1 |
Ciaran McCreesh napsal(a): |
2 |
> On Sat, 05 May 2007 15:15:55 +0200 |
3 |
> Jakub Moc <jakub@g.o> wrote: |
4 |
>> How does it matter exactly whether it's paludis users, gcc users, php |
5 |
>> users, apache users or whoever else who will use the news item? Let's |
6 |
>> not misuse news framework for stuff that |
7 |
> |
8 |
> Er, that's one of the main points of GLEP 42's design: it allows |
9 |
> delivery of news items that are important to a subset of users only to |
10 |
> that subset. The size of the subset is entirely irrelevant. |
11 |
|
12 |
That's not what I've been pointing out at all, you've completely snipped |
13 |
the important part about *unintended* use of this feature. So, once |
14 |
again - this is not an elog replacement and is not intended for trivial |
15 |
stuff (see the 'critical news' in GLEP42 title). |
16 |
|
17 |
>> For this particular case, you can (and do even) advise specifically |
18 |
>> the affected users only on runtime that the config files syntax has |
19 |
>> been changed. (Frankly, you don't even need elog stuff for similar |
20 |
>> things; everyone concerned will get the message paludis spits on the |
21 |
>> screen when they try to use it). |
22 |
> |
23 |
> Experience and user feedback has shown that in situations like this |
24 |
> users want an accompanying news item even if the application does output |
25 |
> deprecation warnings. |
26 |
|
27 |
Well again, what kind of experience? You cannot just carry over an |
28 |
existing practice of heavily abusing such stuff in a particular overlay |
29 |
with a couple of ebuilds and implant it into a tree with thousands of |
30 |
packages - it will produce loads of annoying noise and the feature will |
31 |
become useless if every maintainer starts to use news framework in a |
32 |
similar way. |
33 |
|
34 |
|
35 |
-- |
36 |
Best regards, |
37 |
|
38 |
Jakub Moc |
39 |
mailto:jakub@g.o |
40 |
GPG signature: |
41 |
http://subkeys.pgp.net:11371/pks/lookup?op=get&search=0xCEBA3D9E |
42 |
Primary key fingerprint: D2D7 933C 9BA1 C95B 2C95 B30F 8717 D5FD CEBA 3D9E |
43 |
|
44 |
... still no signature ;) |