Re: [boost] Sun-5.8 failures

From: Simon Atanasyan (Simon.Atanasyan_at_[hidden])
Date: 2006-08-08 07:38:29

John Maddock wrote:
> Simon Atanasyan wrote:
>> John Maddock wrote:
>>> includes the Boost.TR1 version of <memory> which forwards on the real
>>> version as well as including it's own extra parts. However, for some
>>> reason the Boost versions of these headers appear not to be being
>>> picked up.
>> The root of this problem is a surprise for me. Actually to include
>> <memory> Sun C++ searches file named memory.SUNWCCh. This file is a
>> symbolic link to "memory". Please don't ask me to explain the
>> rationale of this "feature" :-) So you can create symbolic links in
>> the tr1
>> folder like memory.SUNWCCh -> memory. It should help. Meanwhile, I
>> will try to persuade my co-workers to implement command line option
>> to turn off this feature.
> 
> It turns out this is harder for us to fix than I imagined: up until now we 
> have had a prohibition on symbolic links in cvs for the obvious portability 
> reasons.  We also require all_lower_case filenames for the same reason.  I 
> note that STLport seems to get around this without resorting to these names, 
> but I can't see how, any ideas?  Or is this a recent compiler feature?

It's old compiler feature.
I suggest to create a shell script to create necessary symlinks. User will
be able to run it by hand. And Jamfile used to run Boost test suite will
able to run this script too. I know this is quite ugly workaround.

--
Simon