1 |
On Tue, 2003-12-02 at 18:27, Daniel Robbins wrote: |
2 |
> Hi All, |
3 |
> |
4 |
> We have updated the Portage project page: |
5 |
> |
6 |
> http://www.gentoo.org/proj/en/portage/ |
7 |
> |
8 |
> It now includes an updated timeline, a Portage roadmap/TODO (there are a |
9 |
> few broken links in here, known issue) and a Portage-ng specification |
10 |
> document which is viewable here: |
11 |
|
12 |
In section 3 (Provide a coherent model for (meta)data representation and |
13 |
storage), how about a single API for accessing portage data that is both |
14 |
efficient and well-designed with the ability to change the back-end used |
15 |
to store the data. |
16 |
|
17 |
i.e. Allow the user/administrator to setup a database (MySQL, |
18 |
PostgreSQL, Oracle, dbm, etc.) either locally or remotely and use it to |
19 |
store portage data. Kinda like emergesql but with the ability to store |
20 |
information about what packages are installed on which systems. |
21 |
|
22 |
There are other things that could be done to with this model. How about |
23 |
using different classes of settings for the Portage system (make.conf). |
24 |
Have a class for workstations, servers, firewalls, x86, ppc, Pentium, |
25 |
Athlon, etc. and then place each Gentoo Linux system in one or more of |
26 |
these classes. |
27 |
|
28 |
|
29 |
-- |
30 |
gentoo-dev@g.o mailing list |