www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - D 64-bit ?

reply Daniel Ribeiro Maciel <daniel.maciel gmail.com> writes:
Who still use a 32-bit computer these days anyways.
Nov 01 2008
next sibling parent reply Robert Fraser <fraserofthenight gmail.com> writes:
Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


Most people.
Nov 01 2008
parent reply Xinok <xnknet gmail.com> writes:
Robert Fraser wrote:

 Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


 
 Most people.


Since the topic is here, I'd like to make a suggestion for when to make 
a 64-bit compiler.
I think the right time is when D 2.0 is stable and we begin to move away 
from D 1.0. Instead of jumping straight to D 3.0, we could spend a 
little time developing a 64-bit compiler for D. It would give us a 
stable and unstable branch like we have already. And developers could 
update their code for D 2.0 and 64-bit hardware at the same time.
Besides, I don't think we should start development on D 3.0 so soon. I 
think it's better if a language doesn't evolve too quickly. It gives 
time for developers to get settled in and write stable code for D, 
without the need to update it every couple of years.
Nov 01 2008
next sibling parent reply dsimcha <dsimcha yahoo.com> writes:
== Quote from Xinok (xnknet gmail.com)'s article

 Robert Fraser wrote:

 Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


 Most people.


 Since the topic is here, I'd like to make a suggestion for when to make
 a 64-bit compiler.
 I think the right time is when D 2.0 is stable and we begin to move away
 from D 1.0. Instead of jumping straight to D 3.0, we could spend a
 little time developing a 64-bit compiler for D. It would give us a
 stable and unstable branch like we have already. And developers could
 update their code for D 2.0 and 64-bit hardware at the same time.
 Besides, I don't think we should start development on D 3.0 so soon. I
 think it's better if a language doesn't evolve too quickly. It gives
 time for developers to get settled in and write stable code for D,
 without the need to update it every couple of years.


Given that LDC is going to be released any day now, and should support 64-bit,
any
estimates of how long it will take for LDC to support D2?  I work with huge
datasets and occasionally run up against the 2 gig address space barrier of 32
bit
on a computer with 4+ gigs of RAM.  Therefore, I for one would really find such
a
thing useful.
Nov 01 2008
parent Christian Kamm <kamm-incasoftware removethis.de> writes:
 Given that LDC is going to be released any day now, and should support
 64-bit


I have to correct that. While we hope to release LDC for x86-32 Linux soon
after LLVM 2.4 is released, x86-64 (and, in fact, any other platform) still
needs work. 

K. Wilson has contributed some big fixes for x86-64 Linux recently (he made
inline asm and exception handling work! yay for him!) and reports that
things basically work, but there's still lots of testing - and most likely
bugfixing - to be done before it's ready.

I've already said it about LDC x86-32 Windows, but it's the same here: we
could really use a few more dedicated x86-64 people!



 any estimates of how long it will take for LDC to support D2?


Neither Tomas nor me are particularly interested in chasing D2 right now.
Providing a cross-platform D1 compiler and fixing bugs there seems more
useful at the moment.
Nov 01 2008
prev sibling parent "Denis Koroskin" <2korden gmail.com> writes:
On Sat, 01 Nov 2008 16:09:39 +0300, Xinok <xnknet gmail.com> wrote:


 Robert Fraser wrote:

 Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


  Most people.


 Since the topic is here, I'd like to make a suggestion for when to make  
 a 64-bit compiler.
 I think the right time is when D 2.0 is stable and we begin to move away  
 from D 1.0. Instead of jumping straight to D 3.0, we could spend a  
 little time developing a 64-bit compiler for D. It would give us a  
 stable and unstable branch like we have already. And developers could  
 update their code for D 2.0 and 64-bit hardware at the same time.
 Besides, I don't think we should start development on D 3.0 so soon. I  
 think it's better if a language doesn't evolve too quickly. It gives  
 time for developers to get settled in and write stable code for D,  
 without the need to update it every couple of years.


Walter is concentrated on the language (specs and a reference  
implementation).
There is, however, a compiler at works that will have 64bit codegen one  
day:

     http://www.dsource.org/projects/ldc
Nov 01 2008
prev sibling next sibling parent =?UTF-8?B?QWxleGFuZGVyIFDDoW5law==?= writes:
Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


I do.
Nov 01 2008
prev sibling next sibling parent "Denis Koroskin" <2korden gmail.com> writes:
On Sat, 01 Nov 2008 15:41:16 +0300, Daniel Ribeiro Maciel  
<daniel.maciel gmail.com> wrote:


 Who still use a 32-bit computer these days anyways.


I bet most Windows users are using 32bit version of Windows, even if their  
CPUs are x64-capable.
Nov 01 2008
prev sibling next sibling parent KennyTM~ <kennytm gmail.com> writes:
Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


I use 32bit Windows ('cause it is bundled) and 64bit Linux.

BTW, gdc is 64 bit (x86-64), though it only supports D1.
Nov 01 2008
prev sibling parent reply Sean Kelly <sean invisibleduck.org> writes:
Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


Donald Knuth does.  He's complained about not being able to leverage the 
additional registers in a 64-bit CPU without having to use 64-bit 
pointers though.  See:

http://www-cs-staff.stanford.edu/~uno/news.html

"A Flame About 64-bit Pointers"


Sean
Nov 01 2008
parent reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
Sean Kelly wrote:

 Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


 
 Donald Knuth does.  He's complained about not being able to leverage the 
 additional registers in a 64-bit CPU without having to use 64-bit 
 pointers though.  See:
 
 http://www-cs-staff.stanford.edu/~uno/news.html
 
 "A Flame About 64-bit Pointers"


What a coincidence. Only yesterday I read that following a link from 
reddit about Knuth giving up sending checks as rewards for bug 
discovery. (It turns out that many people put photocopies of his checks 
on the Web as bragging rights, and a bug in the banking system allows 
anyone who sees an ABA + check number to easily withdraw money from that 
account.)

Andrei
Nov 01 2008
parent Sean Kelly <sean invisibleduck.org> writes:
Andrei Alexandrescu wrote:

 Sean Kelly wrote:

 Daniel Ribeiro Maciel wrote:

 Who still use a 32-bit computer these days anyways.


 Donald Knuth does.  He's complained about not being able to leverage 
 the additional registers in a 64-bit CPU without having to use 64-bit 
 pointers though.  See:

 http://www-cs-staff.stanford.edu/~uno/news.html

 "A Flame About 64-bit Pointers"


 
 What a coincidence. Only yesterday I read that following a link from 
 reddit about Knuth giving up sending checks as rewards for bug 
 discovery. (It turns out that many people put photocopies of his checks 
 on the Web as bragging rights, and a bug in the banking system allows 
 anyone who sees an ABA + check number to easily withdraw money from that 
 account.)


I read it for the same reason :-)  Funny that it would be relevant today.


Sean
Nov 01 2008