www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - DMD 1.072 and DMD 2.057 64bit on Mac OS X

reply Jacob Carlborg <doob me.com> writes:
I just downloaded both 1.072 and 2.057. I see that 2.057 is compiled as 
a 64bit binary and 1.072 as a 32bit binary. Is there a reason for this 
difference?

-- 
/Jacob Carlborg
Dec 14 2011
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 12/14/2011 1:35 AM, Jacob Carlborg wrote:
 I just downloaded both 1.072 and 2.057. I see that 2.057 is compiled as a 64bit
 binary and 1.072 as a 32bit binary. Is there a reason for this difference?
They should have both been 64 bit. Probably something wrong with my build script. My intent is to go all 64 bit for the OS X binaries. Apple stopped shipping 32 bit only Macs several years ago. While I intend to still build and test the 32 bit binaries, I see no point in shipping them.
Dec 14 2011
parent reply Jacob Carlborg <doob me.com> writes:
On 2011-12-14 10:39, Walter Bright wrote:
 On 12/14/2011 1:35 AM, Jacob Carlborg wrote:
 I just downloaded both 1.072 and 2.057. I see that 2.057 is compiled
 as a 64bit
 binary and 1.072 as a 32bit binary. Is there a reason for this
 difference?
They should have both been 64 bit. Probably something wrong with my build script. My intent is to go all 64 bit for the OS X binaries. Apple stopped shipping 32 bit only Macs several years ago. While I intend to still build and test the 32 bit binaries, I see no point in shipping them.
Yeah, I see no point with the 32bit binaries either. But the 64bit binaries target 64bit by default, that's how I noticed it, my code wasn't completely 64bit compatible. -- /Jacob Carlborg
Dec 14 2011
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 12/14/2011 2:03 AM, Jacob Carlborg wrote:
 Yeah, I see no point with the 32bit binaries either. But the 64bit binaries
 target 64bit by default, that's how I noticed it, my code wasn't completely
 64bit compatible.
The default is set to match gcc's default.
Dec 14 2011
parent reply Jacob Carlborg <doob me.com> writes:
On 2011-12-14 11:13, Walter Bright wrote:
 On 12/14/2011 2:03 AM, Jacob Carlborg wrote:
 Yeah, I see no point with the 32bit binaries either. But the 64bit
 binaries
 target 64bit by default, that's how I noticed it, my code wasn't
 completely
 64bit compatible.
The default is set to match gcc's default.
Yeah, it's the only reason this matters on Mac OS X. If it was a "regular" application I probably never would have noticed. -- /Jacob Carlborg
Dec 14 2011
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 12/14/2011 2:37 AM, Jacob Carlborg wrote:
 Yeah, it's the only reason this matters on Mac OS X. If it was a "regular"
 application I probably never would have noticed.
I find I don't notice. (I think that's good!) Unless I'm debugging an issue specific to 32 or 64 bits. It makes me wonder if we need to support 32 bit generation on OSX at all.
Dec 14 2011
next sibling parent reply =?ISO-8859-1?Q?Alex_R=F8nne_Petersen?= <xtzgzorex gmail.com> writes:
On 14-12-2011 11:54, Walter Bright wrote:
 On 12/14/2011 2:37 AM, Jacob Carlborg wrote:
 Yeah, it's the only reason this matters on Mac OS X. If it was a
 "regular"
 application I probably never would have noticed.
I find I don't notice. (I think that's good!) Unless I'm debugging an issue specific to 32 or 64 bits. It makes me wonder if we need to support 32 bit generation on OSX at all.
No point maintaining something that won't be used. I would also imagine that it can't be long before Windows stops supporting 32-bit. - Alex
Dec 14 2011
next sibling parent reply "Vladimir Panteleev" <vladimir thecybershadow.net> writes:
On Wednesday, 14 December 2011 at 11:07:22 UTC, Alex Rønne 
Petersen wrote:
 No point maintaining something that won't be used. I would also 
 imagine that it can't be long before Windows stops supporting 
 32-bit.
Windows still supports 16-bit apps (even 64-bit versions do, for some rare exceptions)!
Dec 14 2011
parent reply =?UTF-8?B?QWxleCBSw7hubmUgUGV0ZXJzZW4=?= <xtzgzorex gmail.com> writes:
On 14-12-2011 14:11, Vladimir Panteleev wrote:
 On Wednesday, 14 December 2011 at 11:07:22 UTC, Alex Rønne Petersen wrote:
 No point maintaining something that won't be used. I would also
 imagine that it can't be long before Windows stops supporting 32-bit.
Windows still supports 16-bit apps (even 64-bit versions do, for some rare exceptions)!
Hehe, fun fact. Though I doubt Microsoft actually cares enough to maintain that support these days... - Alex
Dec 14 2011
next sibling parent Walter Bright <newshound2 digitalmars.com> writes:
On 12/14/2011 8:51 AM, Alex Rønne Petersen wrote:
 On 14-12-2011 14:11, Vladimir Panteleev wrote:
 On Wednesday, 14 December 2011 at 11:07:22 UTC, Alex Rønne Petersen wrote:
 No point maintaining something that won't be used. I would also
 imagine that it can't be long before Windows stops supporting 32-bit.
Windows still supports 16-bit apps (even 64-bit versions do, for some rare exceptions)!
Hehe, fun fact. Though I doubt Microsoft actually cares enough to maintain that support these days...
I rely on that to test the 16 bit support of Digital Mars C/C++. It's the only modern compiler that supports the 16 bit x86 world, and there are a few customers that like it. I.e. you don't have to get a compiler from some "museum".
Dec 14 2011
prev sibling parent Kai Meyer <kai unixlords.com> writes:
On 12/14/2011 09:51 AM, Alex Rønne Petersen wrote:
 On 14-12-2011 14:11, Vladimir Panteleev wrote:
 On Wednesday, 14 December 2011 at 11:07:22 UTC, Alex Rønne Petersen
 wrote:
 No point maintaining something that won't be used. I would also
 imagine that it can't be long before Windows stops supporting 32-bit.
Windows still supports 16-bit apps (even 64-bit versions do, for some rare exceptions)!
Hehe, fun fact. Though I doubt Microsoft actually cares enough to maintain that support these days... - Alex
I work for a company who has large enough contracts in Brazil who still run Windows NT to justify the continued use of Visual Studio 2005 to build software for NT and XP (though most of our Windows developers develop in VS2010 and then make sure it still works in VS2005). Apple has the opportunity to depreciate entire architectures because they aren't as ubiquitous as Microsoft products. They don't have billion dollar contracts that request they spend an extra 200 million a year to support their old stuff. At least that's the only justification I can think of to stay on older software.
Dec 14 2011
prev sibling parent reply Somedude <lovelydear mailmetrash.com> writes:
Le 14/12/2011 11:07, Alex Rønne Petersen a écrit :
 
