[Rivet] Imminent YODA 1.5.0 release planned

[Andy Buckley]

No complaints, so I'm going to make a last little check (related to 
parser forward compatibility with a block format version) and then 
release YODA 1.5.0. A test build of Rivet 2.3.0 on lxplus worked fine, 
so I will also update the Rivet bootstrap script to use this version.

Andy



On 26/08/15 16:04, Andy Buckley wrote:
> Hi all,
>
> I'm planning to make a release of YODA 1.5.0 in the next day or so,
> unless there are objections.
>
> The main change in this version is that the Boost Spirit parser
> dependence is now gone, replaced by a *much* simpler state-machine
> parser which also adds the content sensitivity that our 1000s of lines
> of Spirit/Fusion overload magic wasn't complicated enough to handle ;-)
>
> I've also changed the YODA formatting a little, to remove unnecessary
> "#" characters. This should be backward compatible. I would like to add
> some format versioning indication for each block, to allow us to extend
> and improve how we represent data objects e.g. so we can at some point
> add multiple error support to the Point lines in Scatter blocks... but
> that would *not* be backward compatible so we'll have to wait until this
> new parser is ubiquitous before making that leap.
>
> Other than persistency, I've improved/fixed a few things in the
> histogram division & efficiency routines which make them more robust
> against unfortunate weighting histories. This solves some crashing bugs
> for ATLAS tuning, and is the main motivation to get this version of YODA
> released asap. Current versions of Rivet can already use it, and we'll
> aim as promised for a new Rivet 2.4.0 release in September.
>
> Everyone happy? Shout now, please, before I release this thing.
>
> Andy
>


-- 
Dr Andy Buckley, Lecturer / Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow