Delphi 2006 thoughts and .NET 2.0 challenges Jan 27, 2006 12:05
A while back (December 7th, 2005), I went to see John Kaster show off Delphi 2006. I liked what I saw. I really liked the live templating system; the fact that it was customizable and very extensible. IMO, this embraced the classic Delphi attitude of extensibility that brought forth component-based development.
I also really liked the direction Borland was going with ECO. The ASP.NET autoforms and the state machines are pretty cool. (Autoforms reminds me of Ruby On Rails scaffolding; ECO itself, like ActiveRecord...the nice thing about Ruby on Rails is that it's free and open source. I digress, but I'll be talking more about Ruby for sure.) But the best thing about ECO IMO, is that Borland's product marketing people pushed ECO features down in the SKU levels. This will help a lot to make ECO a more widespread technology in the Delphi community. I do think that it would really help if they pushed the ASP.NET stuff down, but it's a start. Hopefully, they'll consider doing that in Highlander, but I'm not holding my breath.
Finally, BDP gets connection pooling; finally it becomes usable for production environments. Yes, yes, I know there are already 800 people out there that have been using it in production environments, but I would have never used it without connection pooling.
There are a lot of extra enhancements packed into Delphi 2006: C++ Builder, more complete Together integration, Delphi Win32 language improvements, debugger enhancements, VCL visual guidelines, general IDE things, etc. But perhaps the best thing that Delphi 2006 brings to the table is that they've addressed tons of quality issues. I have found that Delphi 2006 is more stable and performs better than Delphi 2005 ever did.
The largest hole in the Delphi 2006 release is .NET 2.0 support. I understand why, (.NET 2.0 just being released, etc.) however from a developer perspective presented with a competitive product (Visual Studio 2005) that does support it, this is definitely a disadvantage. Especially when MS gives away its Express editions and the Standard Edition is very inexpensive (BTW, MS gave away the Standard edition free at their launch event). It also underscores the nature of the .NET IDE marketplace: Borland will *always* be one step behind Microsoft. Whether it's compact frameworks, generics, linq, new databinding, etc.; it doesn't matter.
Many will argue that Borland does innovate in particular areas where .NET 2.0 is now catching up, like databinding. And sure that's one area, but is it enough? Would you stick with Borland's .NET databinding now that the "standard" framework databinding is out? I wouldn't. ;)
The question that will be interesting to see is just how quickly Borland releases Highlander and how well its .NET support is received. ISTM, most in the Delphi community are quite content hacking away in Win32. I am as well, for those applications that are best suited to Win32 (Rich GUI clients..."Smart Clients" is the buzzword). OTOH, for .NET work, if I have a choice (yes, I work for a Delphi shop) I'm going to use Visual Studio.
The bottom line on Delphi 2006 is this: I have confidence upgrading my projects from Delphi 7 to Delphi 2006. I do recommend this upgrade and I'm glad to see that the Delphi team at Borland is recommitted to quality of implementation. Go Delphi!
Write a comment
- Required fields are marked with *.