[Rivet] [Rivet-svn] r2970 - trunk/bin

[Hendrik Hoeth]

Thus spake blackhole at projects.hepforge.org (blackhole at projects.hepforge.org):

> Author: fsiegert
> Date: Thu Feb 24 20:05:19 2011
> New Revision: 2970
> 
> Log:
> Add nicer sort logic to rivet-mkhtml: Sort analyses by Spires ID and
> put everything without one at the back

Honest question: Why do you prefer this over the old sorting?

I don't know about you, but my brain doesn't work particularly well with
spires IDs. I think in terms of experiments and years. The old sorting
was a bit cumbersome, because it was reverse. But now ... let's have a
look at one of the examples from my world:

ATLAS_2010_S8591806
CDF_2009_S8233977
CDF_2008_S8093652
STAR_2006_S6870392
STAR_2006_S6860818
STAR_2006_S6500200
CDF_2005_S6217184
D0_2004_S5992206
CDF_2002_S4796047
CDF_2001_S4751469
CDF_2000_S4155203
STAR_2009_UE_HELEN
CDF_2009_NOTE_9936
CDF_2008_NOTE_9351
CDF_2008_LEADINGJETS
ATLAS_2010_CONF_2010_081
ATLAS_2010_CONF_2010_049
ATLAS_2010_CONF_2010_031

The first block is Spires IDs in reverse (!) order. The second block is
lexicographic in reverse (!) order. For me this means combining the
worst of all options (okay, purely random would top this). Just have a
look at how scattered for example the CDF analyses are. I would
definitely prefer the old sorting, or even better the old sorting but
not reverse.

Cheers,

   Hendrik

-- 
It pays to be obvious, especially if you have
a reputation for subtlety.    -- Isaac Asimov