[#3861] super — ts <decoux@...>
[#3862] Marshal.dump'ing OpenStruct objects — Mauricio Fern疣dez <batsman.geo@...>
Hi,
[#3881] mkdir, mkdir_p in FileUtils and mode — Florian Frank <flori@...>
Hello,
[#3907] Obtaining mode information on an IO object — Jos Backus <jos@...>
The attached patch implements IO#mode. This method returns the mode the IO
Hi,
On Tue, Dec 07, 2004 at 09:25:13AM +0900, nobu.nokada@softhome.net wrote:
Jos Backus wrote:
Hi,
On Thu, Dec 09, 2004 at 10:47:48AM +0900, nobu.nokada@softhome.net wrote:
On Thu, Dec 09, 2004 at 02:40:33PM +0900, James Britt wrote:
[#3914] Pathname needs a makeover — "Berger, Daniel" <Daniel.Berger@...>
Hi all,
[#3922] Incorrect escaping in strings produced by String::inspect — noreply@...
Bugs item #1173, was opened at 2004-12-08 17:35
[#3966] unknown node type 0 — Andrew Walrond <andrew@...>
I still get this happening a lot with my Rubyx linux ruby script.
This is a long standing bug in Ruby, and has been reported hundreds of times
Hi,
[#3975] Patches to test/unit — Ryan Davis <ryand-ruby@...>
I believe these are the minimal patches needed to make it possible to
[#3982] Win32: rb_sys_fail() - errno == 0 — Florian Gro<florgro@...>
Moin!
[#4000] 1.8.2 preview4 — Yukihiro Matsumoto <matz@...>
Hello,
[#4009] cgi.rb -- more GET/POST stuff — mde@...26.com
First of all, I think it would be great, as Eustaquio suggests, to
GETs and POSTs are defined to be fairly different actions. I'd read
-----BEGIN PGP SIGNED MESSAGE-----
Francis Hwang wrote:
-----BEGIN PGP SIGNED MESSAGE-----
First of all, the entire discussion of when GET is appropriate
mde@state26.com wrote:
[#4027] Allowing custom number literal suffixes? — Florian Gro<florgro@...>
Moin!
Hi,
Mathieu Bouchard wrote:
Mathieu Bouchard wrote:
I'm not sure I would advocate making Ruby's grammar even more
>
Brent Roman wrote:
> Brent Roman wrote:
Brent Roman wrote:
> Florian Gross wrote:
Mathieu Bouchard wrote:
Mathieu Bouchard wrote:
[#4033] Garbage collection trouble — Christian Neukirchen <chneukirchen@...>
Hello,
>>>>> "C" == Christian Neukirchen <chneukirchen@gmail.com> writes:
ts <decoux@moulon.inra.fr> writes:
>>>>> "C" == Christian Neukirchen <chneukirchen@gmail.com> writes:
[#4040] Extensions, Internal — Jgen Mangler <juergen.mangler@...>
Hi,
Re: mkmf enhancement for Win32
On Sat, 11 Dec 2004 09:32:42 +0900, nobu.nokada@softhome.net
<nobu.nokada@softhome.net> wrote:
> At Sat, 11 Dec 2004 07:10:54 +0900,
> Berger, Daniel wrote in [ruby-core:03944]:
>> It turns out that some Win32 functions, such as AttachConsole(),
>> are only conditionally available, and depend on the values of
>> specific macros.
>>
[...]
> It doesn't seem meaningful to check the compiling environment.
> Rather you should check the target, the environment to run that
> extension, no?
Yes, it should check the target environment as a native win32
compile, whether done by mingw32 or a Linux cross-compiler or Visual
C++, because the problem exists in the Windows API header files.
>> This will be more of an issue when 64-bit Windows becomes more
>> popular, though I am even hitting it currently with some of the
>> Win32Utils extensions I've been working on.
>>
>> Now, we could leave it up to individual extension authors to set
>> this themselves manually, but that's a pain, and error prone
>> (it's easy to forget, and some may not know to do it in the first
>> place). I thought it would be nice if mkmf would set it for us.
>> So, I submit the following patch for consideration:
> I guess that developers who intend to restrict the target
> should add such macros by themselves.
I disagree completely. The developers may not be intending to
restrict the target, but are using features that appear to be
available in the Win32 API and are often silently not available
based on the version of this _WIN32_WINNT or _WIN32_WINDOWS value.
Additionally, differing behind-the-scenes implementations may be
made present depending on this value (oddly enough). At work, we
have had to set _WIN32_WINNT to 0x0501 in order to get the proper
version of Winsock2.
This is no different than setting optimisation levels based on the
compiler accepts. Or maybe setting macros to ensure that BSD-style
semantics are followed.
Now, what I'll actually suggest is that this value should be set in
the compiling of Ruby, not in mkmf and that this value should be
used by mkmf from config.rb. That way, extensions are (by default)
compiled to match the expectations of the compiled version of Ruby.
If, then, the extension maintainer wants to compile to a different
Windows target, it can be done easily.
-austin
--
Austin Ziegler * halostatue@gmail.com
* Alternate: austin@halostatue.ca