Is there any chance that GHC 6.04, or any GHC version in the next few months, will support the following on Windows: 1) use of native threads so that the world won't be stopped every time you wait for a character; 2) the new system-independent runProcess interface; 3) Network.Socket. (Maybe it does already?) I have a feeling the answer will be "Yes, if you implement it", but sadly my time is limited ...
Is there any chance that GHC 6.04, or any GHC version in the next few months, will support the following on Windows: 1) use of native threads so that the world won't be stopped every time you wait for a character;
Should already be in 6.02.1 - add the -threaded flag when linking, and you'll get the "threaded rts" which uses native thread to make sure that (safe) ffi calls don't block. Waiting for a character sounds like a job for the standard input routines, and they should support concurrency even without the threaded RTS, but I don't know about the windows version. Cheers, Wolfgang
Not sure what 2) refers to, but 1) and 3) is already in 6.2.x;
no need to use the heavier-weight -threaded stuff to enable
it.
--sigbjorn
----- Original Message -----
From: "George Russell"
Is there any chance that GHC 6.04, or any GHC version in the next few months, will support the following on Windows: 1) use of native threads so that the world won't be stopped every time you wait for a character; 2) the new system-independent runProcess interface; 3) Network.Socket. (Maybe it does already?)
I have a feeling the answer will be "Yes, if you implement it", but sadly my time is limited ...
Sigbjorn Finne wrote:
Not sure what 2) refers to, but 1) and 3) is already in 6.2.x; no need to use the heavier-weight -threaded stuff to enable it.
What I mean is a Windows version of Simon M's: http://www.haskell.org//pipermail/libraries/attachments/20040317/609414e2/Pr... or alternatively http://tinyurl.com/2yx2g I'm glad that (1) (world not stopping on IO) and (3) (networking) are already present in 6.2.x though, since it increases the chance of a uniform interface to processes working on Windows. Is there any chance of this in 6.4?
participants (3)
-
George Russell -
Sigbjorn Finne -
Wolfgang Thaller