1 |
On 27/07/14 16:47, Micha³ Górny wrote: |
2 |
> Dnia 2014-07-27, o godz. 14:42:24 |
3 |
> Samuli Suominen <ssuominen@g.o> napisa³(a): |
4 |
> |
5 |
>> On 26/07/14 15:49, Ciaran McCreesh wrote: |
6 |
>>> On Sat, 26 Jul 2014 12:41:16 +0000 (UTC) |
7 |
>>> Martin Vaeth <martin@×××××.de> wrote: |
8 |
>>>> hasufell <hasufell@g.o> wrote: |
9 |
>>>>> Dynamics deps are already broken, not consistently enabled (e.g. |
10 |
>>>>> when subslots are in use) |
11 |
>>>> Just to make it clear: No, dynamic deps are not broken. |
12 |
>>> Yes they are. |
13 |
>> We just succesfully converted ~300 ebuilds in tree without revision |
14 |
>> bumps from virtual/udev[gudev,introspection,static-libs] |
15 |
>> to virtual/libudev and virtual/libgudev |
16 |
>> Tested it on multiple boxes, went fine. |
17 |
> How did you exactly test them? |
18 |
|
19 |
That you could still emerge packages, even world, without Portage |
20 |
complaining about unsatisfied |
21 |
deps |
22 |
|
23 |
> Did you at least bother checking if |
24 |
> portage actually uses the deps you added? |
25 |
> |
26 |
|
27 |
When you `cd /var/db/pkg` and run `grep virtual/udev.*gudev |
28 |
*/*/*.ebuild`, you can indeed still see some: |
29 |
|
30 |
app-cdr/xfburn-0.5.2/xfburn-0.5.2.ebuild: udev? ( virtual/udev[gudev] )" |
31 |
gnome-base/gvfs-1.20.2/gvfs-1.20.2.ebuild: virtual/udev[gudev] ) |
32 |
media-gfx/gimp-2.8.10-r1/gimp-2.8.10-r1.ebuild: udev? ( |
33 |
virtual/udev[gudev] )" |
34 |
sys-fs/udisks-1.0.5-r1/udisks-1.0.5-r1.ebuild: >=virtual/udev-208[gudev] |
35 |
sys-fs/udisks-2.1.3/udisks-2.1.3.ebuild: |
36 |
>=virtual/udev-${UDEV_VERSION}[gudev] |
37 |
virtual/libgudev-208/libgudev-208.ebuild:# $Header: |
38 |
/var/cvsroot/gentoo-x86/virtual/libgudev/libgudev-208.ebuild,v 1.7 |
39 |
2014/06/18 20:55:21 mgorny Exp $ |
40 |
xfce-extra/thunar-volman-0.8.0/thunar-volman-0.8.0.ebuild: |
41 |
virtual/udev[gudev] |
42 |
|
43 |
But if you try to emerge these packages, like, for example: |
44 |
|
45 |
$ sudo emerge -pv xfburn thunar-volman |
46 |
|
47 |
These are the packages that would be merged, in reverse order: |
48 |
|
49 |
Calculating dependencies... done! |
50 |
[ebuild R ] xfce-extra/thunar-volman-0.8.0 USE="libnotify -debug" 0 kB |
51 |
[ebuild R ] app-cdr/xfburn-0.5.2 USE="udev -debug -gstreamer" 0 kB |
52 |
|
53 |
Portage is using the fresh copies from gentoo-x86 |
54 |
|
55 |
I'm _not_ a Portage, the package manager, developer, so I'd really |
56 |
appericiate some *examples* where this would become a problem, binary |
57 |
packages is known one, we have lived with that problem since forever, so |
58 |
something else, please. |
59 |
For everything I do with Portage, this is fine with me, and I expect |
60 |
it's fine for the vast majority of users as well. |
61 |
And this majority of users won't appericiate the suggested solution of |
62 |
"lets always revision bump, despite of triggering rebuild for everyone" |
63 |
|
64 |
To clarify, I know dynamic deps is not perfect, but it's the best |
65 |
solution we have implemented to avoid the rebuilding problem, and long |
66 |
as that's true... And seems like you, yourself, have thought about the |
67 |
same issue: |
68 |
http://bugs.gentoo.org/show_bug.cgi?id=486778 |
69 |
As in, you can't skip that bug, and go directly to |
70 |
http://bugs.gentoo.org/show_bug.cgi?id=516612 |
71 |
|
72 |
- Samuli |
73 |
|
74 |
(Excuse my grammar, woke up five minutes ago, to make some coffee now...) |