1 |
Stuart Bouyer wrote: |
2 |
> I think you misunderstand the complaint here. The problem (which has |
3 |
> been brought up this list previously) is that there is no way to |
4 |
> guarantee that I can get my server back to it's current configuration if |
5 |
> I have to reinstall at a later date. Not only will new versions of |
6 |
> ebuilds have been added to the portage tree, but there is a great chance |
7 |
> that ebuild for the version of the package that I'm happy using will no |
8 |
> longer be in portage tree. What I install using the 1.3 install CD today |
9 |
> will be very different from what I installed 3 months ago. |
10 |
|
11 |
A long-standing problem is certainly the trigger-happy nature of so many |
12 |
developers when it comes to removing "old" ebuilds. I've seen it to the |
13 |
extreme where a new build was committed and all previous builds removed |
14 |
before the new version (which already had Bugzilla reports) had even |
15 |
been tested, letalone assured to be working. |
16 |
|
17 |
I've been pushing, and plan to continue to push as hard as possible for |
18 |
a policy change where old, stable, ebuilds are concerned. |
19 |
|
20 |
In a typical situation, we'll have builds versioned like so; |
21 |
|
22 |
1.0 |
23 |
1.0-r1 |
24 |
1.0-r2 |
25 |
1.0-r3 |
26 |
1.1 |
27 |
1.1-r1 |
28 |
1.1-r2 |
29 |
1.2 |
30 |
1.3 |
31 |
2.0 |
32 |
2.0-r1 |
33 |
2.0-r2 |
34 |
2.0-r3 |
35 |
|
36 |
etc. ad nauseum. Current practises seem to revolve around removing all |
37 |
but the last, say, two or three ebuilds. Unfortunately, that leaves a |
38 |
/LOT/ of users in the lurch. |
39 |
|
40 |
My contention is this; -r* builds are intended to fix problems with |
41 |
previous -r* builds (and the initial ebuild for a given version), so |
42 |
rather than removing everything <2.0, why not remove all the lowest |
43 |
numbered -r builds, thus leaving us with; |
44 |
|
45 |
1.0-r3 |
46 |
1.1-r2 |
47 |
1.2 |
48 |
1.3 |
49 |
2.0-r3 |
50 |
|
51 |
If an ebuild is, for whatever reason, broken enough to warrant a |
52 |
revision bump, does it really conpute to leave it in the tree while |
53 |
removing a perfectly viable, if only older build in its stead? |
54 |
|
55 |
Such a system would allow us to maintain upwards of a years' worth of |
56 |
builds in the tree without severe bloating and would permit such a |
57 |
"static version", as you've said, and would make server administrators |
58 |
rest a little easier. |
59 |
|
60 |
A lot of talk recently about migrating Gentoo from Apache 1.3 to 2 has |
61 |
me a little frighteend. Despite assurances that 1.3 will remain in the |
62 |
tree, my observations over this past year and a half don't comfort me, |
63 |
and I, for one, am not ready (or able) to migrate all my hosting servers |
64 |
to Apache 2 just yet. |
65 |
|
66 |
> If I don't want to update to the new package - and there are many |
67 |
> reasons why I would not want to - then my only optinos are not to emerge |
68 |
> sync (and miss out on the update I do need) or to manually find the |
69 |
> ebuilds I want in the attic of the web cvs gateway. |
70 |
|
71 |
Agreed completely. All too often a version change involves many late |
72 |
nights and heartburn weeding through new features and depracated |
73 |
functionality. |
74 |
|
75 |
Projects, more often than not, will maintain security and critical |
76 |
updates for their old(er) versions for this very reason. ISC and the |
77 |
Apache Foundation are encouraging their userbase to migrate to the |
78 |
latest and greatest, for many reasons, but are in no way leaving their |
79 |
legacy customers in the lurch. Security updates to Apache 1.3 and BIND 8 |
80 |
will continue indefinately (or until the use of such products is |
81 |
completely not viable). For that matter, BIND 4 still sees security |
82 |
updates when required. |
83 |
|
84 |
If we're to move off the desktop and onto servers (or, Tux forbid, the |
85 |
Enterprise Platform) we need to allow users to stick with what works and |
86 |
track critical/security updates. |
87 |
|
88 |
In the meantime, our best bet is to emerge -bk and store the resultant |
89 |
binaries in a centrally available location. The problems, however, with |
90 |
ebuilds dissapearing from the tree remain, so even that is imperfect. |
91 |
|
92 |
-- |
93 |
Stewart Honsberger |
94 |
http://blackdeath.snerk.org/ |
95 |
"Capitalists, by nature, organize to protect themselves. |
96 |
-- Geeks, by nature, resist organizaion." |
97 |
|
98 |
|
99 |
-- |
100 |
gentoo-dev@g.o mailing list |