Fix memory errors in Subsurface

[Lubomir I. Ivanov]

On 30 November 2013 19:31, Dirk Hohndel <dirk at hohndel.org> wrote:
> On Sat, 2013-11-30 at 19:30 +0200, Lubomir I. Ivanov wrote:
>
>> yeah, marble is a goldmine for leaks.
>
> I don't think it is "gold" that you can mine with that code...
>
> Looks pretty, but never try to read it. And never ask Thiago to talk
> about that code, either...
>

sorry if sounded ironic right there, but that's just the case, often...
leaks are leaks - persistent or program heap lifespan we are dealing
with slightly irresponsibly written software.
and it the long run this does not make any sense to be fixes
realistically, but this has to be fixed for clarity - given the power
of open source and the fact that such might clobber debuggers, even
where filtering is an option...

something leaking 16 bytes for an user interaction, where he/she has 8
gigs of ram does not mean this is a good programming practice!
so my point is, if it's not *really* leaking in something core/OS like
MSRVC or some MINGW/GCC depending implementation that can hardly be
touched for a particular version, then we can perhaps still suggest
changes to the third party tools to improve...

hope that makes sense.
lubomir
--