| 1 |
El lun, 29-07-2013 a las 21:07 -0400, Rich Freeman escribió: |
| 2 |
> On Mon, Jul 29, 2013 at 9:01 PM, Dustin C. Hatch <admiralnemo@×××××.com> wrote: |
| 3 |
> > I think the point is that users may have an initramfs (that they built |
| 4 |
> > manually or using some tool besides dracut or genkernel) that makes use of |
| 5 |
> > cryptsetup/lvm2 built statically, or perhaps they just like it that way, so |
| 6 |
> > why take away that ability and make them change? |
| 7 |
> |
| 8 |
> Presumably because it is harder to maintain? If somebody wants to |
| 9 |
> maintain (proxy or otherwise) the needed changes to support the static |
| 10 |
> USE flag my opinion is that they should be able to do so. They would |
| 11 |
> need to be responsive on bugs/etc and not be a burden on the other |
| 12 |
> maintainers. |
| 13 |
> |
| 14 |
> However, if nobody wants to step up and do the work, then those who |
| 15 |
> are doing the work basically get the last word in how it gets done. |
| 16 |
> That's just how things roll around here. |
| 17 |
> |
| 18 |
> Besides, you could make the same argument about every binary in |
| 19 |
> /(s)bin. Initramfs builders manage to deal with a dynamically-linked |
| 20 |
> bash, so they should be able to handle lvm+cryptsetup. |
| 21 |
> |
| 22 |
> Rich |
| 23 |
> |
| 24 |
> |
| 25 |
|
| 26 |
It also causes some problems (some of them broke during udev updated |
| 27 |
from, for example, 200 to 204): |
| 28 |
https://bugs.gentoo.org/buglist.cgi?quicksearch=lvm2% |
| 29 |
20static&list_id=1914334 |
| 30 |
https://bugs.gentoo.org/buglist.cgi?quicksearch=cryptsetup% |
| 31 |
20static&list_id=1914332 |
Archives