Bug#793285: closed by Martin Quinson <martin.quinson at ens-rennes.fr> (Widelands builds without any problem with gcc5)

[Matthias Klose]

On 13.03.2017 21:01, Martin Quinson wrote:
> On Mon, Mar 13, 2017 at 07:50:42PM +0100, Matthias Klose wrote:
>>> as the title says, this bug does not affect widelands, thus closing.
>>
>> yes, it wouldn't have taken 16 months to confirm that. And no, none of these
>> patches results in a build error.  So your reasoning is plain wrong. But who
>> cares ...
> 
> I'm not sure I understand what you are talking about. I don't see no
> reasonning on my side nor patch in this whole bug report, but I may be
> blind. Could you please enlight me?
> 
> As for the long delay, yes. I'm doing my best, but it's not always
> enough, sorry. When you first reported that bug, I did not feel
> skilled enough to install an unstable build chain and test things in a
> certain manner. Plus, I was thinking that a swamp recompilation of the
> whole archive with gcc5 (once I've sorted out how to make an
> experimental image) may have been more efficient than a swamp of bug
> reporting "warning, *possible* issue ahead". But that's not easy
> either. I did not know how to react about your bug report, and then I
> forgot about it, sorry. 

The problem does not manifest itself with just a rebuild.  If you rebuilt the
package, and checked it's (not rebuilt) rdeps still work with the rebuilt
package, then sure, the package doesn't seem to be affected.  But the subject
"Widelands builds without any problem with gcc5" doesn't seem to suggest that.

> I must however confess that I feel surprised to be bashed for finally
> taking care of the problem at my tiny scale... But no offense taken: 
> being the maintainer of such an amount of important packages is a huge
> burden, and I'm still grateful for your work.

sorry, if you did that testing above; it didn't look like it from the terse
closing message.

Matthias