[kplato] Summary of Data Discussion

[Thomas Zander]

On Tue, Jun 19, 2001 at 05:34:32PM -0500, Jim Sabatke wrote:
> They would be calculated just like standard PERT values.  I think the 
> best way to describe it is to put together a diagram of what gets done. 
>  I'll see if I can generate something meaningful in the next day or so.
> 
> bilbo wrote:
> 
> >On 19 Jun 2001, at 16:43, Jim Sabatke wrote:
> >
> >>Every line should have a confidence (risk) associated with it.  However, 
> >>with tasks that have subtasks, the values should be computed, not entered.
> >>
> >On what basis? Are you going to sum all the confidences for all the 
> >subtasks? Just those on the critical path? Or perform some 
> >statistical calcualtion like "half the confidence for each subtask on 
> >the critical path". I think that we should at least consider having a 
> >separate confidence even for a task with subtasks, but it could be 
> >optional.
> >In many projects I have been involved with the 
> >contingency/confidence/risk has been an uncertainty, and thus a 
> >plus/minus value, not just possible overrun. This is why I am 
> >suggesting a separate value for task - if you add up all the 
> >plus/minus values you can get back to zero.
> >
> >It seems that the terms Contingency, Confidence, Risk and 
> >Uncertainty are possibly all referring to the same thing, but maybe 
> >not. I see it as either underrun or overrun and not just overrun.
> >But then I am only familiar with software projects.
> >Yes even these sometimes come in ahead of time!
> >
> >
> >
> >regards,
> >        Bill


The thing I am still not sure about is what the user should enter 
for risk.

If risk is a plus, should the duration of the task be noted (10% or so for
slip?)
If there is a large risk (this person is know to be sick often) should we store 
this in the task? And what could the diagram show at this time? A slippage
of a predefined time, a precentage time etc..

In summery; risks are factors we can't put our finger on, I want to know
how the software should handle these things in calculations and display.

-- 
Thomas Zander                                            zander@earthling.net
The only thing worse than failure is the fear of trying something new