Graham wrote
| > I note that the GHC kit for Windows includes some of the (I presume) | > MinGW | > binutil components, but does not include 'ar'. I find that 'ar' is | > needed | > for building HaXml. May I suggest that 'ar' also be included in this | > kit?
Sigbjorn says
| Adding 'ar' is clearly trivial, but the tools that GHC installers | include aren't meant to be used by the public, only by GHC | to avoid having to rely on external tools when compiling& | linking (e.g., 'gcc' and 'ld' don't work as-is, but need the right | -B setting). I'm not strongly against it, but adding other dev | tools does have potential support costs
I'm inclined to agree with Sigbjorn here. Where would we stop? GHC does include some Mingw stuff "in its belly", but you are not supposed to know or care how GHC does its work. You certainly aren't expected to invoke those tools directly. If you want anything more than GHC, I think the best thing is to get Mingw or Cygwin or SFU.
Fair point. My counter argument would be that the 'ar' utility is (I think) needed to create a GHC package (from just Haskell source files). Am I right about this? Maybe a useful touchstone might be: would be it appropriate to be able create a GHC package using *just* the GHC command?
OTOH, if the GHC package structure is to be displaced by the new library infrastructure project, then I'd fully agree with you.
Here's a possible simple compromise: in the documentation of GHC packages, include a link to the MinGW kit [1] that is needed to build a package on windows? (It took me some time to track down what I needed.)
#g --
[1] http://www.mingw.org/ http://www.mingw.org/download.shtml http://prdownloads.sf.net/mingw/MinGW-3.1.0-1.exe?download
------------ Graham Klyne For email: http://www.ninebynine.org/#Contact
_______________________________________________ Glasgow-haskell-users mailing list [EMAIL PROTECTED] http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
