1 |
On Tuesday 20 July 2004 21:43, Dylan Carlson wrote: |
2 |
> My thoughts on this are: |
3 |
> |
4 |
> 1/ Many Gentoo users want the ability to update packages for fixes only |
5 |
> (and not just for servers) |
6 |
|
7 |
That's simple - just don't do 'emerge sync; emerge -u world' on a daily basis. |
8 |
Only upgrade when you definitely need a fix. |
9 |
</troll> |
10 |
|
11 |
There's a serious point to that. There should be some discussion to make sure |
12 |
we actually understand the problem we're trying to solve first ;-) |
13 |
|
14 |
> 2/ Maintaining separate trees in CVS is probably asking too much of our |
15 |
> current roster, plus requires adding infrastructure and getting everyone |
16 |
> to mirror the new tree(s) |
17 |
|
18 |
Agreed. |
19 |
|
20 |
> 3/ Users already have a good understanding of what profiles are. This is |
21 |
> useful. |
22 |
|
23 |
I believe that's an assumption not based in fact. |
24 |
|
25 |
> 4/ Creating separate CVS trees while also using profiles seems superfluous. |
26 |
|
27 |
Not sure about that. I believe a profile is something that should never |
28 |
change once it's marked stable. Right now, I don't see how that fits in with |
29 |
our ever-changing tree. |
30 |
|
31 |
> Therefore I believe another possible solution is to change the way we use |
32 |
> profiles (both in practice and in QA policy): |
33 |
> |
34 |
> Implications: |
35 |
> |
36 |
> 1/ archs will need to have a regular, predictable release schedule, |
37 |
> probably at least every six months. |
38 |
|
39 |
Why? Not saying I don't agree, but it'd help if this assertion was justified. |
40 |
|
41 |
> 2/ profiles will list specific package versions whenever possible. e.g.: |
42 |
> =sys-libs/glibc-2.3.2 |
43 |
|
44 |
|
45 |
|
46 |
> 3/ we will need to prepare a list of packages which should not change |
47 |
> unless the user switches profiles. those packages (e.g., toolchain) in |
48 |
> the current, and older profiles do not get updated except for fixes. |
49 |
|
50 |
I don't think this is sustainable. Practically every new version of a package |
51 |
contains fixes compared to the one that it replaces. |
52 |
|
53 |
> 4/ we will need to have a policy about how long we'll support a profile, |
54 |
> and a procedure for end-of-lifing profiles. (probably don't want to |
55 |
> support a single profile for more than 2 years) |
56 |
|
57 |
Not *less* than 2 years, I'd have thought. Some commercial systems are |
58 |
supported for 5 or even 10 years. |
59 |
|
60 |
> 5/ we will need to have documentation for users who are upgrading from one |
61 |
> profile to another (upgrade warnings, instructions and considerations) |
62 |
> 6/ repoman will need to be enhanced to check the profiles before removing |
63 |
> any ebuilds from the tree that might be needed. specific versions pinned |
64 |
> in profiles need to stay until the profile is changed. |
65 |
> |
66 |
> Potential problems: |
67 |
> |
68 |
> 1/ Makes KEYWORDS redundant |
69 |
|
70 |
So every single one of our 8,000+ packages that's stable will have to be |
71 |
listed in one or more profiles? That sounds impractical. |
72 |
|
73 |
Best regards, |
74 |
Stu |
75 |
-- |
76 |
Stuart Herbert stuart@g.o |
77 |
Gentoo Developer http://www.gentoo.org/ |
78 |
http://stu.gnqs.org/diary/ |
79 |
|
80 |
GnuPG key id# F9AFC57C available from http://pgp.mit.edu |
81 |
Key fingerprint = 31FB 50D4 1F88 E227 F319 C549 0C2F 80BA F9AF C57C |
82 |
-- |