Delphi Roadmap for 2005 and 2006

[Sascha Willems]

Danny Thorpe just recently posted Borland's Roadmap for the Delphi product line. You can read up on it here.

So basically they have ''DeXter'' planned for 2005 which is like D2005 still only Win32 and .NET 1.1 (which is a shame, since MS will soon release VC2005 with support for .NET 2.0).
For 2006 they have ''Highlander'' sheduled, which then finally aims to get Delphi to .NET 2.0.

And the future? There was a Borland broadcast some days ago where it was about Delphi's future (WILL posted the transcript in this topic). It was talked about the above two products, but what most Delphi developers are interested in, is (at least that's what I look forward to) native 64-Bit support. And to my disappointment it's planned for a release in 2007 or even 2008. That's two or even three years from now. And when you think about the fact that AMD's 64-Bit CPUs are selling like hell, 2006 will already have a large installed base of 64-Bit CPUs, but we Delphi developers will either have to wait until 2007/08 or use FreePascal. Rather sad if you ask me.

[WILL]

I was unfortuantely unable to listen to the live podcast due to some technical difficulties, but The two main questions I was able to get answered were:\r\n\r\n1) davidi: __WILL__: "Will we see the return of CLX?" - This has already been answered by Michael. We will continue to work the the Kylix Community Project - freeclx.sourceforge.net.\r\n\r\nAND\r\n\r\n2) jkaster: __WILL__: "Will we be seeing Delphi for the Mac(considering the Intel/Apple issue)?" - Michael: Boy, IA'd love to do something like that. Right now weA're focused on the Windows platforms.\r\ndavidi: Michael has a MacMini !\r\n\r\nI did not get a chance to ask about Kylix or what they thought about Lazarus/Free Pascal. The latter might have been a bit inapropreate aswell.\r\n\r\n\r\nA small point to note: Lazarus has really improved over the last few releases and is really starting to maintain alot of the Delphi functionality. Some of the perks that I myself have complained about have been fixed. A serious alternative to consider if you find yourself without a solution from Borland.

[LP]

This news post is somehow focused on .NET, but you did not mention some important improvements for Win32 platform:

1) Operator overloading
2) Unicode in VCL
3) Faster IDE startup
4) Improved run-time performance of compiled applications

In my opinion, .NET is not very applicable to game development considering performance issues, so whether new version will support .NET 2.0 or not doesn't concern our game studio much

I would argue with Sacha's opinion about native 64-bit support:

1) Personally, AMD has been quite a disappointment with the lifespan of their Athlon CPUs and previously, with the performance of their AMD K6-2 CPUs. This can be either related to my *extreme bad luck* or the fact that I bought most of them in US, but I wouldn't buy a single AMD CPU in the near future. I've never had any problems with Intel (and yes, I only use Intel's CPU and motherboards ) Just don't think I'm an Intel freak or something... ops:

2) "AMD's 64-Bit CPUs are selling like hell": We'll have to verify the statistics about this, but take the fact that 32-bit CPUs are *still* sold (and I suspect more than 64-bit CPUs at the moment) and I can bet at least 9 of every 10 today's PC gamers still have 32-bit CPU (and probably *will*, for quite some time). So the question is whether native 64-bit support really worth losing more than 9/10 of your market (at the moment).

3) It's not enough just *compile* the code for 64-bit CPUs - it should be efficient (consider code optimization). I think Borland should take their time to implement this carefully, without rushing things as they did with Delphi 2005. Who knows... maybe in 2007 we'll be coding in C# for handhelds and won't care about 64-bit support anyway... :lol:

There were some of my thoughts. Any opinions on this?