[omniORB] issues with omniorb(py) on solaris

[Tim Docker]

Hi,

My testing indicates that the problem has been fixed in the CVS
snapshots:

	omniORB-4.1-latest.tar.gz
	omniORBpy-3-latest.tar.gz

This is good. However, I'm hesitant to release a system that depends
on a CVS snapshot. When do you plan to make a new release? Assuming
it's not very soon, could you point me at the appropriate code where
the bug fix would lie, so that I can manually create a patch over
the existing 4.0.6 release?

Thanks for your help.

Tim


-----Original Message-----
From: Duncan Grisby [mailto:duncan at grisby.org]
Sent: Monday, 12 December 2005 11:28 PM
To: Tim Docker
Cc: omniorb-list at omniorb-support.com
Subject: Re: [omniORB] issues with omniorb(py) on solaris 


On Friday 9 December, "Tim Docker" wrote:

> I am having problems using omniorbpy under solaris. The
> software versions are:

[...]
> The problem I am seeing is that there is some sort of alignment
> error with sequences of doubles. As you can see below, the server
> recieves the sequence incorrectly from the client, and the sequence
> is further corrupted on transfer back to client.

I expect it's a strict aliasing issue to do with the marshalling code.
It has been fixed in CVS, so try the latest CVS version, or the
snapshots from here:

  http://omniorb.sourceforge.net/snapshots/omniORB-4.0-latest.tar.gz
  http://omniorb.sourceforge.net/snapshots/omniORBpy-2-latest.tar.gz

Cheers,

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --