| 1 |
On Wed, 11 Jun 2008 14:49:19 +0200 |
| 2 |
Luca Barbato <lu_zero@g.o> wrote: |
| 3 |
> > Why is "Create tests for EAPI=1 stuff." not a way to describe how |
| 4 |
> > to reproduce a problem? |
| 5 |
> |
| 6 |
> because EAPI1 isn't specified completely so you don't have a large |
| 7 |
> field to cover but you also do not know the bounds of it. |
| 8 |
|
| 9 |
EAPI 1 is entirely specified in terms of a diff against EAPI 0. |
| 10 |
Checking every part that's changed before releasing an EAPI 1 package |
| 11 |
manager is the least any responsible person would do. That they would |
| 12 |
release a version without doing such basic tests shows you just how |
| 13 |
much they care about Gentoo... |
| 14 |
|
| 15 |
> Assuming that Ciaranm isn't just lying knowingly it's just plainly |
| 16 |
> rude, otherwise it is pure malice. |
| 17 |
|
| 18 |
What, asking the pkgcore people to test their code before releasing a |
| 19 |
version that claims to support EAPI 1 but actually doesn't, forcing |
| 20 |
people to avoid using some of EAPI 1 to avoid breaking pkgcore, is |
| 21 |
malice? |
| 22 |
|
| 23 |
The whole "EAPI lets us do upgrades cleanly" process is broken when |
| 24 |
people release a package manager that claims to support a certain EAPI |
| 25 |
but doesn't. If pkgcore had any actual users we'd have to consider |
| 26 |
banning EAPI 1 in the tree and releasing EAPI 2 as being identical to |
| 27 |
EAPI 1 just to work around this. |
| 28 |
|
| 29 |
-- |
| 30 |
Ciaran McCreesh |
Archives