bug-commoncpp
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bayonne-devel]Compile error on commoncpp2-1.0.0


From: Federico Montesino Pouzols
Subject: Re: [Bayonne-devel]Compile error on commoncpp2-1.0.0
Date: Tue, 17 Sep 2002 23:43:04 +0200
User-agent: Mutt/1.4i

        This is now fixed in CVS. I am afraid that there will be more
issues around HAVE_SSTREAM in the future...

        Both XMLRPC and URLStream have had to be modified.

On Sun, Sep 15, 2002 at 07:11:34AM -0400, David Sugar wrote:
> I think a lot of other Common C++ related things (in particular, Bayonne and 
> ccscript) still use strstream.  I will need to update these to work with 
> HAVE_SSTREAM also, but to enable use of sstream :).
> 
> On Friday 13 September 2002 14:29, Federico Montesino Pouzols wrote:
> > Yes, HAVE_SSTREAM seems necessary. I think this only involves
> > modifying the XMLRPC class. The new sstream was incorporated not too
> > much ago, so adding #ifdef's right now is not too dificult. However,
> > it may become a maintenance headache in the future :)
> >
> >     I have added this to the queue of fixes and additions that I
> > should incorporate soon. Let me know if there is a better solution.
> >
> > On Wed, Sep 11, 2002 at 06:33:11AM -0400, David Sugar wrote:
> > > Yes, we have to have a warning.  We may also need later to have a
> > > HAVE_SSTREAM check that does either old or new style, which means a lot
> > > of conditional code :)...This will be needed because we do sometimes need
> > > to target C++ compilers that do not have sstream yet...
> > >
> > > On Monday 09 September 2002 14:49, Jason Spence wrote:
> > > > On Sun, Sep 08, 2002 at 05:29:07PM +0200, Federico Montesino Pouzols 
> wrote:
> > > > > On Sun, Sep 08, 2002 at 02:17:49AM -0700, Jason Spence wrote:
> > > > > > On Sun, Sep 08, 2002 at 05:23:49AM +0300, Vitaly wrote:
> > > > > > > Hi,
> > > > > > >
> > > > > > > I am having trouble compiling commoncpp2-1.0.0 on Red Hat 6.2
> > > > > > > (Kernel 2.2.14-5.0).
> > > > > > > I get to a point where I receive the following error:
> > > > > > >
> > > > > > > /bin/sh ../libtool --mode=compile
> > > > > > > c++ -DHAVE_CONFIG_H -I. -I. -I../include -I../src
> > > > > > > -DCCXX_EXPORT_LIBRARY  -I/ usr/local/include -D_GNU_SOURCE
> > > > > > > -I../include -c socket.cpp
> > > > > > > rm -f .libs/socket.lo
> > > > > > > c++ -DHAVE_CONFIG_H -I. -I. -I../include -I../src
> > > > > > > -DCCXX_EXPORT_LIBRARY -I/u sr/local/include -D_GNU_SOURCE
> > > > > > > -I../include -c socket.cpp   -fPIC -DPIC -o .libs/socket.lo
> > > > > > > socket.cpp:51: sstream: No such file or directory
> > > > > > > make[1]: *** [socket.lo] Error 1
> > > > > > > make[1]: Leaving directory `/root/Work/commoncpp2-1.0.0/src'
> > > > > > > make: *** [all-recursive] Error 1
> > > > > >
> > > > > > This is a bit different from the problem I'm having, but I do know
> > > > > > that the commoncpp2 build is currently broken in at least one other
> > > > > > way.  You must apply the attached patch to get it to work.  I've
> > > > > > already sent it to David for application.
> > > > > >
> > > > > > Unfortunately, David is stuck in some third world country with
> > > > > > spotty Network access at the moment and hasn't seem to have applied
> > > > > > it yet
> > > > > > (http://savannah.gnu.org/cgi-bin/viewcvs/commoncpp/commoncpp2/inclu
> > > > > >de/c c%2B%2B/)
> > > > >
> > > > >       Hi, this is now fixed in CVS. My apologies, since I introduced
> > > > > this problem.
> > > > >
> > > > >       However, this bug was introduced after 1.0.0 was released. The
> > > > > make errors above complain that the sstream header is missing. I see
> > > > > that Vitaly uses Red Hat 6.2, which probably comes with a gcc release
> > > > > older than 2.9x that does not have the newer sstream C++ header.
> > > > >
> > > > >       We are using sstream to follow the newer C++ standard and gcc
> > > > > 3.x releases. Perhaps we will have to add some configuration stuff to
> > > > > keep compatibility with old systems.
> > > >
> > > > Do you think you could put a warning in configure.in if the check for
> > > > sstream fails as well as instructions to the user on how to upgrade to
> > > > 3.x if necessary?  I'm sure it would cut down on the "Common C++ is
> > > > broken" posts to the list :)




reply via email to

[Prev in Thread] Current Thread [Next in Thread]