1 |
On 01/28/2018 08:34 PM, Alec Warner wrote: |
2 |
> |
3 |
> |
4 |
> On Sun, Jan 28, 2018 at 8:10 PM, Zac Medico <zmedico@g.o |
5 |
> <mailto:zmedico@g.o>> wrote: |
6 |
> |
7 |
> On 01/28/2018 04:17 PM, Michael Orlitzky wrote: |
8 |
> > Since ::gentoo is the only repository governed by the PMS, can't we make |
9 |
> > repoman do this? The problem with requiring "repoman --force" for an |
10 |
> > in-place dependency change is that repoman generally won't have access |
11 |
> > to the unedited ebuild; but for ::gentoo, we can probably hack something |
12 |
> > together in git. Have it source the old and new ebuilds, and compare |
13 |
> > their dependency lists. |
14 |
> |
15 |
> Yes I'm in favor of that, but it doesn't accomplish my current goal, as |
16 |
> I'll explain below. |
17 |
> |
18 |
> > It won't work outside of a git repo, but it will work in the one place |
19 |
> > that really matters. |
20 |
> |
21 |
> My goal for this patch is to provide a safety net for users, such that |
22 |
> they'll be automatically notified that the --changed-deps option is |
23 |
> available when needed to solve problems involving stale dependencies. |
24 |
> The reason behind this goal is that git commits flow directly to rsync |
25 |
> without any restriction, which makes users vulnerable to stale |
26 |
> dependencies, in the form of failed dependency calculations. |
27 |
> |
28 |
> |
29 |
> What I would go along with is: |
30 |
> |
31 |
> 1) If we can determine the depgraph failed and |
32 |
|
33 |
If we use depgraph failure as a condition, then we're ignoring |
34 |
potentially missed updates due to stale dependencies. So, it's safer to |
35 |
show the message regardless of depgraph failure, and that's what it does. |
36 |
|
37 |
> 2) We can determine that dependencies changed. |
38 |
|
39 |
Yes, _changed_deps_report returns silently if there are no changed |
40 |
dependencies. Also, it returns silently with --quiet if none of the |
41 |
packages with changed dependencies are in the graph. |
42 |
|
43 |
Actually, if none of the packages with changed dependencies are in the |
44 |
graph, then we can be silent even without --quiet because the emphasis |
45 |
on filing bugs has been removed in PATCH v2. So, in PATCH v3, I'll make |
46 |
it silent regardless of --quiet in this case. |
47 |
|
48 |
> 3) Show a NOTE telling users about --changed-deps=y |
49 |
|
50 |
This is in the HINT section, which is displayed if both --changed-deps |
51 |
and --dynamic-deps are disabled in PATCH v2. |
52 |
|
53 |
> |
54 |
> But I'm not sure the patch does this today? I think it will trigger too |
55 |
> aggressively. But ultimately I'm not against helping users work around this |
56 |
> sort of breakage. |
57 |
|
58 |
Do my responses above sound reasonable? |
59 |
-- |
60 |
Thanks, |
61 |
Zac |