1 |
On Tue, Feb 28, 2012 at 07:45:51AM +0000, Ciaran McCreesh wrote: |
2 |
> On Tue, 28 Feb 2012 08:40:05 +0100 |
3 |
> Ulrich Mueller <ulm@g.o> wrote: |
4 |
> > >>>>> On Mon, 27 Feb 2012, David Leverton wrote: |
5 |
> > > I think the logic behind that is that an EAPI defined outside PMS |
6 |
> > > wouldn't be bound by PMS's rules anyway, and the EAPIs defined |
7 |
> > > inside PMS are explicitly listed so there's not much point giving a |
8 |
> > > general syntactic rule as well [...] |
9 |
> > |
10 |
> > If we follow this logic, then the following sentence should be removed |
11 |
> > from PMS: |
12 |
> > |
13 |
> > "EAPIs whose value starts with the string paludis- are reserved for |
14 |
> > experimental use by the Paludis package manager." |
15 |
> |
16 |
> Don't tell anyone, but that's mostly just in there because some people |
17 |
> insisted that EAPIs were numbers (and thus comparable), so I wanted an |
18 |
> explicit mention of one that wasn't. |
19 |
|
20 |
A hard rule there is a bit dumb, although frankly I don't hugely |
21 |
care one way or another- if we have it, extend it to the three |
22 |
rather than just paludis. Or drop it, don't hugely care. |
23 |
|
24 |
EAPI *does* need to be tightened to make clear the first char of EAPI |
25 |
can't be '-'; it's used now via the cache as a marker to indicate that |
26 |
it pulled the EAPI, but didn't know how to handle it. It's live and |
27 |
in use, and has been an undocumented requirement basically since day |
28 |
1. |
29 |
|
30 |
No one is going to conflict with it, just best to explicitly note it's |
31 |
offlimits as the first char of an EAPI name. |
32 |
~harring |