| 1 |
On Mon, 2006-01-09 at 18:58 -0800, Drake Wyrm wrote: |
| 2 |
> Lares Moreau <lares.moreau@×××××.com> wrote: |
| 3 |
> > media-gfx/graphicsmagick:gs - enable ghostscript support |
| 4 |
> > media-gfx/imagemagick:gs - enable ghostscript support |
| 5 |
> > media-libs/urt:gs - Add support for postscript |
| 6 |
> > |
| 7 |
> > Looking in these ebuilds, all: |
| 8 |
> > gs? ( virtual/ghostscript ) |
| 9 |
> |
| 10 |
> For how many of these might it make sense to use "ps - enable postscript |
| 11 |
> support" rather than "gs - enable ghostscript support"? If I understand |
| 12 |
> correctly, the consensus is that use flags should reflect the capability |
| 13 |
> enabled rather than the name of the library or package which provides |
| 14 |
> that capability. |
| 15 |
|
| 16 |
In doing some poking around and talking to a couple more devs, It seems |
| 17 |
that the _unofficial_ consensus is that there be 5 or more packages that |
| 18 |
use a particular flag before it is to move to global. There are |
| 19 |
exception as always. Previously there has been great debate about where |
| 20 |
the proverbial line is. Wherever the line, it must be posted to -dev |
| 21 |
before the flag is committed to global. |
| 22 |
|
| 23 |
And as for the flag reflecting the capability, you are absolutely |
| 24 |
correct. The urt package seems to have mislabelled the entry in |
| 25 |
use.local.desc. |
| 26 |
|
| 27 |
I'm just letting this global use die a quiet death. |
| 28 |
-- |
| 29 |
Lares Moreau <lares.moreau@×××××.com> | LRU: 400755 http://counter.li.org |
| 30 |
lares/irc.freenode.net | |
| 31 |
Gentoo x86 Arch Tester | ::0 Alberta, Canada |
| 32 |
Public Key: 0D46BB6E @ subkeys.pgp.net | Encrypted Mail Preferred |
| 33 |
Key fingerprint = 0CA3 E40D F897 7709 3628 C5D4 7D94 483E 0D46 BB6E |
Archives