[omniORB] Windows98 Problems
Christopher Fahey
christopher@faheys.org
Wed Feb 12 19:54:02 2003
this problem is now manifesting itself on Windows2000. i have moved to the
lastest release of omniORB4 but the problem persists. i turned the tracing
on the server to trace level 40 and the data being output is definitely
different from what was supposed to have been sent. i have written numerous
test programs to determine the problem but from what i can tell, it is
within the marshalling of the data that the problem exists.
one of the programs i have written has the client and server embedded in the
same application. there was NO problem sending the data through the
interface.
again, here is the interface:
struct ImageFile
{
string name;
sequence<octet> data;
};
interface ImageService
{
boolean addImage( in long id, in ImageFile file, in long index );
};
i am really at a loss and need to get this working asap.
chris fahey
***************************
>yes, i have confirmed that the data is being read correctly in all
>instances. i will try your suggestion. thanks.
>
>chris
>
>On Tue, 2003-02-04 at 05:17, Duncan Grisby wrote:
> > On Thursday 30 January, Christopher Fahey wrote:
> >
> > > i am hoping someone can help me with this unusual situation. i have
an
> > > application that sends images from a workstation to a server using
> > > omniORB3.0.5. (i plan on going to 4 but not right now since i am near
> > > release). the client code is compiled in c++ for linux, windows xp,
windows
> > > nt, and windows 9x. the problem i am having is that the windows 9x
client
> > > is corrupting the data when it sends it. the data is fine until it is
> > > loaded into a sequence<octet> object that is being sent, but when it
reaches
> > > the server, the the data in the sequence is corrupt.
>
> > Mysterious. I can't see anything wrong with your code. Have you
> > confirmed that the data is correct just before you send it? Try
> > running with trace level 40 to see the data transmitted. That will
> > show if it's the sending side or the receiving side getting it wrong.
> >
> > Cheers,
> >
> > Duncan.