M
Michael C
Mr. Arnold said:The thing is that a lot of developers don't want to be bothered with
inline T-SQL or sprocs.
I'd suggest that these developers might want to get a job somewhere else. As
a developer writing database apps known SQL is part of the job and is a
requirement.
That's the bottom line. It's kind of clear here that you have not done any
N-Tier application development that thrives on using a technology such as
Linq-2-SQL
Do a search on something called "Attacking the man, not the arguement"
You ever work with compiled linq queries? You ever work with precompiled
views of the model? What you say is simply not correct. And in todays
database environment such as SQL server, a T-SQL batch is created by the
Linq query and is submitted to SQL server to be executed with a sproc
functionality that is cached on SQL Server.
So, the Linq translation only happens once with a compiled linq query, as
the compiled linq query stays in memory to be executed again, again and
again with no translation needed to T-SQL only parameter substitution on
the tree, because it has been translated onetime to T-SQL and doesn't not
need to be translated over and over.
Um, what are you responding to here? I never said anything about things
being compiled multiple times. What you've written here has no relationship
to anything I mentioned. Hence, I don't know how I could be "simply not
correct"
????????
It's difficult to respond to your confusions when you just post question
marks. What I was saying was the article ignores all the newer features of
sql in sqlserver 2005 where there are other solutions to some of the issues
he raised. For example, the page issue can be solved with the rownumber
function.