Subject:
Re: DRAFT minutes of the reco data model meeting
From:
David Rousseau <rousseau@lal.in2p3.fr>
Date:
Mon, 8 Oct 2001 17:50:59 +0200 (CEST)
To:
David Adams <dladams@bnl.gov>

----------------------------------------------------------------------------
On Mon, 8 Oct 2001, David Adams wrote:


> David:
> 
> I understand the requirement for partial unpacking. My question is about 
> the input stream. Do we need to impose the requirement just for raw data 
> or also for digits that have been written to StoreGate (presently 
> objectivity or root)? I have assumed only the former.
> 
> I assume the digits will not be saved for most events and the typical 
> job of the digit container is to unpack raw data rather than fetch 
> previously saved digits from disk.

I don't think you can make this assumption. People are very nervous
about it. I've not heard anyone revoice this idea recently.


> 
> I get back to you on the documentation.

Not necessary any more in fact, Hong has just promised to release a
better documentation within a few days, once the code is in the cern sw
repository


> Cheers
David