Hi folks -

A bit of  bad news - it would appear the lack of an automated  
regression test for the static dance stuff has caused it to break  
somewhere between x.5.5 and x.6.0.

I am tracking the problems now - I am not sure of the extent of the  
breakage, I will be tracking the problem tonight and will kepp yall  
up to date on my progress.  Needless to say, once I get it fixed I  
will port the changes forward to x.6.1 and add tests for all extant  
examples.

Thanks,
/-Will

Will,

Indeed. I've been trying to track it down for the last hour or so. Actually, I had contributed $CIAO_ROOT/examples/Null_Component, which doesn't seem to work anymore :(

It should be possible to add this to autobuild. Johnny and I worked on adding a static CIAO build under MSVC, so perhaps we can add that test to it.

Thanks,
Abdul

Hi Abdul -

The attached patch will allow the Hello example to deploy, but it is  
probably not the Right Thing To Do for the long term.   Longer term  
solution would be to re-architect some of the RT-CCM stuff that  
Nanbor has added so we can take advantage of it in static  
deployments, which is not currently possible given how the static  
dance bits are implemented.

That will probably require a bit more time than we have for the  
current demo, however.

Thanks,
/-Will

Hi Will,

We've been aware that RT-CCM stuff doesn't work for Static DAnCE for a while now. We opened a ticket for this (Bugzilla 2736):
http://deuce.doc.wustl.edu/bugzilla/﷒0﷓

Even it is not the "right thing to do" [TM] for the long term, would it be possible to commit the short term fix for Static DAnCE and add note to the bugzilla entry above (or start a new one)? That would at least enable us at LM to continue our work with CIAO until a long term solution is reached (the dynamic DAnCE is not an option for us since we're working with legacy hardware that doesn't have much memory).

Thanks,
Abdul 

> Even it is not the "right thing to do" [TM] for the long term, would it
> be possible to commit the short term fix for Static DAnCE and add note
> to the bugzilla entry above (or start a new one)? That would at least
> enable us at LM to continue our work with CIAO until a long term
> solution is reached (the dynamic DAnCE is not an option for us since
> we're working with legacy hardware that doesn't have much memory).

Thats my intention - I have also been able to get the BasicSP example deployed.  I will get this fix in for x.6.1, as well as daily tests so it does not happen again.

Thanks,
/-Will

Tue Sep 18 15:13:37 UTC 2007  Abdullah Sowayan  <abdullah.sowayan@lmco.com>

        * DAnCE/NodeApplication/NodeApp_Configurator.cpp:

          StaticDAnCE applications stopped working since the introduction
          of this file. This change fixes the problem. For details look
          into Bugzilla 3073. Thanks to William R. Otte for providing 
          the patch.

We'll keep this ticket open until we add some regression test to make sure that StaticDAnCE doesn't break again.