1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA1 |
3 |
|
4 |
On 2009.06.07 10:34, Ulrich Mueller wrote: |
5 |
> >>>>> On Sun, 07 Jun 2009, Steven J Long wrote: |
6 |
> |
7 |
> > I'd just like to know what the implications would be for users if |
8 |
> we |
9 |
> > kept the .ebuild extension, and a new PMS were rolled out stating |
10 |
> > that the mangler were allowed to find the EAPI without sourcing |
11 |
> (and |
12 |
> > giving the restrictions) once portage 2.2 was stable, or the |
13 |
> ability |
14 |
> > to handle this backported to 2.1.6, and issued in a release? |
15 |
> |
16 |
> Even if we do only a one-time change of the file extension, can we |
17 |
> ever get rid of the old extension? Or are we then stuck with two |
18 |
> extensions in the tree until the end of time? |
19 |
> |
20 |
> Let's assume for the moment that we change from ".ebuild" to ".eb". |
21 |
> Then we obviously cannot change all ebuilds in the tree to ".eb", |
22 |
> otherwise old Portage versions would see an empty tree and there |
23 |
> would |
24 |
> be no upgrade path. |
25 |
> |
26 |
> Or am I missing something? |
27 |
> |
28 |
> Ulrich |
29 |
|
30 |
First, lets be clear that the upgrade path related problems are driven |
31 |
by the EAPI change, not the .ebuild to .eb change. That serves only to |
32 |
allow the new EAPI to be introduced immedately and hide the change from |
33 |
older package managers |
34 |
|
35 |
The upgrade path issue remains even if we do the usual Gentoo introduce |
36 |
new features to the package manager then wait a year or so before they |
37 |
are deployed. Without the .ebuild to .eb change. late upgraders will |
38 |
get the error messages in Appendix A of the GLEP when these features |
39 |
are relesed. |
40 |
|
41 |
To keep an upgrade path, package managers and their dependencies need |
42 |
to use EAPI 0, 1 or 2. (EAPI 3 is not yet out) How far back should the |
43 |
upgrade path extend? |
44 |
|
45 |
As you suggest, even with .ebuild to .eb change.its not a clean break |
46 |
with the past but would happen over time, with ebuilds for new package |
47 |
versions being migrated to the new format. For example we would |
48 |
have |
49 |
foo-2.1-r1.ebuild |
50 |
foo-3.0.eb |
51 |
portage-2.3.ebuild |
52 |
portage-3.0.eb |
53 |
|
54 |
Portage-2.3 will understand the later EAPI but will itself, use only |
55 |
EAPI, 0, 1 or 2. |
56 |
|
57 |
With the .ebuild to .eb change. users who do not upgrade portage will |
58 |
not see newer versions of foo. Without it, they will get the error |
59 |
messages in Appendix A of the GLEP. This will have the side effect of |
60 |
driving the adoption of the newer package managers. |
61 |
|
62 |
It is not suffcient to have portage-2.3.ebuild as understanding later |
63 |
EAPI files. All of its dependencies need to keep to EAPI 0, 1 or 2. |
64 |
These must be the last packages to migrate to later EAPIs. |
65 |
|
66 |
Thats just portage. The same reasoning applies to any other package |
67 |
manager and there are at least three. This begs the question of how |
68 |
friendly do we want to be to derivitive distros that use our tree but |
69 |
their own package manager? |
70 |
|
71 |
Upgrade path issues need to be addressed in the GLEP. I will add |
72 |
something like the above in the next version. |
73 |
|
74 |
- -- |
75 |
Regards, |
76 |
|
77 |
Roy Bamford |
78 |
(NeddySeagoon) a member of |
79 |
gentoo-ops |
80 |
forum-mods |
81 |
treecleaners |
82 |
trustees |
83 |
|
84 |
|
85 |
|
86 |
-----BEGIN PGP SIGNATURE----- |
87 |
Version: GnuPG v2.0.11 (GNU/Linux) |
88 |
|
89 |
iEYEARECAAYFAkoryrUACgkQTE4/y7nJvashtQCgkO+feHG2BBGLOTObrwq72iOx |
90 |
nI4AoNZ67Mhq4yEKYTzfBOPmu98Py9Mc |
91 |
=GB62 |
92 |
-----END PGP SIGNATURE----- |