关注Delphi9/10的如何发展----分析Danny Thorpe(andrers之后的Delphi编译器设计师)的精彩blog（1）！

///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

Thursday, April 08, 2004

Technology Goals for 2004  

# 12:09 PM

Here's the synopsis of compiler work on the boards for calendar year 2004:

• Win32 compiler enhancements, including codegen improvements and language enhancements to improve parity with the .NET side of the house.
• IL compiler enhancements to broaden the .NET push started with Delphi 8 for .NET. This will include more language enhancements, codegen improvements, and reorganized metadata output to better align with informal conventions found elsewhere in CLR.
• Prepare for "Whidbey" (.NET 2.0). Since the release of Whidbey has slipped out to 2005, we can't deliver a commercial product to target Whidbey this year. Nevertheless, compiler work to fully support Whidbey's new IL features has to start now. The biggest item we're currently exploring is Delphi syntax for parameterized types (generics). We've had general syntax ideas cooking for several years now, so fleshing out the details in the spec and catching the caveats has moved very quickly.
• .NET Compact Framework. Make sure the compiler can import CF symbols and emit CF compatible IL. CF is quite a different platform target than desktop .NET.

While there's been vocal interest in creating a new Delphi for AMD64 project, it's just not in the cards for this year. We can't afford to drop everything to build yet another platform up from scratch, particularly when native AMD64 codegen will be coming under extreme pressure by .NET within the next year or so. Delphi 8 for .NET code already runs as native machine code on AMD64 and Itanium hardware.

More specific details will be posted here as we get things firm enough to publically commit to. Subscribe to the ATOM or RSS feeds so you won't miss anything. ;>

Plans are forward-looking speculation subject to continuous change. As these plans evolve and take shape, you will read about it here.
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

这是从Danny Thorpe的blog摘出来的“Technology Goals for 2004 ”(2004技术目标)

Delphi.net现在面临的是一个由Andrer率领的Microsoft .Net和C#小组打造.Net战略，当然了.Net的利剑是指向Java平台，但C#语言可能要扫荡所有其他的程序语言，，C#的大压下C++已经没落了，:)，不过就算没有.Net，凭MFC库，C++也无法在Windows平台一般应用程序开发中占主要地位。（不过据说Scott建议B.S修改C++，以让C++适应在虚拟平台上，B.S对此似乎抱着不反对的态度）.

我最关心的是，RAD中最重要的语言Delphi如何应付这样的局面。因为程序语言的大部分功能已经集成到OS核心（CLR也算是核心之一），所以，一般程序语言顿时失去卖点！Delphi.net,如何在提供系统提供的功能的基础上增加附加值加呢？？这关系到Delphi.net的重要生存发展!

背景说完后，我们来到Danny Thorpe的Blag来，看看Delphi如何增加它的卖点：

Danny Thorpe在下面列出了2004技术目标的一些大纲如下,我们要从中嗅出D9,D10的发展方向！

1：:Win32 compiler enhancements（Win32编译器的增强）

让delphi同时支持win32和il编译器，这在一段时间内，可以说是delphi的重要卖点！

2：IL compiler enhancements （IL编译器的增强）

这个说到Delphi.net语言本身的增强，也就是让Delphi.net在.net中取得与C#同等的First Class地位。这，确实是与C#的竞争中不可缺少的一部分，

3：Prepare for "Whidbey" (.NET 2.0). （为.Net2.0准备）

包括为delphi.net增加范型的功能，

4：.NET Compact Framework （.Net framework简缩版）

让Delphi.net产生可移植的IL，支持CF!

从2004年的计划中，看出Danny Thorpe的计划还是相当保守的。主要是提供.Net下必须提供的功能。这是这些功能还不足与可以击败C#,但是必须承认的是，用1年的时间击败C#无疑是可笑的。更何况C#的领军人物是Andrers。且看Delphi在今后几年的表现了。

下一篇，我们来看看，在Danny Thorpe的blog中,用户对Danny Thorpe和/D9,D10有什么期待，

to be continued

--- by littleroy 2004-7-3