[
Date Prev][Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [tsf-dev] TSF process feedback, part 3: aggregate scores, changing evidence
|
Sam,
does have links to other statements, then Trudag generates a score
based on some maths that I don't fully understand:
<https://pages.eclipse.dev/eclipse/tsf/tsf/model/scoring/implementation.html#calculating-scores-for-all-items>.
It treats the graph as a Markov chain.
I assume that it requires the graph to be acyclic because
this would require an infinite number of multiplications.
There is a matrix multiplication identity that can be used
to handle the infinite multiplication requirement.
See example in section 9.4.1 here http://knosof.co.uk/ESEUR/
The Markov chain approach assumes that nodes are independent.
Some are probably not independent, so something else is
needed.
In our experience, the further away you get from edge nodes in the
graph, the less useful these aggregate scores become. The top level
"TRUSTABLE-SOFTWARE" number can vary wildly. Imagine you have 2
Wild variation implies that the system is highly dependent on
a few nodes. A sensitivity analysis ought to highlight which nodes
these are, and they can be marked as critical failure points and
the actions taken to 'protect' them.
Lesson: be very wary of taking any meaning from the aggregate scores.
You mean they are totally useless? So why bother calculating them?
--
Derek M. Jones Evidence-based software engineering
blog:https://shape-of-code.com
