
hi,
When i try to compile the code in listing 1 below using MSVC7.1 i get a
whole host of errors of the same form as that given in listing 2. I've
looked around on google but either i'm not concentrating or this issue
hasn't come up yet. I've also had a quick look in boost/format but would
rather not hack around with stuff i don't know. Does anyone know what i
need to do to get this working?
Thanks in advance,
Ian Whittley
Listing 1:
#include

Compiles with MSVC 6.0.
"Witz"
hi,
When i try to compile the code in listing 1 below using MSVC7.1 i get a whole host of errors of the same form as that given in listing 2. I've looked around on google but either i'm not concentrating or this issue hasn't come up yet. I've also had a quick look in boost/format but would rather not hack around with stuff i don't know. Does anyone know what i need to do to get this working?
Thanks in advance,
Ian Whittley
Listing 1:
#include
int main() { boost::numeric::ublas::triangular_matrix<int> m; }
Listing 2:
'C:\Developer\boost_1_30_0\boost\numeric\ublas\matrix_expression.hpp(781): error C2027: use of undefined type 'boost::numeric::ublas::vector_matrix_binary
::const_iterator2' Info: http://www.boost.org Wiki: http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl Unsubscribe: mailto:boost-users-unsubscribe@yahoogroups.com
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Hi Ian, you wrote:
When i try to compile the code in listing 1 below using MSVC7.1 i get a whole host of errors of the same form as that given in listing 2. I've looked around on google but either i'm not concentrating or this issue hasn't come up yet.
John Maddock reported and explained this problem to me some time ago (thanks!) and I've (hopefully) implemented a *ugly* workaround for it. The problem is due to the questions of DR 135 and its (new ;-) interpretation through MS. The correct solution probably would be to separate class member declarations and definitions, but then the code probably wouldn't work with MSVC 6.0 anymore :-(.
I've also had a quick look in boost/format but would rather not hack around with stuff i don't know. Does anyone know what i need to do to get this working?
You can download one of the newer version from http://groups.yahoo.com/group/ublas-dev/files and try to enable // #define BOOST_UBLAS_NO_NESTED_CLASS_RELATION in the MSVC section of ublas' config.hpp. But be aware that you're probably one of the first to give this a try ;-) [snip] Thanks, Joerg

Hi Ian,
you wrote:
When i try to compile the code in listing 1 below using MSVC7.1 i get a whole host of errors of the same form as that given in listing 2. I've looked around on google but either i'm not concentrating or this issue hasn't come up yet.
John Maddock reported and explained this problem to me some time ago (thanks!) and I've (hopefully) implemented a *ugly* workaround for it. The problem is due to the questions of DR 135 and its (new ;-) interpretation through MS. The correct solution probably would be to separate class member declarations and definitions, but then the code probably wouldn't work with MSVC 6.0 anymore :-(.
I've also had a quick look in boost/format but would rather not hack around with stuff i don't know. Does anyone know what i need to do to get this working?
You can download one of the newer version from http://groups.yahoo.com/group/ublas-dev/files and try to enable
// #define BOOST_UBLAS_NO_NESTED_CLASS_RELATION
in the MSVC section of ublas' config.hpp. But be aware that you're
I've been trying to compile the Bayes++ code which requires ublas, so I've
been in the learning-about-boost phase since yesterday and ran into this
myself using VS.NET 2003.
I've downloaded the 5/19 release of ublas and placed it in the 1.30.0
release of boost, and had to:
1) uncomment the #define BOOST_UBLAS_NO_NESTED_CLASS_RELATION
2) Add "#define BOOST_UBLAS_TYPENAME typename" to the MSVC section
('typename' being what's different) (yeah, I know I'm breaking VC6, and it
should be #if'd in there if it's going to stay, but I'm just trying to get
this to compile)
#1 got rid of a lot of errors, and #2 got rid of everything else but 3
errors.
So, I'm left with trying to get my head above all the template usage. If
anyone can provide some insight, I'd appreciate it. This is probably a
little more towards the Bayes++ side, not boost, but thought I'd ask here
since I did want to mention the 'typename' issue and VC7.1 which I hadn't
seen mentioned yet:)
Here's the error output:
c:\Documents and Settings\dswallow\My Documents\Visual Studio
Projects\Bayes++\BayesFilter\uBLASmatrix.hpp(544) : error C2664:
'boost::numeric::ublas::matrix_matrix_binary
one of the first to give this a try ;-)

Hi Doug, you wrote:
I've been trying to compile the Bayes++ code which requires ublas, so I've been in the learning-about-boost phase since yesterday and ran into this myself using VS.NET 2003.
I've downloaded the 5/19 release of ublas and placed it in the 1.30.0 release of boost, and had to:
1) uncomment the #define BOOST_UBLAS_NO_NESTED_CLASS_RELATION 2) Add "#define BOOST_UBLAS_TYPENAME typename" to the MSVC section ('typename' being what's different) (yeah, I know I'm breaking VC6, and it should be #if'd in there if it's going to stay, but I'm just trying to get this to compile)
Thanks, I'll use this information to prepare a corresponding config section in the next days.
#1 got rid of a lot of errors, and #2 got rid of everything else but 3 errors.
So, I'm left with trying to get my head above all the template usage. If anyone can provide some insight, I'd appreciate it. This is probably a little more towards the Bayes++ side, not boost,
You can verify this with the ublas regression tests (libs/numeric/ublas/...).
but thought I'd ask here since I did want to mention the 'typename' issue and VC7.1 which I hadn't seen mentioned yet:)
Here's the error output:
[snip diagnostics] Thanks for your feedback, Joerg
participants (4)
-
Doug Swallow
-
jhr.walter@t-online.de
-
Patrick Kowalzick
-
Witz