1 |
On Mon, 2004-03-01 at 21:34, Jean-Francois Patenaude wrote: |
2 |
> Ok for the 'typos' kind of changes ... maybe I had exagerated my examples. |
3 |
> But still, it would be nice to have some level of criticality related |
4 |
> to updates. |
5 |
|
6 |
There is already a group working on the ability to ONLY merge in |
7 |
critical updates which have a GLSA released for them. |
8 |
|
9 |
emerge sync |
10 |
emerge glsa |
11 |
|
12 |
You simply don't do a world update until your maintenance window. |
13 |
|
14 |
> A highly important security bug fix would need to be updated in a |
15 |
> matter of hours while a functionality enhancement could wait for a |
16 |
> maintenance window a few weeks later. I love the idea of being able |
17 |
> to have constant up-to-date servers but I'd love to be able to update |
18 |
> the important things quickly and do all the other updates at a later time. |
19 |
> |
20 |
> One could say "how can we judge if an update is important/critical" ... |
21 |
> I'd answer that the one that is doing the ebuild surely has a good idea |
22 |
> of the importance of the change. |
23 |
> |
24 |
> It's late here too, I'm not sure I'm pretty clear ... let me give a |
25 |
> visual example : |
26 |
> |
27 |
> |
28 |
> emerge --critical --pretend --update world |
29 |
> [ebuild U ] sys-libs/pam-0.77 [0.75-r11] |
30 |
> [ebuild U ] sys-apps/pam-login-3.14 [3.11] |
31 |
> [ebuild U ] net-misc/openssh-3.7.1_p2-r2 [3.7.1_p2-r1] |
32 |
> |
33 |
> |
34 |
> emerge --pretend --update world |
35 |
> [ebuild U ] sys-devel/libperl-5.8.2 [5.8.0] |
36 |
> [ebuild U ] dev-lang/perl-5.8.2-r1 [5.8.0-r12] |
37 |
> [ebuild U ] sys-libs/ncurses-5.4 [5.3-r5] |
38 |
> [ebuild U ] sys-devel/bison-1.875 [1.35] |
39 |
> [ebuild U ] sys-devel/gcc-3.3.2-r5 [3.2.3-r3] |
40 |
> [ebuild U ] sys-libs/pam-0.77 [0.75-r11] |
41 |
> [ebuild U ] sys-apps/pam-login-3.14 [3.11] |
42 |
> [ebuild U ] app-shells/bash-2.05b-r9 [2.05b-r7] |
43 |
> [ebuild U ] net-misc/openssh-3.7.1_p2-r2 [3.7.1_p2-r1] |
44 |
> |
45 |
> |
46 |
> |
47 |
> |
48 |
> |
49 |
> |
50 |
> |
51 |
> On Tue, 2 Mar 2004, André Ventura Lemos wrote: |
52 |
> |
53 |
> > Yhea, I kind of remember that, things just get quietly fixed in the cvs. |
54 |
> > |
55 |
> > seems ok to me |
56 |
> > |
57 |
> > On Tue, 2004-03-02 at 01:33, Michael Cummings wrote: |
58 |
> > > Technically speaking (ok, so this is in policy somewhere at some point in |
59 |
> > > time, it's late, I'm tired) if it's a typo fix, the ebuild isn't bumped - |
60 |
> > > bumps (should) only occur when the change affects the code that is produced, |
61 |
> > > or otherwise alters/enhances functionality. |
62 |
> > > |
63 |
> > > -mike, just a perl guy |
64 |
> > |
65 |
> > |
66 |
> > -- |
67 |
> > I/O, I/O, |
68 |
> > It's off to disk I go, |
69 |
> > A bit or byte to read or write, |
70 |
> > I/O, I/O, I/O... |
71 |
> > |
72 |
> |
73 |
> |
74 |
> -- |
75 |
> gentoo-dev@g.o mailing list |
76 |
-- |
77 |
Chris Gianelloni |
78 |
Developer, Gentoo Linux |
79 |
Games Team |
80 |
|
81 |
Is your power animal a pengiun? |