| 1 |
On Thu, Jul 26, 2012 at 4:44 PM, Peter Alfredsen |
| 2 |
<peter.alfredsen@×××××.com> wrote: |
| 3 |
> On Wed, Jul 11, 2012 at 11:04 PM, Michał Górny <mgorny@g.o> wrote: |
| 4 |
>> On Wed, 11 Jul 2012 15:27:41 -0400 |
| 5 |
>> Mike Gilbert <floppym@g.o> wrote: |
| 6 |
>> |
| 7 |
>>> Personally, I think a consolidated systemd/udev package is the best |
| 8 |
>>> way to go here. |
| 9 |
>> |
| 10 |
>> A consolidated package means that: |
| 11 |
>> |
| 12 |
>> - every change made by udev developers would have to be reviewed by |
| 13 |
>> systemd team to make sure it doesn't break systemd. udev developers |
| 14 |
>> don't use systemd; |
| 15 |
>> - every change made by systemd developers would have to be reviewed by |
| 16 |
>> udev team to make sure it doesn't break openrc. systemd developers |
| 17 |
>> usually don't run openrc; |
| 18 |
>> - udev developers will force me to use eclasses they like and force |
| 19 |
>> their coding style on me; |
| 20 |
>> - i will force eclasses I like and my coding style on udev developers; |
| 21 |
>> - new udev wouldn't be able to be stabilized without systemd being |
| 22 |
>> stabilized at the same time (and I don't really think systemd is in |
| 23 |
>> any condition to go stable), |
| 24 |
>> - there will be a few random flags which will either work or not, |
| 25 |
>> depending on a state of magical switch flag, |
| 26 |
>> - and after all, the ebuild will be basically one big use-conditional. |
| 27 |
> |
| 28 |
> So, since this is blocking up development and people actually solving |
| 29 |
> things, could we just have virtual/udev and be done with it? Upstream |
| 30 |
> obviously reneged on their promise to make the component parts |
| 31 |
> buildable separately, so the virtual seems like the only sane choice |
| 32 |
> right now. |
| 33 |
|
| 34 |
Just to clarify, udev/systemd never promised "to make the component |
| 35 |
parts buildable separately". They promised: |
| 36 |
|
| 37 |
"we will be supporting this for a long time since it is a necessity to |
| 38 |
make initrds (which lack systemd) work properly. Distributions not |
| 39 |
wishing to adopt systemd can build udev pretty much the same way as |
| 40 |
before, however should then use the systemd tarball instead of the |
| 41 |
udev tarball and package only what is necessary of the resulting |
| 42 |
build." |
| 43 |
|
| 44 |
Where "package only what is necessary" being the important part for Gentoo. |
| 45 |
|
| 46 |
http://lwn.net/Articles/490413/ |
| 47 |
|
| 48 |
Certainly they don't care about source-based distributions like |
| 49 |
Gentoo, but they never promised "to make the component parts buildable |
| 50 |
separately". |
| 51 |
|
| 52 |
Anyway, I also support the virtual/udev, since it's the only way for |
| 53 |
us systemd users to not build udev twice. |
| 54 |
|
| 55 |
Regards. |
| 56 |
-- |
| 57 |
Canek Peláez Valdés |
| 58 |
Posgrado en Ciencia e Ingeniería de la Computación |
| 59 |
Universidad Nacional Autónoma de México |
Archives