There were a lot of comments about the supposed demise of LINQ for SQL people expected it to evolve into a nHibernate style framework (and rightly object to EF Entity Frameworks change tracking http://weblogs.asp.net/fbouma/archive/2008/11/21/baby-sitter-framework-2-0-change-tracking-in-the-ef-v2-it-s-still-your-problem.aspx ) .
I think people miss the point here , LINQ for SQL is and always will be a light weight solution for smaller or RAD style projects. As Damian Guard ( from the LINQ for SQL team ) explains it will be maintained and changes required by the community will be implemented. http://damieng.com/blog/2008/10/31/linq-to-sql-next-steps. As it is now it is fine for most current and future needs.
I admit this leaves some people in the hole ( especially large scale OO apps) however EF and nHibernate will duke this out. I have used nHibernate before and personally i dont think the complexity is worth the benefits , I spent a lot of times solving nHibernate problems rather than business or even architectural or software design prolems and i also had an issue with the way it handles sessions if you have lots of small stateless calls ( eg web services) it tends to be quite heavy. For EF (Entity Framework) the jury is out and i will see how it evolves.
Print | posted on Wednesday, January 14, 2009 10:17 PM