|
|
Boost Testing : |
From: Victor A. Wagner Jr. (vawjr_at_[hidden])
Date: 2005-05-09 23:37:36
At Sunday 2005-05-08 13:29, you wrote:
>The situation re this platform is characterised by the following facts.
>
>a) The serialization libraries fail to build on the VC 8.0 platform.
>b) With my local copy of VC 8.0 on my local machine all libraries build and
>all tests results are the same as with VC 7.1
>c) up until a short time a go there was anohter VC 8.0 platform on the test
>matrix which showed all tests passing.
>
>I checked the library build failure and traced into spirit. Spirit tests
>also fail. I'm quite confident that addressing the spirit failures will fix
>the serialization library build failure for VC 8.0
>
>What, if anything, should be done about this? There are a couple of options
>
>a) Presume that its a quirk with VC 8.0 and just wait for it to go away in
>the next beta
>b) Persume its an error in Spirit and fix spirit so it passes the spirit
>tests.
>c) Presume its an error in the serialization library.
>
>I'm not in a position to do b). So if that's the right thing I would hope
>someone would address it.
>
>I could do c) but would want to wait until b) isdone.
>
>So, for now, I'm doing a)
no, it's not a quirk in vc8.0 it's an illegal use of auto_ptr which I
didn't have a chance to chase further before my trip (specifically, someone
is doing a &(*someautoptr<blah>) where the current "value" of the pointer
is 0 .... I believe I posted a message which shows the section of the
specification which this violates.
I'm back from my trip now and will see if there is a relatively simple fix
to this architectural decision that &*someptr is a well-defined construct
(which I don't believe it is)
>Robert Ramey
>
>
>
>_______________________________________________
>Boost-Testing mailing list
>Boost-Testing_at_[hidden]
>http://lists.boost.org/mailman/listinfo.cgi/boost-testing
Victor A. Wagner Jr. http://rudbek.com
The five most dangerous words in the English language:
"There oughta be a law"