1 |
Steve Long wrote: |
2 |
> Ciaran McCreesh wrote: |
3 |
>> Steve Long wrote: |
4 |
>> | How serious an issue is it in terms of deps on sys pkgs? |
5 |
>> |
6 |
>> Very. It means we can't realistically handle packages that, by using |
7 |
>> autotools, depend upon the fifty odd system packages that are used by |
8 |
>> autotools-generated code. |
9 |
>> |
10 |
>> | > | > DEPEND="virtual/c-toolchain" would indeed be nice, but it's a |
11 |
>> | > | > rather large change... |
12 |
>> | > | > |
13 |
>> | > | How so? Isn't it simply a new meta? |
14 |
>> | > |
15 |
>> | > And an entire tree to update before it becomes meaningful. |
16 |
>> | > |
17 |
>> | Sure, but the changes can propagate thru as people update their |
18 |
>> | ebuilds, no? |
19 |
>> |
20 |
>> The tricky part then is figuring out whether something doesn't dep upon |
21 |
>> c-compiler because it doesn't need one or because the ebuilds haven't |
22 |
>> been updated. |
23 |
>> |
24 |
> I'm out of my depth here- I can't see where that would be a problem? |
25 |
> |
26 |
|
27 |
Er, his point being that if you don't do the upgrade all at once, you |
28 |
have two classes of package. |
29 |
|
30 |
1. Packages that don't require C-compiler |
31 |
2. Packages that don't yet depend upon C-compiler |
32 |
|
33 |
When doing the upgrade over a period of time the two classes of package |
34 |
become indistinguishable. Does Foo not need a C compiler, or has it |
35 |
just not gotten updated yet, it's impossible to tell without looking, so |
36 |
it's very difficult for people to report 'problem packages' because they |
37 |
have to unpack and examine the package every time (at worst). |
38 |
|
39 |
-- |
40 |
gentoo-dev@g.o mailing list |