[Insight-developers] Integrating the New Statistics framework into ITK

[Luis Ibanez]

Bill,


We *are* providing backward compatibility by leaving the existing
Statistics Framework there and carrying it along for many years to come.


We are already going to go through the pain of supporting and
maintaining two directories (and possibly two namespaces).


Modifying the new design to make it look closer to the old one, is
contrary to what motivated the design changes in the first place.


   We are getting to the point where ITK is looking like DOS...
   and that is a sad state of affairs.


   Progress happens through cycles of evolution and revolution.
   We have kept a steady evolution for a decade now. And the leaks
   and patches on the hull of our ship are starting to make water.


   The fact that we have been spending more time in discussion how
   a recent patch will not break the patch of a previous patch,
   rather than spending time on adding new algorithms to the
   toolkit by moving contributions from the Insight Journal...
   is a bad sign.


ITK doesn't impose automatic updates on its customers.
They upgrade to new versions of ITK out of their own will,
and in some cases there is an adaptation cost associated
with it.


I see the point that we shouldn't make capricious changes in the
toolkit just for the sake of change... but we are reaching the
point were the advantages of Agile Development have been sacrificed
for the benefit of the backward compatibility... and what it probably
of most concern is that we are not making a good use of the time that
volunteer developers are dedicating to the toolkit.


At some point the software should be revamped or we will risk
that a community of progressive developers will decide that it
is simpler to develop a brand new toolkit. GDCM already moved
ahead and we are missing to take advantage of a better DICOM
support by sticking to their older version.


A system lacking flexibility will end up breaking...


      Luis


--------------------
Bill Lorensen wrote:
> I would not give up too hastily the backward compatibility potential
> of the new statistics. A t first glance, it seems that some additional
> effort should be made to make the new statistics more compatible than
> it is currently.
> 
> A simple replacement is simple for us developers, but does not
> consider our large community of customers that rely on a stable API to
> build their applications and products.
> 
> Bill
> 
> 2009/4/9 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>:
> 
>>Hi,
>>
>>Le 9 avr. 09 à 20:19, Karthik Krishnan a écrit :
>>
>>
>>>I think from developer feedback, Option A has the widest support. Is that
>>>the general consensus ?
>>
>>
>>I think it would be great to consider the simple non-backward compatible
>>replacement — maybe at list to discuss it.
>>If non-backward compatible changes is something we can think about for ITK
>>4.0, doing a simple replacement with the new framework would save a lot of
>>valuable time and complexity.
>>
>>Regards,
>>
>>Gaëtan
>>
>>--
>>Gaëtan Lehmann
>>Biologie du Développement et de la Reproduction
>>INRA de Jouy-en-Josas (France)
>>tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
>>http://voxel.jouy.inra.fr  http://www.mandriva.org
>>http://www.itk.org  http://www.clavier-dvorak.org
>>
>>
> 
> _______________________________________________
> Powered by www.kitware.com
> 
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
> 
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
> 
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>