1 |
On 2016-01-25 22:34, Jonathan Callen wrote: |
2 |
> On 01/25/2016 07:29 AM, Aaron W. Swenson wrote: |
3 |
> > On 2016-01-24 18:44, Michael Orlitzky wrote: |
4 |
> >> On 01/24/2016 06:29 PM, Aaron W. Swenson wrote: |
5 |
> >>> Okay, provided that the new USE_EXPAND is okay for |
6 |
> >>> POSTGRES_TARGETS, attached are the eclasses that I'll commit to |
7 |
> >>> the tree. |
8 |
> >>> |
9 |
> >> |
10 |
> >>> case ${EAPI:-0} in 0|1|2|3|4) die "postgres-multi.eclass |
11 |
> >>> requires EAPI 5 or higher" ;; *) ;; esac |
12 |
> >> |
13 |
> >> Does this really work with EAPI=6? I didn't try, but it looks |
14 |
> >> like it would need an eapply_user somewhere in src_prepare. And, |
15 |
> >> pedantry warning, there's no such thing as "EAPI 5 or higher." |
16 |
> >> The lawyers will tell you to do something like this instead |
17 |
> >> (stolen from git-r3): |
18 |
> >> |
19 |
> >> case "${EAPI:-0}" in 5|6) ;; *) die "Unsupported EAPI=${EAPI} |
20 |
> >> (unknown) for ${ECLASS}" ;; esac |
21 |
> >> |
22 |
> >> That will require an edit for every new EAPI, but prevents weird |
23 |
> >> crashes from things like a missing eapply_user. |
24 |
> >> |
25 |
> >> |
26 |
> > |
27 |
> > Thank you. |
28 |
> > |
29 |
> > I've added the eapply_user to postgres-multi and modified the case |
30 |
> > condition to match the hot goods you're selling on the sly. |
31 |
> > |
32 |
> |
33 |
> You missed the fix for EAPI in postgres-multi.eclass (it looks okay in |
34 |
> the other eclass, though). |
35 |
|
36 |
Fixed. Thanks! |