Bug#660162: RFS: tack/1.07-2
On Sun, Feb 26, 2012 at 10:17 PM, Samuel Bronson <email@example.com> wrote:
> On Sun, Feb 26, 2012 at 7:21 PM, Jakub Wilk <firstname.lastname@example.org> wrote:
>> * Samuel Bronson <email@example.com>, 2012-02-26, 17:49:
>>> * Fix "hyphen-used-as-minus-sign" warning from lintian.
>> This patch...
>>> * New patch 03-allow-echoing-compilation-commands.patch, which enables
>>> printing of compilation commands by default.
>> ...and this patch have Forwarded headers pointing to gmane. Could you use
>> links to the "official" web archive (i.e.,
>> http://lists.gnu.org/archive/html/bug-ncurses/) instead? Not every one likes
>> gmane (e.g. I cannot stand its UI); lists.gnu.org would be more neutral.
>> (I won't be insistent about this, feel free to ignore me.)
> I was actually thinking maybe I should link to the official GNU
> archives, but then I noticed what they did to anything resembling an
> email address and decided not to do that -- gmane's mangling is at
> least *almost* reversible.
>> Also, if you wanted the patch headers to follow DEP-3, then you probably
>> need to remove stuff like "diff -Naurp tack.orig/tack.1 tack/tack.1" or
>> "Index: tack-1.06/tack.1". Otherwise such lines could be misinterpreted
>> by a parser as a part of patch header.
>>  I said "could" because the specification is far being clear (or maybe I
>> should say s/clear/unambiguous/).
> Well, *quilt* certainly doesn't misinterpret these lines thus. I don't
> really feel like removing things like this merely because of what
> DEP-3 neglects to specify; it really seems more like an issue with
> DEP-3 to me.
> (Also, patch #03 was originally produced by dpkg-source; I did remove
> some templates, but kept the overall format the same. I suppose
> *maybe* I should have left that line of three dashes?)
> If and when DEP-3 were to be clarified on this point, I'd be happy to
> change the patches to comply with it.
> -- Samuel Bronson
But note that Thomas Dickey has prepared a snapshot which is supposed
to do patch #03's job a different way, and fix that dpkg-shlibdeps
(I'll take a look at this in the morning.)