 No point maintaining something that won't be used. I would also imagine
 that it can't be long before Windows stops supporting 32-bit.
 
 - Alex
You imagine wrong. A large percentage of Windows users are still on Win XP 32 bits (like me).
Dec 15 2011
parent reply =?ISO-8859-1?Q?Alex_R=F8nne_Petersen?= <xtzgzorex gmail.com> writes:
On 15-12-2011 18:04, Somedude wrote:
 Le 14/12/2011 11:07, Alex Rønne Petersen a écrit :
 No point maintaining something that won't be used. I would also imagine
 that it can't be long before Windows stops supporting 32-bit.

 - Alex
You imagine wrong. A large percentage of Windows users are still on Win XP 32 bits (like me).
Amount of users has nothing to do with whether Microsoft will continue to support it. Microsoft has well-defined lifecycles for support on their products. I'm not sure what you being on a 32-bit OS from 2001 has to do with my imagination of the future in 2011. :) There are many reasons a company like Microsoft would want to let 32-bit x86 die; only having to maintain one platform being one of such. It is undeniable that the trend is going towards 64-bit (and this is a Good Thing, obviously). - Alex
Dec 15 2011
parent "Nick Sabalausky" <a a.a> writes:
"Alex Rønne Petersen" <xtzgzorex gmail.com> wrote in message 
news:jcdep3$2gs3$1 digitalmars.com...
 being on a 32-bit OS from 2001
https://www.semitwist.com/articles/article/view/why-use-a-10-year-old-os-!
Dec 16 2011
prev sibling next sibling parent reply Jacob Carlborg <doob me.com> writes:
On 2011-12-14 11:54, Walter Bright wrote:
 On 12/14/2011 2:37 AM, Jacob Carlborg wrote:
 Yeah, it's the only reason this matters on Mac OS X. If it was a
 "regular"
 application I probably never would have noticed.
I find I don't notice. (I think that's good!) Unless I'm debugging an issue specific to 32 or 64 bits. It makes me wonder if we need to support 32 bit generation on OSX at all.
Yeah, I wonder that too. But does it hurt/cause problems to do? I mean, it's already supported. -- /Jacob Carlborg
Dec 14 2011
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 12/14/2011 3:09 AM, Jacob Carlborg wrote:
 On 2011-12-14 11:54, Walter Bright wrote:
 It makes me wonder if we need to support 32 bit generation on OSX at all.
Yeah, I wonder that too. But does it hurt/cause problems to do? I mean, it's already supported.
There's the ongoing effort to support dynlibs.
Dec 14 2011
next sibling parent reply Jacob Carlborg <doob me.com> writes:
On 2011-12-14 19:28, Walter Bright wrote:
 On 12/14/2011 3:09 AM, Jacob Carlborg wrote:
 On 2011-12-14 11:54, Walter Bright wrote:
 It makes me wonder if we need to support 32 bit generation on OSX at
 all.
Yeah, I wonder that too. But does it hurt/cause problems to do? I mean, it's already supported.
There's the ongoing effort to support dynlibs.
But as Michel mentioned, iPhone is 32bit and I think that is a good point. -- /Jacob Carlborg
Dec 14 2011
parent Michel Fortin <michel.fortin michelf.com> writes:
On 2011-12-15 07:22:47 +0000, Jacob Carlborg <doob me.com> said:

