| 1 |
Enrico Weigelt wrote: |
| 2 |
> > > say I have very lean kernel on the target, how I can compile |
| 3 |
> > > against the target kernel? |
| 4 |
> > |
| 5 |
> > This is a non-issue for most packages because the kernel interface |
| 6 |
> > can be considered stable. |
| 7 |
> |
| 8 |
> No, they can't, especially in embedded world. |
| 9 |
|
| 10 |
Obviously there is a limit to every definition of stable. But it is |
| 11 |
my impression that Linux does not change API at a whim. |
| 12 |
|
| 13 |
|
| 14 |
> The biggest problem is certain ill-designed packages which try to |
| 15 |
> guess something from the *running* system. This can only be solved |
| 16 |
> in the source. |
| 17 |
|
| 18 |
Sure, and this is a problem with individual packages. Not so much |
| 19 |
with the kernel. The packages can be fixed. |
| 20 |
|
| 21 |
|
| 22 |
> > Do you have a problem with some package? |
| 23 |
> |
| 24 |
> Just from the tip of my head, in recent years: network utils, |
| 25 |
> drbd, etc, ... |
| 26 |
|
| 27 |
Because the recommended API to use was changed, or? |
| 28 |
|
| 29 |
|
| 30 |
//Peter |
Archives