| 1 |
On Sun, Oct 15, 2006 at 08:22:01PM +0100, Ciaran McCreesh wrote: |
| 2 |
> On Sun, 15 Oct 2006 15:09:32 -0400 Mike Frysinger <vapier@g.o> |
| 3 |
> wrote: |
| 4 |
> | On Sunday 15 October 2006 14:16, Ciaran McCreesh wrote: |
| 5 |
> | > On Sun, 15 Oct 2006 02:09:58 -0400 Mike Frysinger |
| 6 |
> | > <vapier@g.o> |
| 7 |
> | > | what are you talking about ? the point of having per-package |
| 8 |
> | > | defaults is so that you can enable a flag by default in one |
| 9 |
> | > | package only |
| 10 |
> | > |
| 11 |
> | > package != ebuild. |
| 12 |
> | > |
| 13 |
> | > | to take the oss example, we would want to remove that from our |
| 14 |
> | > | profiles and only enable it by default on say libsdl |
| 15 |
> | > |
| 16 |
> | > That's removing it from one place and adding it to five others. |
| 17 |
> | |
| 18 |
> | ... and you can have different default USE for some versions so if a |
| 19 |
> | feature changes between them, you can easily control it |
| 20 |
> |
| 21 |
> Which is very much not the normal case, so it's not worth requiring |
| 22 |
> five times as much duplication just for the occasional time when it is |
| 23 |
> necessary. |
| 24 |
|
| 25 |
Ebuilds already have a boatload of duplication; bit of a red herring |
| 26 |
however complaining about a single char in IUSE to indicate a flag |
| 27 |
defaults to on (seriously, you're bitching about 5 chars of wasted |
| 28 |
space for a single flag forced on, switch to a better arguement). |
| 29 |
|
| 30 |
~harring |
Archives