 On 2011-12-14 19:28, Walter Bright wrote:
 On 12/14/2011 3:09 AM, Jacob Carlborg wrote:
 On 2011-12-14 11:54, Walter Bright wrote:
 It makes me wonder if we need to support 32 bit generation on OSX at
 all.
Yeah, I wonder that too. But does it hurt/cause problems to do? I mean, it's already supported.
There's the ongoing effort to support dynlibs.
But as Michel mentioned, iPhone is 32bit and I think that is a good point.
And I believe DMD can already create executables that would work on the iPhone simulator, since those are simply 32-bit Mac OS X executables linked to a different set of frameworks. -- Michel Fortin michel.fortin michelf.com http://michelf.com/
Dec 15 2011
prev sibling parent reply Michel Fortin <michel.fortin michelf.com> writes:
On 2011-12-14 18:28:07 +0000, Walter Bright <newshound2 digitalmars.com> said:

 On 12/14/2011 3:09 AM, Jacob Carlborg wrote:
 On 2011-12-14 11:54, Walter Bright wrote:
 It makes me wonder if we need to support 32 bit generation on OSX at all.
Yeah, I wonder that too. But does it hurt/cause problems to do? I mean, it's already supported.
There's the ongoing effort to support dynlibs.
It's one thing to ask developers to use 64-bit machines for development, it's another to say to developers they can't target 32-bit users if they choose to use DMD. Even Apple ships everything with dual architecture binaries these days, and some of Apple's apps also keep PowerPC support, iTunes for instance. And also iOS is 32-bit. -- Michel Fortin michel.fortin michelf.com http://michelf.com/
Dec 15 2011
parent Walter Bright <newshound2 digitalmars.com> writes:
On 12/15/2011 4:40 AM, Michel Fortin wrote:
 It's one thing to ask developers to use 64-bit machines for development, it's
 another to say to developers they can't target 32-bit users if they choose to
 use DMD. Even Apple ships everything with dual architecture binaries these
days,
 and some of Apple's apps also keep PowerPC support, iTunes for instance.
I think it's fair to continue with 32 bit dev support for now.
 And also iOS is 32-bit.
I didn't know that.
Dec 15 2011
prev sibling parent reply Michel Fortin <michel.fortin michelf.com> writes:
On 2011-12-14 10:54:55 +0000, Walter Bright <newshound2 digitalmars.com> said:

 On 12/14/2011 2:37 AM, Jacob Carlborg wrote:
 Yeah, it's the only reason this matters on Mac OS X. If it was a "regular"
 application I probably never would have noticed.
I find I don't notice. (I think that's good!) Unless I'm debugging an issue specific to 32 or 64 bits. It makes me wonder if we need to support 32 bit generation on OSX at all.
If I were to develop a Mac application in D at this time, not being able to target older 32-bit macs would make me reconsider using D. The app I'm working on still has PowerPC support! Perhaps I'm not the typical Mac developer though. But as long as Apple supports compiling for 32-bit with the provided developer tools I'd keep it in DMD too, if only because it'd make D look cheep otherwise. Also, I think it'd make sense that druntime and phobos continue to support 32-bit OS X in case someone wants to target iOS one day, which is basically 32-bit OS X on ARM. -- Michel Fortin michel.fortin michelf.com http://michelf.com/
Dec 14 2011
parent Jacob Carlborg <doob me.com> writes:
On 2011-12-14 14:39, Michel Fortin wrote:
 Also, I think it'd make sense that druntime and phobos continue to
 support 32-bit OS X in case someone wants to target iOS one day, which
 is basically 32-bit OS X on ARM.
Never thought of that, it's a good point. -- /Jacob Carlborg
Dec 14 2011