1 |
On Tue, 2004-09-21 at 09:13, Chris Gianelloni wrote: |
2 |
> On Mon, 2004-09-20 at 12:41, Benjamin Judas wrote: |
3 |
> > There is a slight problem regarding the USE-Settings in the profiles: |
4 |
> > |
5 |
> > GRP_STAGE23_USE="ipv6 pam tcpd readline nls ssl gpm perl python berkdb acl ncurses" |
6 |
> > |
7 |
> > As you can see, this defines acl. Now: |
8 |
> > |
9 |
> > USE="x86 oss apm arts avi berkdb bitmap-fonts crypt cups encode foomaticdb gdbm |
10 |
> > gif gpm gtk gtk2 imlib jpeg kde gnome libg++ libwww mad mikmod motif mpeg ncurses |
11 |
> > nls oggvorbis opengl pam pdflib png python qt quicktime readline sdl slang spell |
12 |
> > ssl svga tcpd truetype X xml2 xmms xprint xv zlib" |
13 |
> > |
14 |
> > This doesn't define acl. |
15 |
> > |
16 |
> > This could break things if users don't set USE="acl" on bootstrap/emerge system. |
17 |
> > Possible solutions could be to either remove acl from GRP_STAGE23_USE or add it to |
18 |
> > USE. The first solution would cause everybody using acl to not be able to install without |
19 |
> > any problems. Putting it into USE would cause an overhead for everybody. The overhead |
20 |
> > doesn't seem to be that big, but...well...it's an overhead. |
21 |
> > |
22 |
> > What do you think about it and how do you think this should be solved? |
23 |
> |
24 |
> We should not remove the ability to install using acl. I also think |
25 |
> that it is required for somethings (hardened?) and should stay. I think |
26 |
> my main problem is I don't understand how the situation is broken. |
27 |
> Could you perhaps paint a better picture for me (and all the other |
28 |
> devs)? |
29 |
|
30 |
|
31 |
drop it++ |
32 |
|
33 |
Some USE flags almost need maintainers (and this is one of them). There |
34 |
are special patches that have to be applied to packages that don't seem |
35 |
to be going upstream. Sometimes these are patches that conflict with the |
36 |
natural course of upgrading/version bumps of packages and the complete |
37 |
logic has to be reworked. When these conflicts happen where the acl flag |
38 |
conflicts said patches are simply dropped. So I'm thinking that a full |
39 |
bootstrap is currently only half arse with USE=acl and not a complete |
40 |
solution. |
41 |
|
42 |
If somebody/existing dev is willing to take on the responsibility for |
43 |
said flag then hey great leave it in the stages but as is it's |
44 |
incomplete as I've pointed out. |
45 |
|
46 |
And as josh has pointed out hardened does not need it in anyway shape or |
47 |
form. |
48 |
|
49 |
Furthermore it's quite a confusing USE flag name to begin with. |
50 |
What's it really needed for? |
51 |
|
52 |
Gentoo supports atleast 3 types of access control systems that I can |
53 |
think of so not sure why this one has to be so descriptive. |
54 |
|
55 |
-- |
56 |
Ned Ludd <solar@g.o> |
57 |
Gentoo (hardened,security,infrastructure,embedded,toolchain) Developer |