1 |
2006/3/20, Gustavo Sverzut Barbieri <barbieri@×××××.com>: |
2 |
> |
3 |
> I do think you're overcomplicating things where you shouldn't. |
4 |
> |
5 |
> Declaring stuff manually will always break, and to ensure a safe |
6 |
> system, it's better to use compiler information. |
7 |
|
8 |
In all cases, dependancy should be based on interfaces, not code. |
9 |
|
10 |
All packages may: |
11 |
* Provide an interface |
12 |
* Use an interface |
13 |
|
14 |
Depending on useflags, OS and other compile options, it differs, which |
15 |
interfaces are provided and used. |
16 |
|
17 |
This is, abstractly, what portage does with interfaces. |
18 |
|
19 |
If portage uses some interface, it may need it's header files when |
20 |
building. It may also need another lib for static build. This means |
21 |
that binary check is not possible in all cases. |
22 |
|
23 |
Now, the problem is: |
24 |
* How to get an information about a package, which specifies exactly, |
25 |
which interface is needed. How to get it before building in case when |
26 |
this interface is needed to be emerged before compilation [before |
27 |
linking everything together, at least]. Which is a form of this |
28 |
information and what could be read out from that? |
29 |
* How to get information about which interfaces are provided by which |
30 |
packages *not yet emerged* -- by their current use flags(?). This |
31 |
means that it must be possible to know, which interfaces are provided |
32 |
by packages, without first building it -- and the form given by binary |
33 |
check must be the same as the form of descriptor used by this package |
34 |
check. |
35 |
|
36 |
So, how to get correct provider together with correct client? |
37 |
|
38 |
> So I wouldn't mind fixing it to one package instead of a slot. |
39 |
> |
40 |
> I mean, if user compiled software X-1.0 and it depends on library |
41 |
> Y-2.0, provided at the moment of X-1.0 compilation by package Z-3.0, |
42 |
> then make X-1.0 depend on Z-3.0. |
43 |
> |
44 |
> If you were to remove Z-3.0 or upgrade it or even add other option by |
45 |
> means of USE, have X-1.0 to be recompiled too... you could be even |
46 |
> more correct to make it check CFLAGS too. |
47 |
> |
48 |
> Of course this correctness could piss users, then you could have a |
49 |
> --nodeps or something like that to avoid this and use the old |
50 |
> behaviour. |
51 |
> |
52 |
> -- |
53 |
> Gustavo Sverzut Barbieri |
54 |
> -------------------------------------- |
55 |
> Jabber: barbieri@×××××.com |
56 |
> MSN: barbieri@×××××.com |
57 |
> ICQ#: 17249123 |
58 |
> Skype: gsbarbieri |
59 |
> Mobile: +55 (81) 9927 0010 |
60 |
> Phone: +1 (347) 624 6296; 08122692@××××××××××××××.com |
61 |
> GPG: 0xB640E1A2 @ wwwkeys.pgp.net |
62 |
> |
63 |
> -- |
64 |
> gentoo-portage-dev@g.o mailing list |
65 |
> |
66 |
> |
67 |
|
68 |
|
69 |
|
70 |
-- |
71 |
tvali |
72 |
(e-mail: "qtvali@×××××.com"; msn: "qtvali@×××××.com"; |
73 |
icq: "317-492-912") |
74 |
|
75 |
I like net more than life, cause if i do something wrong, then people |
76 |
in net will tell me that i do, so that i can fix it -- people in life |
77 |
will tell others that i do. |
78 |
|
79 |
-- |
80 |
gentoo-portage-dev@g.o mailing list |