Re: [PATCH] guix: refresh: Use bags.

Eric Bavier skribis: > Ludovic Courtès writes: > >> Eric Bavier skribis: >> >>> Ludovic Courtès writes: >>> >>> From 1d22367e0806cea004631e22a782b7db3ffe65b0 Mon Sep 17 00:00:00 2001 >>> From: Eric Bavier >>> Date: Mon, 13 Oct 2014 13:46:09 -0500 >>> Subject: [PATCH] guix: refresh: Use bags. >>

Re: [PATCH] guix: refresh: Use bags.

Ludovic Courtès writes: > Eric Bavier skribis: > >> Ludovic Courtès writes: >> >> From 1d22367e0806cea004631e22a782b7db3ffe65b0 Mon Sep 17 00:00:00 2001 >> From: Eric Bavier >> Date: Mon, 13 Oct 2014 13:46:09 -0500 >> Subject: [PATCH] guix: refresh: Use bags. >> >> * guix/packages.scm (bag-dire

Re: [PATCH] guix: refresh: Use bags.

Eric Bavier skribis: > Ludovic Courtès writes: > >> Eric Bavier skribis: >> >>> This patch may need to go in core-updates. >> >> Why? > > Would changes in gnu/packages.scm not trigger rebuilds for most > packages? No, because this file is just a “host-side” module, it is not used as an input to

Re: [PATCH] guix: refresh: Use bags.

Ludovic Courtès writes: > Eric Bavier skribis: > >> This patch may need to go in core-updates. > > Why? Would changes in gnu/packages.scm not trigger rebuilds for most packages? > It would be nice to show this info on the output of > build-aux/list-packages.scm (used to build >

Re: [PATCH] guix: refresh: Use bags.

Eric Bavier skribis: > Ludovic Courtès writes: > >> Eric Bavier skribis: >> >>> Ludovic Courtès writes: [...] >> Of course it would be good to use the full DAG. For that we’d need >> something that uses ‘fold-packages’ to get the “entry points” of the >> DAG, and then traverses it all to also

Re: [PATCH] guix: refresh: Use bags.

Ludovic Courtès writes: > Eric Bavier skribis: > >> Ludovic Courtès writes: >> >>> I’m surprised gzip, diffutils, patch, and tar are very low, even though >>> they’re part of %final-inputs. Conversely, bzip2 (also part of >>> %final-inputs) is higher. >>> >>> Any idea? >> >> I think this is bec

Re: [PATCH] guix: refresh: Use bags.

Eric Bavier skribis: > Ludovic Courtès writes: > >> I’m surprised gzip, diffutils, patch, and tar are very low, even though >> they’re part of %final-inputs. Conversely, bzip2 (also part of >> %final-inputs) is higher. >> >> Any idea? > > I think this is because fold-packages only looks at packa

Re: [PATCH] guix: refresh: Use bags.

Ludovic Courtès writes: > I’m surprised gzip, diffutils, patch, and tar are very low, even though > they’re part of %final-inputs. Conversely, bzip2 (also part of > %final-inputs) is higher. > > Any idea? I think this is because fold-packages only looks at packages that are exported by a module

Re: [PATCH] guix: refresh: Use bags.

On Oct 10, 2014 3:42 PM, "Ludovic Courtès" wrote: > > Eric Bavier skribis: > > > This patch allows the `guix refresh --list-dependent` option to be more > > accurate in its dependency reporting, mostly by taking into account > > implicit dependencies generated at the build-system level (thanks to

Re: [PATCH] guix: refresh: Use bags.

Eric Bavier skribis: > This patch allows the `guix refresh --list-dependent` option to be more > accurate in its dependency reporting, mostly by taking into account > implicit dependencies generated at the build-system level (thanks to our > new wonderful bags). Cool! > From ebc348a37addd53bab2

Re: [PATCH] guix: refresh: Use bags.

Eric Bavier writes: > Out of curiosity, I also produced this table of packages and their total > number of dependent packages. Something similar might be useful for us, > e.g., when deciding whether a package update should go in master or > core-updates. > Whoa, that's neat! Thanks for sharing