| 1 |
On Sun, 17 May 2009 21:57:40 +0200 |
| 2 |
Thomas de Grenier de Latour <tom.gl@××××.fr> wrote: |
| 3 |
> On 2009/05/17, Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> wrote: |
| 4 |
> > You don't define it quite like that. You define it by mapping EAPI X |
| 5 |
> > _p onto super-EAPI _p0, and EAPI Y _p onto super_EAPI _pINFINITY. |
| 6 |
> > That way the ordering's well defined. |
| 7 |
> |
| 8 |
> I understand the idea, but I still don't think it's viable to allow |
| 9 |
> such definitions, because there are too many contexts in which we |
| 10 |
> manipulate pure version strings, without decorating them with an EAPI, |
| 11 |
> and without reference to a concrete package from which we could get |
| 12 |
> it. Take ">=bar/foo-1_p" as an "emerge" command line argument for |
| 13 |
> instance, is my foo-1_p1.ebuild a candidate? |
| 14 |
|
| 15 |
Conceptually, you'd define a 'user' EAPI for those things, so you can |
| 16 |
define it any way you want (including in such a way that the _p thing |
| 17 |
works both ways depending upon the EAPI used for creating the thing |
| 18 |
you're comparing it to -- for the user EAPI, you'd define it as being |
| 19 |
_pUNSPECIFIED rather than _p0 or _pINFINITY and use the other side of |
| 20 |
the comparison to decide the result). But yes, if you do something |
| 21 |
silly like your example, things get very complicated. |
| 22 |
|
| 23 |
> > > As a consequence, the algorithm for picking best version of a |
| 24 |
> > > package can be as simple as the following: |
| 25 |
> > > 1- among all ebuilds filenames, filter out the ones with |
| 26 |
> > > unrecognized version string |
| 27 |
> > |
| 28 |
> > You don't know whether you recognise the version string until you |
| 29 |
> > know the EAPI, though. |
| 30 |
> |
| 31 |
> Under my previously stated restrictions, you know: |
| 32 |
> - which one can be rejected for sure (the ones not recognized by any |
| 33 |
> of your implemented EAPI). |
| 34 |
> - how to correctly order the remaining ones (even the incorrect ones |
| 35 |
> which may remain and would be rejected only at step 4), and thus |
| 36 |
> where to start to find the "best" correct one. |
| 37 |
|
| 38 |
Your previously stated restrictions are too strong, though. And when it |
| 39 |
turns out a future change breaks those restrictions, we'd be back to |
| 40 |
yet another extension change. |
| 41 |
|
| 42 |
-- |
| 43 |
Ciaran McCreesh |
Archives