Re: [bug-gsrc] Recommended way of generating sha256sum

[Henrik Sandklef]

On 11/01/2010 10:06 PM, Brian Gough wrote:
> At Mon, 01 Nov 2010 20:53:22 +0100,
> Henrik Sandklef wrote:
> > I wouldn't mind the option of being able to test BOTH one alpha AND one
> > "real" release. I'd prefer one Makefile (to rule them all) without
> > conditionals, e.g
> >
> > GARNAME = xnee
> >
> > GARVERSION = 3.07
> > GARVERSION_ALPHA = 3.07.91
> >
> > MASTER_SITES = $(MASTER_GNU)
> > MASTER_SITES_ALPHA = $(MASTER_GNU_ALPHA)
> >
> > MASTER_SUBDIR = xnee/
> > MASTER_SUBDIR_ALPHA = xnee-alpha/
>
> I agree it's nicer to have it in one file but I'm not sure how to make
> it work, because the Makefiles are hard-coded to use the variables
> GARVERSION etc and there are many of them so I am not sure how to
> translate the _ALPHA version to be used.

Thought so, but wanted to express my wishes just in case they were 
possible to (be easily) implement(ed).

> What can work is to create a separate directory tree
>
>     alpha/xnee/Makefile:
>
>     override GARVERSION = 3.07.91
>     override MASTER_SITES = $(MASTER_GNU_ALPHA)
>     override MASTER_SUBDIR = xnee/
>
>     include ../../gnu/xnee/Makefile
>
> and then make -C alpha/xnee

That'll work fine with me.

> I've just checked in a test version of that if you want to try it.

I'll do it asap. Some more work in prep for tomorrow's work first.

Is the code on the trunk?


/hesa