown itself ill prepared to understand linq since it long ago gave up thinking that computer science had any relevancy to the relational model. The prevailing view in MS is that sql is legacy code. So they are throwing it out and with it the relational model!. This is a classic example of throwing the baby out with the bathwater. There is no evidence that there is anyone in MS who knows what the relational model even is. One can only assume that they make the simple minded assumption, like so many others, that sql is the same thing! To say that sql is 'cumbersome' or views 'litter' a database must be their idea of humor. They must explain to users in the clearest language possible why they determined many years ago that sql sucks. MS assumes users really understand the concept of 'impedance mismatch'. Well they do not! MS assumes users understand the difference between the strings of (N)Hibernate and Ruby on Rails vs. the idea of an 'integrated' query language. Well they do not. Do you really think users understand what Jim Gray meant by 'traditional relational databases never being designed to allow for the commingling of data and algorithms.'? And why linq is an answer to this issue? Do sql users really understand why stored procedures are part of the problem in O/R and not part of the solution? The fact is few sql users understand the computer science of net. How can they be expected to understand the concept of enumerable, its significance in representing a sequence let alone the importance of an anonymous type. The whole idea of the significance of  'types' in a query that can be represented by a 'variable' is alien to sql users.

Writing about linq performance, sql mapping that is 'impenetrable', 'entities' as part of an object model and even the syntax design of linq/esql seems to be putting the cart before the horse. It would be much more appropriate to first write about the basic issues involved, what it means to represent a query within an object oriented runtime environment and why it's an advancement over sql. Why such an environment could just as easily support a 'relational' model.  

Some thirty years ago IBM concocted System-R, a mess called sql, that is still with us in the form of DB2, Oracle and Sql Server. But then, as now, the relational model was a mystery so users bought the farm. Now history has a chance to repeat itself with an object model. But much less so with informed users, no?

February 22, 2008 4:58 AM

Bill McCarthy said:

thanks  !!


February 22, 2008 7:54 AM

Chuck Heinzelman said:

Thanks for throwing this out there, Greg.  I've heard a lot of incorrect first impressions about this technology, and this will help to clear some of those up.

February 22, 2008 10:44 AM

alphatross said:

Great article, Greg!  Anyone had experience with the old Progress 4GL language (now OpenEdge ABL I think) ? While Progress' RDBMS and 4GL Language made SQL look like Tutorial-D in terms of not adhering to the Relational Model, it *was* nice to have the database querying language seamlessly integrated with the Procedural code.  I'm more of a DBA than a Coder, so I may be out of order here, but DLinq reminds me of the Progress 4GL in that I believe MS are trying to move away from the current situation where SQL queries\stored proc calls are separated as strings that need to passed to objects within code, making the code look like two languages (e.g. C# + T-SQL) gaffer-taped together.

February 22, 2008 5:03 PM

The Bit Bucket (Greg Low) said:

Alphatross posted an interesting reply to my blog entry about LINQ and Entity Framework terminology.

February 24, 2008 6:36 PM

alex hatcher said:

Are the developers going to be on call when the server spikes due to a bad inline (Linq) sql query that starts being called?

no. they call the DBA who will be able to figure out pretty quickly through profiler the issue.  and guess what, they won't be able to fix it, since it's hard coded in someone's .net code.

yes this happens, don't kid yourself.

January 13, 2010 7:48 PM

Leave a Comment


This Blog



No tags have been created or used yet.


Privacy Statement