<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Nov 11, 2014 at 12:42 PM, Akim Demaille <span dir="ltr"><<a href="mailto:akim@lrde.epita.fr" target="_blank">akim@lrde.epita.fr</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Le 11 nov. 2014 à 21:37, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> a écrit :<br>
<span class=""><br>
> Fixed in r221704 - again, apologies for the breakage... :/<br>
<br>
</span>Great, thanks!<br>
<br>
But...<br>
<span class=""><br>
>> On Mon, Nov 3, 2014 at 8:32 AM, Akim Demaille <<a href="mailto:akim@lrde.epita.fr">akim@lrde.epita.fr</a>> wrote:<br>
>><br>
</span><span class="">>> Is there some work around, such as asking for less accurate<br>
>> debugging symbols with some magic compiler flags?<br>
<br>
</span>Would you know such a means for pre-r221704?<br></blockquote><div><br></div><div>Unfortunately not... :( Really sorry about that. You would have to avoid this construct (in any context where the type might not be deduced immediately - so any class template or an out of line definition of an auto return type function). In those places you can still use a late return type specification ("auto x() -> foo").</div><div> </div></div><br></div></div>