[vtkusers] Creating an ImageData from the bounds of a PolyData

[David Doria]

On Tue, Dec 8, 2009 at 7:56 PM, Bill Lorensen <bill.lorensen at gmail.com> wrote:
> Also, we must take into account the cost of each line of code we add
> to the toolkits. A few years ago I estimated that each line of code
> (including comments) cost about $30. I just checked ohloh.org and
> vtk's cost is estimated at about $45 million assuming $100K per person
> year (which is conservative I think). That works out to a bit over $20
> per line.
>
> paraview: $63 million
> itk: $24 million
>
> Bill

@David G. - I hear ya, I'll proceed with making filters for these
things. I've had a hard enough time getting small functions added -
proposing a total interface redesign seems just about out of the
question.

@Bill - Two things - 1) you've assumed that all lines of code are
created equally. It clearly takes a long time to do the software
planning and algorithm design/implementation and hardly any time to
add a little accessor. 2) If you add the RIGHT lines of code to the
toolkit, you can likely REDUCE the total number of lines. I'm not sure
how many places need to get an image data from a point set, so maybe
this is not the best example. But if you have the same big messy loop
appearing in multiple places and you can just move it to the right
place and call it from each of the places it used to be, the code gets
shorter, cleaner, and most of all easier to read and use.

Thanks,

David