1 |
On Thu, 2020-12-17 at 16:08 -0500, Mike Gilbert wrote: |
2 |
> On Thu, Dec 17, 2020 at 3:50 PM Mike Gilbert <floppym@g.o> |
3 |
> wrote: |
4 |
> > |
5 |
> > On Thu, Dec 17, 2020 at 3:38 PM William Hubbs <williamh@g.o> |
6 |
> > wrote: |
7 |
> > > |
8 |
> > > On Sat, Dec 12, 2020 at 09:22:06PM -0500, Mike Gilbert wrote: |
9 |
> > > > On Sat, Dec 12, 2020 at 9:09 PM William Hubbs |
10 |
> > > > <williamh@g.o> wrote: |
11 |
> > > > > |
12 |
> > > > > On Sat, Dec 12, 2020 at 04:25:48PM -0500, Mike Gilbert wrote: |
13 |
> > > > > > On Sat, Dec 12, 2020 at 3:48 PM William Hubbs |
14 |
> > > > > > <williamh@g.o> wrote: |
15 |
> > > > > > > If both /usr/bin/python and /usr/bin/python3 are going |
16 |
> > > > > > > away, the best |
17 |
> > > > > > > choice would be to add functionality to python-exec or |
18 |
> > > > > > > eselect python to tell us |
19 |
> > > > > > > the path to the default python interpretor. Once we know |
20 |
> > > > > > > that we call it |
21 |
> > > > > > > directly. |
22 |
> > > > > > |
23 |
> > > > > > I don't think they are "going away". There is a USE flag on |
24 |
> > > > > > dev-lang/python-exec that makes them optional, and I think |
25 |
> > > > > > it will be |
26 |
> > > > > > forcibly enabled for the foreseeable future. |
27 |
> > > > > > |
28 |
> > > > > > > Please do not apply this patch to meson; I think we can |
29 |
> > > > > > > figure something |
30 |
> > > > > > > out that is better. |
31 |
> > > > > > |
32 |
> > > > > > I think installing a small script to help translate |
33 |
> > > > > > arguments from one |
34 |
> > > > > > format to another is a reasonable solution. |
35 |
> > > > > |
36 |
> > > > > I think we should look at the eclass to see if we can |
37 |
> > > > > provide functions |
38 |
> > > > > that can be used by consumers to handle this. |
39 |
> > > > |
40 |
> > > > I don't really understand what you mean by this. I am |
41 |
> > > > converting one |
42 |
> > > > internal bash function into an external script so that its |
43 |
> > > > python |
44 |
> > > > dependencies can be better defined and managed. |
45 |
> > > |
46 |
> > > What I mean is, ebuilds should not be calling _meson_env_array at |
47 |
> > > all |
48 |
> > > since it is defined and documented as an eclass internal |
49 |
> > > function. |
50 |
> > > |
51 |
> > > I would like to know more about what the gallium-nine-standalone |
52 |
> > > ebuild |
53 |
> > > is doing and why it needs to call a meson.eclass internal |
54 |
> > > function. |
55 |
> > > |
56 |
> > > On the other hand, if _meson_env_array is meant to be called by |
57 |
> > > ebuilds, |
58 |
> > > we need to rename it and improve the documentation for it in the |
59 |
> > > eclass. |
60 |
> > |
61 |
> > I do not really care about gallium-nine-standalone and its abuse of |
62 |
> > the private _meson_env_array function. That's an issue that should |
63 |
> > be |
64 |
> > separate from the change I am proposing. I am only touching the |
65 |
> > ebuild |
66 |
> > because my patch removes the _meson_env_array function and I want |
67 |
> > to |
68 |
> > avoid breaking the ebuild. |
69 |
> > |
70 |
> > > > > Also, I don't think your script will run if native-symlinks |
71 |
> > > > > is disabled since in |
72 |
> > > > > that setting /usr/bin/python would not exist. |
73 |
> > > > |
74 |
> > > > python_doscript updates the shebang before installing the |
75 |
> > > > script. |
76 |
> > > |
77 |
> > > Ok, I didn't know python_doscript does this, but couldn't we |
78 |
> > > just |
79 |
> > > change line 129 in the eclass to "python3 -c ..."? |
80 |
> > |
81 |
> > No, that will not help in any way. |
82 |
> > |
83 |
> > > > > I question the value of the native-symlinks use flag on |
84 |
> > > > > python-exec |
85 |
> > > > > unless there is a way to query the path of the default python |
86 |
> > > > > interpretor. |
87 |
> > > > |
88 |
> > > > Regardless, I don't see how that makes my solution a bad thing. |
89 |
> > > > It |
90 |
> > > > ensures that the code will be executed by a |
91 |
> > > > known/support/tested |
92 |
> > > > version of python. |
93 |
> > > > |
94 |
> > > |
95 |
> > > I'm not sure how useful the script is as a command, so I don't |
96 |
> > > think it |
97 |
> > > should be installed that way, but I do want to hear more about |
98 |
> > > this, |
99 |
> > > both from you and chewi. :-) |
100 |
> > |
101 |
> > I don't see any reasonable way to make it work otherwise. If you |
102 |
> > have |
103 |
> > no better suggestion, please refrain from further comments. |
104 |
> |
105 |
> I gave it a little more thought, and I suppose we could use "eselect |
106 |
> python show" to get a valid python interpreter. I'll put together a |
107 |
> patch for that. |
108 |
> |
109 |
|
110 |
Do you want all meson packages to depend on eselect-python? |
111 |
|
112 |
-- |
113 |
Best regards, |
114 |
Michał Górny |