| 1 |
On Mon, 2005-10-24 at 21:37 -0700, Donnie Berkholz wrote: |
| 2 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 3 |
> Hash: SHA1 |
| 4 |
> |
| 5 |
> Spider (D.m.D. Lj.) wrote: |
| 6 |
> | If your package, libFoo, installs .h files that directly require header |
| 7 |
> | files from libBar, then you have a Runtime dependency on libBar, not |
| 8 |
> | only a compile time dependency |
| 9 |
> | |
| 10 |
> | |
| 11 |
> | Why? Because libFoo should be usable at runtime, even for compiling |
| 12 |
> | things not managed by our tree. And if BazPack has a configure check, |
| 13 |
> | requires explicitly libFoo, tests for the presence of said header, and |
| 14 |
> | all checks out.. Then BazPack should be able to work, and libFoo's |
| 15 |
> | header certainly shouldn't be broken for use in compile. |
| 16 |
> |
| 17 |
|
| 18 |
> Now, the other side of the story. It's not true runtime dependence |
| 19 |
> because it's not required for programs to run, only to compile. And the |
| 20 |
> way I see it, things required for programs to compile are by definition |
| 21 |
> DEPEND rather than RDEPEND. |
| 22 |
|
| 23 |
exactly. |
| 24 |
|
| 25 |
> The consequences of the two sides are like this, from what I can see: |
| 26 |
> |
| 27 |
> 1) Headers are run-time and build-time deps |
| 28 |
> |
| 29 |
> - - Headers have to be installed even when you're using purely binary |
| 30 |
> packages, because they are supposedly needed at "runtime" for your |
| 31 |
> packages to work. |
| 32 |
> |
| 33 |
> - - Also, header packages can't be uninstalled after the build via |
| 34 |
> depclean because they're specified as run-time dependencies. |
| 35 |
> |
| 36 |
|
| 37 |
|
| 38 |
> 2) Headers are build-time deps only |
| 39 |
> |
| 40 |
> - - Binary packages don't require the header packages. |
| 41 |
> |
| 42 |
> - - Header packages can be unmerged after builds. |
| 43 |
> |
| 44 |
> - - Packages requiring the headers have to DEPEND on them directly, |
| 45 |
> because DEPENDs don't cascade. (Although this brings to mind the concept |
| 46 |
> of some sort of cascadable DEPEND.) |
| 47 |
> |
| 48 |
> |
| 49 |
> I'd like to hear what some other people think about this. |
| 50 |
|
| 51 |
Well put Donnie. |
| 52 |
|
| 53 |
-- |
| 54 |
|
| 55 |
emerge -B virtual/libc busybox |
| 56 |
ROOT=/mnt/foo emerge -K virtual/libc busybox |
| 57 |
|
| 58 |
# if this were to pull in linux-headers on the second step into the |
| 59 |
$ROOT then all embedded support would be shot. |
| 60 |
|
| 61 |
+1 vote to ignore Spiders mail and revert such changes that put headers |
| 62 |
in rdeps as it has undesirable side effects to existing supported |
| 63 |
projects. |
| 64 |
|
| 65 |
|
| 66 |
-- |
| 67 |
Ned Ludd <solar@g.o> |
| 68 |
|
| 69 |
-- |
| 70 |
gentoo-dev@g.o mailing list |
Archives