| 1 |
On Fri, Feb 12, 2016 at 1:36 AM, Kent Fredric <kentfredric@×××××.com> wrote: |
| 2 |
> On 12 February 2016 at 18:56, Duncan <1i5t5.duncan@×××.net> wrote: |
| 3 |
>> So my USE="-* ..." (without letting portage do autounmasking) would |
| 4 |
>> continue to work just like it does now, correct? |
| 5 |
> |
| 6 |
> I would hope so. |
| 7 |
|
| 8 |
That would be my proposal. |
| 9 |
|
| 10 |
> And obviously, this feature would be potentially |
| 11 |
> tenous, and might be wise to only |
| 12 |
> activate its mechanics with a parameter to emerge. |
| 13 |
>,,, |
| 14 |
> That way, people who do want the new behaviour can stick it in the |
| 15 |
> default emerge options and \o/ |
| 16 |
|
| 17 |
Whether you call it a FEATURE or an option it would almost certainly |
| 18 |
have to be defaulted. |
| 19 |
|
| 20 |
If you run with that option for six months and then run emerge without |
| 21 |
that option while using something like --newuse, it will immediately |
| 22 |
detect a bazillion dependency conflicts and if using autounmask it |
| 23 |
will want to add 1000 lines to your package.use - which would be the |
| 24 |
same 1000 lines that many are running with today. |
| 25 |
|
| 26 |
I think it would make more sense in general for this to be the default |
| 27 |
and to have a flag to disable it, in part for this reason. It |
| 28 |
wouldn't affect people running -* and such anyway, so this is targeted |
| 29 |
mostly at users who don't care a great deal about micromanaging their |
| 30 |
USE flags. |
| 31 |
|
| 32 |
-- |
| 33 |
Rich |
Archives