1 |
On Tue, 24 Feb 2009 21:59:39 +0530 |
2 |
Nirbheek Chauhan <nirbheek@g.o> wrote: |
3 |
> On Tue, Feb 24, 2009 at 9:26 PM, Ciaran McCreesh |
4 |
> <ciaran.mccreesh@××××××××××.com> wrote: |
5 |
> > ...and it means we can't change name or version rules. |
6 |
> |
7 |
> And has such a case come to light recently where it was *essential* to |
8 |
> do so? Why solve problems that don't exist? |
9 |
|
10 |
Because they do exist, which is why name and version rules have been |
11 |
changed the hard way at least twice previously. The version format is |
12 |
still considerably less flexible than what upstreams use, and a lot of |
13 |
the current limitations on its format are purely historical. |
14 |
|
15 |
> > ...and it means we can't make arbitrary format changes. |
16 |
> |
17 |
> What? Why are we over-engineering this? Does anyone seriously want to |
18 |
> convert ebuilds to XML? I honestly think anything beyond incremental |
19 |
> changes is not relevant for Gentoo |
20 |
|
21 |
You appear to be confusing arbitrary format changes with doing a Zynot. |
22 |
The two are not the same. |
23 |
|
24 |
> > Developers already have to stop and think and consult the |
25 |
> > conveniently available table of features for EAPIs. By splitting |
26 |
> > the EAPI concept in two you're doubling the amount of data to be |
27 |
> > learnt. |
28 |
> |
29 |
> That's a documentation problem. |
30 |
|
31 |
No, it's a design problem. Good design looks for ways to minimise the |
32 |
amount of unnecessary arbitrary information the user has to remember. |
33 |
|
34 |
-- |
35 |
Ciaran McCreesh |