1 |
On Fri, Jan 5, 2018 at 6:55 PM, Michał Górny <mgorny@g.o> wrote: |
2 |
|
3 |
> W dniu pią, 05.01.2018 o godzinie 23∶09 +0100, użytkownik Kristian |
4 |
> Fiskerstrand napisał: |
5 |
> > On 01/05/2018 10:28 PM, Aaron W. Swenson wrote: |
6 |
> > > On 2018-01-05 15:16, William Hubbs wrote: |
7 |
> > > > If we have a default expiration, it should be one year after the date |
8 |
> > > > posted to go along with our current policy of not supporting things |
9 |
> that |
10 |
> > > > are older than a year. |
11 |
> > > > |
12 |
> > > > William |
13 |
> > > |
14 |
> > > I thought it was three years. |
15 |
> > > |
16 |
> > > At any rate, I think a year is too short. |
17 |
> > > |
18 |
> > > How about 18 months? |
19 |
> > > |
20 |
> > |
21 |
> > I might sound like a broken CD here, but why define the expiration as |
22 |
> > part of the news format instead of specifying it in the package manager |
23 |
> > as a user defined variable? Various use cases requires different |
24 |
> > treatment, so leaving it up to user seems more relevant to me, and we |
25 |
> > could allow information to be presented as part of stages to give a hint |
26 |
> > for what dates to look for? |
27 |
> > |
28 |
> |
29 |
> To be honest, I kinda agree with Kristian here. I feel like this header |
30 |
> isn't going to work well. |
31 |
> |
32 |
> While the idea may initially sound good, I'm afraid we'll have the usual |
33 |
> developer split here: some developers will set very long times, some |
34 |
> will set very short times, some will not care and just copy some random |
35 |
> value (default, the value from some other news item). In the end, users |
36 |
> will end up seeing very old news items from dev X, while newer items |
37 |
> from dev Y will disappear. |
38 |
> |
39 |
> So yes, I think having a single predefined time is better, |
40 |
> for consistency at least. And allowing user to adjust that time based |
41 |
> on his own is certainly better than making it only dev-settable. |
42 |
> |
43 |
|
44 |
I'll swerve right then and say this: |
45 |
antarus@nspawntest /var/lib/gentoo/news $ cat news-gentoo.unread |
46 |
2013-09-27-initramfs-required |
47 |
2014-06-15-gcc48_ssp |
48 |
2014-10-26-gcc_4_7_introduced_new_c++11_abi |
49 |
2015-02-04-portage-sync-changes |
50 |
2015-07-25-python-targets |
51 |
2015-08-13-openssh-weak-keys |
52 |
2015-10-22-gcc-5-new-c++11-abi |
53 |
2016-06-23-l10n-use_expand |
54 |
2017-11-30-new-17-profiles |
55 |
|
56 |
Here are my news items. |
57 |
|
58 |
initramfs-required hits every Gentoo box, has 0 Display restrictions, and |
59 |
will never ever ever go away => delete it. |
60 |
gcc48_ssp: Display-If-Installed: >=sys-devel/gcc-4.8.3 (and it includes |
61 |
basically every major arch.) It will also never ever ever ever go away. |
62 |
Either delete it or cap the gcc version. |
63 |
gcc_4_7: Same boat as above, either cap the gcc version or nuke it. |
64 |
portage: display-If-Installed: sys-apps/portage, will display on ever box. |
65 |
So cap that as well. |
66 |
python-targets: has no Display-If Headers, it will display on every box. |
67 |
Delete it or cap it. |
68 |
openssh_weak_keys: Display-If-Installed: net-misc/openssh. This looks like |
69 |
a bug (only people who were on < 7.0 care about it. |
70 |
gcc-5: Display-If-Installed: >=sys-devel/gcc-5 => another uncapped news |
71 |
item. |
72 |
l10n: Another news item with no Display-If-* headers, it too will live |
73 |
forever. |
74 |
new-17-profiles: another >=gcc-X, will never go away. |
75 |
|
76 |
So 3 items have no display headers and are not fixable without adding a new |
77 |
header. Many others can be fixed by capping Display-If-Installed versions. |
78 |
|
79 |
-- Aside, I'm not totally convinced modifying the entries will work based |
80 |
on a bug in portage-news. I'll poke the implementation more. |
81 |
|
82 |
-A |
83 |
|
84 |
|
85 |
> |
86 |
> -- |
87 |
> Best regards, |
88 |
> Michał Górny |
89 |
> |
90 |
> |
91 |
> |