Re: compilation errors from Xcode's new standard library

[Mike Solomon]

On Nov 3, 2013, at 2:03 PM, Mike Solomon <address@hidden> wrote:

> 
> On Nov 3, 2013, at 2:00 PM, Joseph Rushton Wakeling <address@hidden> wrote:
> 
>> On 03/11/13 13:36, Mike Solomon wrote:
>>> Doesn’t work, but all the files in flower compile fine with gcc, so I’m a 
>>> happy camper.
>>> Apple’s home cooked clang is not free software, so there’s no reason to 
>>> expect free software to compile with it.  I don’t mind giving up on it.
>> 
>> It's difficult to see how they could have cooked it to the point where what 
>> is AFAICS fairly standard C/C++ fails to compile with it.
>> 
>> Anyway, if you want to leave it here, no worries.  I just thought it might 
>> be useful for other people if we could tie down what the source of the 
>> problem is.
>> 
>> Just for reference, do you have a log of what happens when you try running 
>> ./configure with CC=clang and CXX=clang++?
>> 
> 
> Same exact problem with CC=clang and CXX=clang++.
> It’s not that I want to leave it here, but rather that I need to spend time 
> engraving. When I have time, I’ll try to make a minimal example and send it 
> to Apple.
> 
> Cheers,
> MS

Another thing (dunno if it’s relevant)…
The lily/out/lexer.cc generated by flex had the type size_t used in definitions 
of LexerInput and LexerOutput.  Something barfed during make and told me these 
had to be int, so I changed them by hand and everything kept chugging along.  
This seems kludgy, though…I’m guessing that multiple versions of flex may be 
being used, but I’m not sure.

Cheers,
MS