Re: [boost] re: [Thread] The last step

From: Martin Wille (mw8329_at_[hidden])
Date: 2004-10-02 15:51:57

Aleksey Gurtovoy wrote:
> Michael Stevens writes:
> 
[...]
>>
>>I think the problem is that the compiler is running on a platform
>>where the version of glibc is unsupported (too recent). I have come
>>across the same problem myself. After much research I found the only
>>workaround was to rename strol in the compilers libcprts.a This is a
>>big hack but gets things compiling.  I have a already manipulated
>>libcprts.a but it is 1M. I can post it to whoever needs it however.
> 
> 
> Martin, should we go down this route, or should I mark up the
> failures as expected with a corresponding note?

That would work for the sake of the tests.

However, this is an extremely ugly hack, and it would give
a user the impression Boost would work without a problem
on that platform while this is not entirely true.

The only option I see would be to run two separate toolsets,
one of them with the hack, the other one without it, and to
markup accordingly. However, this puts more burden on the
those who run the intel 8 tests (like me :o) and I'm not
sure this is worth the effort.

To me, it looks like this is Intel's problem and we should
clearly mark it so. Of what use is a compiler for a platform
that only supports rather outdated environments? Frankly,
I'm quite annoyed by the number of problems we get with
respect to compatibility of the intel compiler. Even more so
when I see it advertised as being compatible. We're facing an
incompatibility regarding to strtol() here!! I hope this problem
will go away with 8.1 (which I still have to install).

In short: I think we should mark up the failures as expected.

Regards,
m