1 |
Dnia 2014-06-15, o godz. 14:45:07 |
2 |
Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> napisał(a): |
3 |
|
4 |
> On Sun, 15 Jun 2014 09:53:02 +0200 |
5 |
> Michał Górny <mgorny@g.o> wrote: |
6 |
> > It is the spec's fault when it sets stupid and mismatched |
7 |
> > requirements. As long as we don't restrict pkg_* phases from calling |
8 |
> > external tools, prohibiting writes to temporary directory is just |
9 |
> > pointless. Every semi-sane system has a writable temporary directory |
10 |
> > that can be used freely by random programs, and so shall ebuild |
11 |
> > environment have. |
12 |
> |
13 |
> But ${T} has to point somewhere special... |
14 |
|
15 |
But ${TMPDIR} doesn't. |
16 |
|
17 |
> > As for pkg_pretend(), I don't have a strong opinion whether it should |
18 |
> > be added to use external tools. However, prohibiting that would limit |
19 |
> > its use a lot. As for remaining pkg_*() phases, I think such |
20 |
> > prohibition would simply make the phases purposeless. |
21 |
> |
22 |
> But it can't rely upon external tools being present, and if they are |
23 |
> present, they can't be relied upon to work (have their dependencies |
24 |
> installed). |
25 |
|
26 |
It can rely on base system tools being present since that's what PMS |
27 |
requires. It relies on bash being present in particular, and bash can |
28 |
use temporarily files too. |
29 |
|
30 |
-- |
31 |
Best regards, |
32 |
Michał Górny |