<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Nov 17, 2015 at 8:07 PM, Jonas Ã…dahl <span dir="ltr"><<a href="mailto:jadahl@gmail.com" target="_blank">jadahl@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5"><br>
> We'll see how that patch is received. If anyone complains it breaks<br>
> their thing, I think we have to revert it, because it is technically<br>
> breaking the stability rules.<br>
<br>
</div></div>The most obvious thing that would "break" is that every client or server<br>
C implementation will suddenly get new compiler warnings and to be able<br>
to build both the version before and after the change without warnings,<br>
they'd need #if #else #fi macros around every such place.<br>
<br>
This is primarily an annoyance anyhow since the warning is mostly<br>
harmless.<br></blockquote><div><br></div><div>The code could use just one #if to define GEOMETRY(x) to either be int(x) or (x) depending on the protocol header version, and then call GEOMETRY(unsigned) everywhere it needs to pass the enumeration to a protocol function.<br><br></div><div>Though I think the best solution is to just fix the code so it does not produce warnings for the newer protocol files and ignore the warnings for old ones.<br><br></div><div><br></div></div></div></div>