1 |
On Mon, 24 Jul 2006 21:22:20 +0200 |
2 |
Marius Mauch <genone@g.o> wrote: |
3 |
|
4 |
> > The problem with having these in the tree is that they should be |
5 |
> > able to take advantage of the unique features of each package |
6 |
> > manager. Otherwise, they really are just eclasses. |
7 |
> |
8 |
> "unique features" such as? What exact benefits does this hooks system |
9 |
> give you over using an eclass? |
10 |
|
11 |
The biggest one is the ability to clean up after myself if the emerge |
12 |
fails (if the fail hooks idea I mentioned in my previous mail makes |
13 |
sense, at least). |
14 |
|
15 |
Basically, what I need to do is: |
16 |
|
17 |
- Run a script to intelligently add / manage package manager-added |
18 |
users. At the moment, it would seem this should just be run before |
19 |
or during the pkg_setup phase, but this may well be sub-optimal, |
20 |
particularly in the case if binary packages or non-standard ROOTs. |
21 |
|
22 |
- Properly let that set of scripts know when an added user or group is |
23 |
safe to remove. This code is intimately tied with the |
24 |
implementation-specific details of the above scripts, which may |
25 |
change from version to version, hence my trepidation about adding |
26 |
this code directly into portage itself. |
27 |
|
28 |
- Have the ability to clean up from myself if a build goes awry, or |
29 |
is aborted by the operator. As far as I know, this particular |
30 |
requirement would exclude the use of eclasses, but I could well be |
31 |
wrong. |
32 |
|
33 |
- Have this user addition handled in an intelligent way when we're |
34 |
only building/installing a binary package. I'm not really sure on |
35 |
what would be a logical way to do this. |
36 |
|
37 |
So, if the answer to the above is to not use some sort of hooks, then I |
38 |
would appreciate pointers on where to look to achieve all of the above |
39 |
with portage. |
40 |
|
41 |
Thanks for all your patience with me. |
42 |
|
43 |
-- |
44 |
Mike Kelly |