[aur-general] Fwd: gnupg-largekeys in AUR

[Ido Rosen]

On Thu, Nov 28, 2013 at 2:04 PM, Jerome Leclanche <adys.wh at gmail.com> wrote:

> On Thu, Nov 28, 2013 at 5:48 PM, Ido Rosen <ido at kernel.org> wrote:
> > On Thu, Nov 28, 2013 at 10:49 AM, Jerome Leclanche <adys.wh at gmail.com
> >wrote:
> >
> >> What's the outcome on this? I'm interested in large keys in default
> gnupg.
> >>
> >> That said, is there a reason why the patch isnt upstream yet?
> >> J. Leclanche
> >>
> >>
> > It was rejected upstream previously a few times.
> >
> > If we want it, it has to be a patch on upstream in our gpg version.  I
> > believe the reasoning that allowing larger key sizes are a performance
> > issue for mobile does not really apply here.
>
> That sounds like the kind of perfect use case for a compile-time option.
>
> J. Leclanche
>
>
If you mean an upstream compile-time option (i.e. not in the PKGBUILD), I
agree wholeheartedly.  Now all we have to do is convince Werner Koch, which
means emailing gnupg-users/gnupg-devel, or submitting a bug report with a
patch (and a copyright assignment).

Closed bugs related to this are:
https://bugs.g10code.com/gnupg/issue1441
https://bugs.g10code.com/gnupg/issue1460

FWIW, the pattern in upstream seems to be that anyone who suggests a larger
max key size gets told no for a few years, and then it happens anyway.
 That is, if you can call 2 data points (2048, then 4096) a pattern...