| 1 |
Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> wrote: |
| 2 |
> On Tue, 24 Feb 2009 10:46:30 -0500 |
| 3 |
> Richard Freeman <rich0@g.o> wrote: |
| 4 |
> > I will certainly concede that putting it inside the ebuild |
| 5 |
> > potentially breaks compatibility with existing package managers. |
| 6 |
> > That is certainly a downside to this approach. However, none of the |
| 7 |
> > other objections that have been raised appear to hold water. |
| 8 |
> |
| 9 |
> ...and it means we can't change name or version rules. |
| 10 |
> |
| 11 |
> ...and it means over doubling the best possible time to work out a |
| 12 |
> dependency tree in the common case where the metadata cache is valid. |
| 13 |
> |
| 14 |
> ...and it means we can't make arbitrary format changes. |
| 15 |
|
| 16 |
Those would all land in the category of "backwards compatibility" |
| 17 |
mentioned below, as they would all break current sourcing rules. |
| 18 |
|
| 19 |
> > And if backwards compatibility were a serious issue you could define |
| 20 |
> > a new ".ebuild2" file spec that incorporates the EAPI inside the |
| 21 |
> > file and current package managers would ignore it. Then you're not |
| 22 |
> > changing the file extension every time a new EAPI comes along, and |
| 23 |
> > the need to do so could be handled via future GLEPs. |
| 24 |
> |
| 25 |
> Developers already have to stop and think and consult the conveniently |
| 26 |
> available table of features for EAPIs. By splitting the EAPI concept |
| 27 |
> in two you're doubling the amount of data to be learnt. |
| 28 |
|
| 29 |
I would think that this is a very small cost, and the benefit would be |
| 30 |
(I hope) that more people would agree on the solution and then we can go |
| 31 |
forward. Is that not a valid consideration? |
| 32 |
|
| 33 |
-- |
| 34 |
Jim Ramsay |
| 35 |
Gentoo Developer (rox/fluxbox/gkrellm/vim) |
Archives