Comments on: .NET 3.0 and C# 3.0 http://blog.mbcharbonneau.com/2006/11/10/net-30-and-c-30/ Code. Cocoa. Technology. Mon, 07 Jun 2010 06:53:26 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: nasrin http://blog.mbcharbonneau.com/2006/11/10/net-30-and-c-30/comment-page-1/#comment-381 nasrin Sun, 18 Nov 2007 07:00:20 +0000 http://blog.downtownsoftwarehouse.com/2006/11/10/net-30-and-c-30/#comment-381 I have a program that had wrote with c#.Net 3.0, now I want to run this program in visuall studio .Net 2.0 ,but I dont know what am I doing,please guide me! I have a program that had wrote with c#.Net 3.0, now I want to run this program in visuall studio .Net 2.0 ,but I dont know what am I doing,please guide me!

]]> By: Scott Stevenson http://blog.mbcharbonneau.com/2006/11/10/net-30-and-c-30/comment-page-1/#comment-380 Scott Stevenson Sat, 11 Nov 2006 05:22:28 +0000 http://blog.downtownsoftwarehouse.com/2006/11/10/net-30-and-c-30/#comment-380 Without doing an in-depth analysis, I believe virtually everything you describe here can be done with NSPredicate + NSArray or NSPredicate + Core Data. Core Data shields you from the data source-specific logic, and you can certainly use tranisient attributes (amongst other things) to modify values before they're returned in the query. Grouping, sorting and aggregation all works as well. <i>even specify joins on different types of in-memory data objects, by specifying which attributes are related</i> Not sure what you mean by this, but Core Data does have an in-memory store type which works as any other store type. It use a managed object model to describe relationships, and the model can change at runtime. I'm not trying to say LINQ is worthless (all I know about it is what I read here), but there's more flexibility in Core Data and NSPredicate than many people realize. Without doing an in-depth analysis, I believe virtually everything you describe here can be done with NSPredicate + NSArray or NSPredicate + Core Data.

Core Data shields you from the data source-specific logic, and you can certainly use tranisient attributes (amongst other things) to modify values before they’re returned in the query. Grouping, sorting and aggregation all works as well.

even specify joins on different types of in-memory data objects, by specifying which attributes are related

Not sure what you mean by this, but Core Data does have an in-memory store type which works as any other store type. It use a managed object model to describe relationships, and the model can change at runtime.

I’m not trying to say LINQ is worthless (all I know about it is what I read here), but there’s more flexibility in Core Data and NSPredicate than many people realize.

]]> By: Marc Charbonneau http://blog.mbcharbonneau.com/2006/11/10/net-30-and-c-30/comment-page-1/#comment-379 Marc Charbonneau Sat, 11 Nov 2006 00:25:51 +0000 http://blog.downtownsoftwarehouse.com/2006/11/10/net-30-and-c-30/#comment-379 I'm a big fan of NSPredicate (there have been many times when I wish I had something similar that worked with .NET's List<T> class), but LINQ is really much more than just filtering. The aim isn't just to work with the .NET collection classes or SQL, but any data source, whether it's XML, an SQL database, ldap, and so on. Any query logic that's specific to those data sources is abstracted through LINQ, and you can also use your own delegate methods to extend any part of the query (say, to modify values before they're returned in the query). There's more to it than just searching though; LINQ also does sorting, grouping, aggregating results into a single value as well as splitting them apart. If I'm understanding it right, you can even specify joins on different types of in-memory data objects, by specifying which attributes are related. I've heard good arguments for and against this kind of abstraction, and I don't know how much of it will be practical in the real world. LINQ has definitely caught my interest though. I'm going to have to spend some time reading more about it once it's officially released. I’m a big fan of NSPredicate (there have been many times when I wish I had something similar that worked with .NET’s List<T> class), but LINQ is really much more than just filtering. The aim isn’t just to work with the .NET collection classes or SQL, but any data source, whether it’s XML, an SQL database, ldap, and so on. Any query logic that’s specific to those data sources is abstracted through LINQ, and you can also use your own delegate methods to extend any part of the query (say, to modify values before they’re returned in the query). There’s more to it than just searching though; LINQ also does sorting, grouping, aggregating results into a single value as well as splitting them apart. If I’m understanding it right, you can even specify joins on different types of in-memory data objects, by specifying which attributes are related.

I’ve heard good arguments for and against this kind of abstraction, and I don’t know how much of it will be practical in the real world. LINQ has definitely caught my interest though. I’m going to have to spend some time reading more about it once it’s officially released.

]]> By: Scott Stevenson http://blog.mbcharbonneau.com/2006/11/10/net-30-and-c-30/comment-page-1/#comment-378 Scott Stevenson Fri, 10 Nov 2006 23:17:53 +0000 http://blog.downtownsoftwarehouse.com/2006/11/10/net-30-and-c-30/#comment-378 NSPredicate.h adds category methods for NSArray and NSMutableArray: -filteredArrayUsingPredicate: -filterUsingPredicate: I think the results are very similar. NSPredicate.h adds category methods for NSArray and NSMutableArray:

-filteredArrayUsingPredicate:
-filterUsingPredicate:

I think the results are very similar.

]]>