1 |
On Tuesday 16 December 2003 22:36, Pieter Van den Abeele wrote: |
2 |
|
3 |
> I would use 'knowledge base' instead of database, because it is a more |
4 |
> generic term. A ports directory is not really a database as we know it |
5 |
> (think SQL etc.). One of the benefits of formulating the core in |
6 |
> prolog, would be to use a 'datalog' syntax, datalog is regarded as a |
7 |
> universal query language. SQL can be expressed in datalog, just like it |
8 |
> can be expressed in relational algebra/calculus. |
9 |
|
10 |
Well, knowledge base makes me think of logic, so settings repository might be |
11 |
even a better word. It could however also be that I had too many encounters |
12 |
with logicians. |
13 |
> each concrete architecture/platform can implement its own |
14 |
> 'instal/uninstall' strategy. |
15 |
> |
16 |
> > Ebuilds wouldn't anymore be a shell scripts, but the build module will |
17 |
> > offer all necessary commands to build something. Ebuild will simply be |
18 |
> > a command sequence. Same goes for the install module. |
19 |
> |
20 |
> This makes eclasses overriding/adding some default behaviour or |
21 |
> 'meta-data' somewhat more difficult to implement. |
22 |
> I'd split ebuilds into two parts: the part detailing 'howto' and the |
23 |
> part detailing 'about'. |
24 |
|
25 |
My idea is basically that there is a structure with embedded shellscripts. |
26 |
This way we don't need bash for the metadata, only for actual unpacking, |
27 |
configuring, compiling and installing |
28 |
|
29 |
> |
30 |
> I certainly wouldn't exclude using a database (like in sql databases, |
31 |
> etc.), but wouldn't exclude other possibilities too. If anyone wants to |
32 |
> create one huge file containing all data in his own syntax, that should |
33 |
> be possible because of the modular design. |
34 |
|
35 |
Me neither, it should be an option however. |
36 |
|
37 |
Paul |
38 |
|
39 |
-- |
40 |
Paul de Vrieze |
41 |
Gentoo Developer |
42 |
Mail: pauldv@g.o |
43 |
Homepage: http://www.devrieze.net |