1 |
On Mon, 2005-08-29 at 15:32 -0500, Brian Harring wrote: |
2 |
> On Mon, Aug 29, 2005 at 12:56:35PM -0400, Chris Gianelloni wrote: |
3 |
> > On Sat, 2005-08-27 at 05:01 -0500, Brian Harring wrote: |
4 |
> > Basically, you've taken then 2005.1 profile and made it useless, since |
5 |
> > the stages weren't built against it anyway. |
6 |
|
7 |
> Via that logic (don't change it lest it negates a release), we would |
8 |
> never be able to do changes, or would be forced to do changes strictly |
9 |
> whenever y'all are doing a new release. |
10 |
|
11 |
Not exactly. |
12 |
|
13 |
I tend to agree with you that the base arch profiles should be a fairly |
14 |
minimal set of defaults, but also, default-linux has always been tied to |
15 |
releases, as much as possible. The point is that we really should *not* |
16 |
be making changes without media that matches it. Take the recent "eds |
17 |
gstreamer" thread as a good example. |
18 |
|
19 |
People expect a release to have changes. They don't expect, and don't |
20 |
*want* their system to change underneath them. |
21 |
|
22 |
If I had my way, there wouldn't ever be changes made to any of the |
23 |
profiles, including the arch profiles, unless absolutely necessary, and |
24 |
all changes would go into versioned (or named and versioned, or just |
25 |
named, etc) sub-profiles. The only thing that has kept us from doing |
26 |
this already is the lack of multiple inheritance. |
27 |
|
28 |
> Profiles aren't bound to the releases, despite how people may view it |
29 |
> and/or the current profile maitnainer's usage of 'em. |
30 |
|
31 |
No, they are not. I never said that they were, either. |
32 |
|
33 |
That being said, it ends up being the release team that ends up getting |
34 |
the bugs for the profiles. While there are a small number of developers |
35 |
that will change profiles under default-linux, it is more often than not |
36 |
left up to each arch's release coordinator. There are also non-release |
37 |
profiles on a few arches. |
38 |
|
39 |
There's nothing stopping you from creating a default-linux/x86/ferringb |
40 |
profile and doing whatever you wish in it, but editing |
41 |
default-linux/x86/2005.1 without speaking with releng would be |
42 |
considered taboo. |
43 |
|
44 |
> > My point is pretty simple, |
45 |
> > why should we spend a bunch of time maintaining something that is |
46 |
> > designed from the start to be customized, and most likely won't even be |
47 |
> > used anyway? |
48 |
> That's the issue; the profiles in their current form are customizable |
49 |
> only in the ability to negate a collection of flags. |
50 |
> Negating the whole beast is another story due to the desktop cruft |
51 |
> being shoved into the arch subprofiles. |
52 |
|
53 |
Sorry, but this didn't make a bit of sense to me. Perhaps you could |
54 |
reword it? |
55 |
|
56 |
> > I would much rather stick with the "2005.1" profile |
57 |
> > meaning "what we used to build 2005.1" than having it mean "some |
58 |
> > variation of 2005.1 is below here and using this profile is minimal and |
59 |
> > likely won't do what you expect". |
60 |
> Again, releases may be bound by available profiles, but available profiles |
61 |
> are not bound by available releases. |
62 |
|
63 |
Nobody ever said that profiles were bound to releases. I said that the |
64 |
versioned profile should match the release it was used to build and |
65 |
nothing more. Please don't insert meaning into what I say. |
66 |
|
67 |
> Aside from that, the comments about variations/minimal/not doing what |
68 |
> you expect, what do you think USE="-* user's actual desired flags" |
69 |
> accomplishes? |
70 |
|
71 |
It accomplishes exactly what I think it does. It turns off all |
72 |
automatically-enabled USE flags. I use it personally, because I run |
73 |
with a much smaller set of USE flags and I don't want things being |
74 |
changed without my knowledge. That being said, I also understand that |
75 |
this means I need to spend some time maintaining my systems and checking |
76 |
for the inclusion of new flags when I merge packages or do upgrades. |
77 |
|
78 |
> Profile customization occurs, /etc/portage/profiles exists for this |
79 |
> reason; the 2005.1 profile (fex) is probably *rarely* ran exactly as |
80 |
> y'all have it specified considering we do have user level use flags, |
81 |
> tweaking the hell out of '05.1. |
82 |
|
83 |
You would be surprised at the number of people that use GRP and rarely, |
84 |
if ever, change their USE flags. I wish I had numbers, but I don't. |
85 |
|
86 |
Anyway, the default set of USE flags seems to be a pretty perfect mix |
87 |
for most people. It gives packages that work as expected, and is geared |
88 |
toward a desktop system. Without any more specific examples of what |
89 |
you're trying to point out, I'm just not seeing it. |
90 |
|
91 |
> Aside from mild disagreement on views, as was stated in previous |
92 |
> emails, multiple inheritance I tend to think is required to minimize |
93 |
> the work for y'all; what I want you guys to do (or I'll do myself) is |
94 |
> chunk the suckers up so people after a minimal base for running |
95 |
> it themselves, or building up their own subprofile can do so. Not |
96 |
> after jamming maintenance nightmares on you, which without multiple |
97 |
> inheritance, might be a bit. |
98 |
|
99 |
I know that I won't be spending *my* time making any profile other than |
100 |
the defaults used for building the release. Anyone is welcome to build |
101 |
profiles for anything else that they might want, but since the release |
102 |
team doesn't use it, we shouldn't be forced to waste our time on it. |
103 |
|
104 |
-- |
105 |
Chris Gianelloni |
106 |
Release Engineering - Strategic Lead/QA Manager |
107 |
Games - Developer |
108 |
Gentoo Linux |