simonmar at microsoft.com
Tue Jun 28 07:18:40 EDT 2005
On 28 June 2005 08:58, John Meacham wrote:
> On Tue, Jun 28, 2005 at 03:00:46AM +0100, Claus Reinke wrote:
>>> I'd like to second this - the Win32 problems in 6.4 are preventing
>>> several projects from moving forward. Any word on 6.4.1? The major
>>> FFI bug in 6.4 is holding back some projects of mine.
>> if you have projects depending on this, I assume you are aware of the
>> nightly development snapshots and build logs (for both HEAD and
>> STABLE; not many platforms, but including linux and windows/mingw)?
>> http://www.haskell.org/ghc/download.html (bottom)
>> (nightly snapshots for windows have not always been available, but
>> that they are, thanks to Sigbjorn, it would be good to make use of
>> them -
>> not just to get rid of bugs in the last release, but to avoid bugs
>> in the next one)
> Yeah, I should have said that I have projects waiting to be released
> until people can compile with a standard version of ghc. Asking people
> to install a cvs snapshot is not something I want to do. :)
You can work around the FFI problem - just use a C function to return
the pointer value instead of foreign import "&". You can even inline
the C function and compile via-C to get the same code.
I had in mind a timescale of 1-2 months for 6.4.1 - there's various
other stuff on my queue first. We want to have a performance shakedown
for one thing, and I'd really like 6.4.1 to have GHCi on x86-64.
More information about the Glasgow-haskell-users