| 1 |
>>>>> On Mon, 14 Dec 2015, Michał Górny wrote: |
| 2 |
|
| 3 |
> As far as I can see, this GLEP predates EAPI and does not meet |
| 4 |
> modern standards. It needs to be updated or killed with fire. |
| 5 |
|
| 6 |
> For a start, relation to EAPI needs to be defined. This will likely |
| 7 |
> require both profiles and ebuilds to use the new EAPI. |
| 8 |
|
| 9 |
Certainly this is true for ebuilds. We could introduce EUSERS and |
| 10 |
EGROUPS variables as well as FEATURES=noautoaccts in EAPI 7. |
| 11 |
|
| 12 |
I'd rather avoid bumping all profiles to EAPI 7, though. From a |
| 13 |
pragmatic point of view, I think nothing bad would happen if we would |
| 14 |
add user and group definition files to an existing (EAPI 5) profile, |
| 15 |
as old package managers would simple ignore these files. |
| 16 |
|
| 17 |
> Also, the contents of 'backwards compatibility' section are |
| 18 |
> unacceptable. But that's probably going to be covered by EAPI. |
| 19 |
|
| 20 |
> The spec itself is hard to follow, though the idea seems simple. |
| 21 |
> It makes me wonder if we aren't missing something important there. |
| 22 |
|
| 23 |
The spec seems incomplete. I cannot find a description of the user and |
| 24 |
group files' format. (But in fact, there is a standard format which |
| 25 |
suggests itself, namely that of the passwd(5) and group(5) files.) |
| 26 |
|
| 27 |
Also having whole directory trees seems wasteful and doesn't fit so |
| 28 |
well into the existing design of profiles. It might be simpler to put |
| 29 |
"user" (or "passwd") and "group" files directly in the profile. |
| 30 |
(If directories are really needed, we could use the scheme foreseen |
| 31 |
in [1] for package.* and use.* files.) |
| 32 |
|
| 33 |
Also a mechanism how a subprofile could undefine a user or group |
| 34 |
defined in its parent seems to be missing. |
| 35 |
|
| 36 |
Ulrich |
| 37 |
|
| 38 |
|
| 39 |
[1] https://bugs.gentoo.org/282296 |
Archives