Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?
From: Thomas Miller (tmiller_at_bss-software.com)
Date: 01/31/05
- Next message: Tim Jarvis: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Previous message: Liz: "Re: Follow up."
- Maybe in reply to: Bob Dawson: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Next in thread: Dennis Landi: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Reply: Dennis Landi: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Mon, 31 Jan 2005 16:35:16 -0500
About the only thing I think you haven't stated is how fast
.Net will actually be adopted. That is an unknown too.
I would suspect 64 bit hardware to easily out pace the
acceptance of .Net by the beginning of 2006 (12 months).
With this said, within 12 to 18 months, the target market
for native 64 bit will be bigger then .Net for the current
installed base. Now once Longhorn SP2 is released, that
might reverse and I have no clue. But that is going to be
in 2010 time frame. I remember reading an article about the
companies that plan on adopting Longhorn immediately and it
was 3%-5% (first 18 months the OS is shipping).
So yes we may being heading that direction, but extremely
slow. And as you point out, it might be so slow that MS
reverses course or at least comes out with a dual point of
attach, both .Net and native API.
I think this is what is more likely and why I think Borland
needs to put as much into native as they are .Net.
Bob Dawson wrote:
> "Will DeWitt Jr." wrote
>
>>Intel plans to move their entire desktop line (from Pentium 4
>>EE to Celeron D) to support EM64T. And, I'll note, very few
>>people responded to that. It's the /truth/. Why are people here
>>afraid of it?
>
>
> No one's afraid of anything in those articles--the fact that 64bit hardware
> is coming on line in the near future is perhaps the one thing that everyone
> here knows. The disagreements concern whether that single fact demonstrates
> the need for a 64bit compiler "real soon now." You appear to think that
> self-evident, whereas others read the same articles and say 'so what?"
>
> Let's summarize the claims for and against:
>
> A. For the general software market, it remains that
> -- the vast majority of the installed base will continue to be 32
> bit for years, completely unable to run 64 bit applications,
> -- in most cases, 32bit software can do all that's needed,
> -- those new machines will still run 32bit software just fine,
> -- in most cases where speed is a problem, either
> -- algorithm or db/network speed is the issue, not bitness, or
> -- a hardware upgrade alone will provide the speedup needed.
>
> B. Additionally, the platform owner (MS)
> -- officially advises that .NET is the future
> -- has discontinued VB (its most used language)as a
> native code compiler
> -- is treating win64 as a sideshow; while
> -- pouring all its efforts into .NET 2.0 and Longhorn.
>
> C. Against these points, one must note that
> -- there are classes of applications where bitness just matters,
> notably databases, real time video/graphics manipulation, etc.
> -- there are classes of apps where .NET isn't (at least so far)
> up to the task (again, commonly media related, but also things
> like high speed math libraries, scientific calculation, etc.).
>
> Initial summarizations:
> C demonstrates the need for a 64bit compiler. It is simply the best
> solution available in certain situations.
> A and B would seem to indicate, however, that the market is currently
> small (because it's supporting specialized, niche needs), may stay small for
> some time (because the installed base will remain predominantly 32bit for
> some time), and in fact may never grow at all (because .NET is continuing to
> evolve, and MS is doing quite a bit to resolve its speed issues).
> On the other hand, .NET might just fail and MS reverse course. And the
> notion of smallness is perhaps debatable--do few programmers need a 64bit
> platform all the time, or do all programmers need a 64bit platform now and
> again?
>
> The questions then, become:
> What is the real size of the 64bit compiler market; how quickly will it
> grow, and how big will it become?
> Given the answer to that, should Borland pay the real and opportunity
> costs of developing a 64bit Delphi compiler, and, if so, how soon does it
> have to do that to secure a reasonable slice of the market?
>
> I think I'm being fair to everyone here--I miss anything salient on either
> side?
>
> bobD
>
>
-- Thomas Miller Wash DC Delphi SIG Chairperson Delphi Client/Server Certified Developer BSS Accounting & Distribution Software BSS Enterprise Accounting FrameWork http://www.bss-software.com http://www.cpcug.org/user/delphi/index.html https://sourceforge.net/projects/uopl/ http://sourceforge.net/projects/dbexpressplus
- Next message: Tim Jarvis: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Previous message: Liz: "Re: Follow up."
- Maybe in reply to: Bob Dawson: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Next in thread: Dennis Landi: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Reply: Dennis Landi: "Re: Windows XP Professional x64 Edition ---> Delphi 2005 x64 Edition?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|