Getting a GHC repository got easier: new instructions

Simon Marlow marlowsd at
Mon Apr 26 10:51:34 EDT 2010

On 26/04/2010 15:31, Max Bolingbroke wrote:
> On 26 April 2010 13:42, Simon Marlow<marlowsd at>  wrote:
>> comments welcome!
> This is great news. However,  I had some problems:
> 1) darcs-all does not add --lazy by default, which contradicts the wiki:
> """
> $ ./darcs-all --testsuite get
> warning: adding --partial, to override use --complete
> warning: . already present; omitting
> == running darcs get --partial
> ghc-tarballs
> --partial: hashed or darcs-2 repository detected, using --lazy instead
> Finished getting.
> """

This doesn't cause any problems - I left it that way on purpose, because 

> 2) Once I added --lazy, I found that the binary repo is not in hashed format:
> """
> == running darcs get --lazy
> libraries/binary
> This is GHC's branch of the main binary repo.
> NB. DO NOT push new patches to this repo, instead push patches
> upstream first.  The main binary repo is at
> Only push patches that are also in main repo to this branch.  Only
> push once the patches pass GHC's validate script.
> Before releasing GHC, the branch must be synced with a released
> upstream version of binary.
> **********************
> Converting old-fashioned repository to hashed format...
> *******************************************************************************
> Fetching a hashed repository would be faster.  Perhaps you could persuade
> the maintainer to run darcs optimize --upgrade with darcs 2.4.0 or higher?
> *******************************************************************************
> Finished getting.
> """

... this happens :-)  But as you discovered, this is not actually a 
problem either.  There was some problem with converting the binary 
package and Ian has been in touch with the maintainers about it, so 
we'll straighten this out in due course.

> Apart from those issues, it worked great: it's certainly a damn sight
> faster than it used to be! Since I'm on (case-insensitive) OS X, this
> is also the first time I've been able to branch the GHC repo without
> using "cp", which is excellent news :-)

Yep, definitely a step forwards.


More information about the Glasgow-haskell-users mailing list