| 1 |
lör 2012-09-01 klockan 13:46 +0200 skrev Amadeusz Żołnowski: |
| 2 |
> I am definitely for (1). The extracted old internal Genkernel initramfs |
| 3 |
> generator would need to be named differently for either case, for |
| 4 |
> example: geninitrd, geninitramfs, genifs, genird or geninitfs. |
| 5 |
> |
| 6 |
> |
| 7 |
|
| 8 |
Whatever you do, do NOT call it anything that reminds of initrd. |
| 9 |
Genkernel has not been able to generate initrd/ramdisk-images since 2008 |
| 10 |
and Dracut has never have had that functionality. And people are really |
| 11 |
confused over what is what, so please, keep use only initramfs to keep |
| 12 |
confusion to a bare minimum. |
| 13 |
|
| 14 |
When it comes to the rest: |
| 15 |
You make som valid points. I am convinced that we need to refactor the |
| 16 |
code in Genkernel, but I am not convinced that we need the split. |
| 17 |
|
| 18 |
And on a side note: I really think that when someone thinks about adding |
| 19 |
a option to Genkernel they should read Documents/kernel-parameters.txt |
| 20 |
from the linux kernel tree. Using anything not documented there without |
| 21 |
a really good reason (like there is no option there for what we try to |
| 22 |
archieve) only creates more need for documentation. I also think that we |
| 23 |
should work towards compability with that text-file to ensure that we do |
| 24 |
not need one document for Genkernel and one for all other cases if it is |
| 25 |
not necessary... |
| 26 |
|
| 27 |
Regards |
| 28 |
Peter Hjalmarsson |
Archives