Le Vendredi 25 Novembre 2005 15:53, gilh a écrit :
> Hi team,
Hi Gil,
> In reading over the wish-list I noticed that there is no mention of two
> areas that I think might be useful, namely 'estimation' and
> 'scheduling'.
>
> Estimation
> ----------------
>
> I find that one of the biggest flaws in any plan isn't so much the
> structure (although that in itself is quite important) but in the
> estimation of effort or duration required for a task. I also find that
> each type of project has tasks whose accuracy of estimation lies on some
> sort of difficulty scale. Let's take two examples[...]
>
> Stated as a 'wish', I think it would be helpful to provide a framework
> for dynamically loading and using 'estimation packages' into oopm. These
> estimate packages will need a standard interface for the oopm tool to
> activate and to retrieve the final estimate.
I agree. Each task should be associated to an "estimation function".
To continue on the same way, we could associate risk management with this
estimation module.
> Scheduling
> ----------------
>
> There are various flavours of scheduling engines that take a directed
> graph (nodes that reflect activities and edges that reflect
> dependencies, see also note 1) and schedule those activities based on
> various criteria (see note 2). Each engine seems to have benefits over
> another and is probably more applicable in one subject area over another
> (with performance differences as well).
>
> Rather than picking one scheduling engine (as most commercial products
> do) we should likely employ a dynamic behaviour in this area as well and
> have plugins that can perform the scheduling.
>
> Stated as a 'wish': oopm should have several scheduling engines to
> choose from and this choice should be controlled at run-time.
I also love this ideas.
While reading this mail, I had some more ideas. I write them while they are in
my mind (even if at the moment, they are not classified). Here is is one of
the ways I would like to use oopm :
I - I start defining resources :
To each resource, I would like to be able to define many parameters. For
example, to define "skills", or the "cost"... ? Some examples of skills:
- Thomas knows C++ and Python, Jack knows Administration and LDAP.
- A machine can produce Cars, or Trucks (but not both at a time).
2 - Then, I define tasks and the work associated to those tasks :
A - Coding an XML parser - work : 30 days - skills : C++, XML
B - Realising a mailing engine - work : 25 days - skills : LDAP, C++
3 - Then, I would like the engine to calculate how to share the work beetween
the different resources to optimize a parameter (time, cost...) and to
produce a planning.
With this, some constraints would be added. example :
- One resource should work at least 10 days in a raw on the same
project (so
that it does not lose time on working one day on a project the next one on
another project and each time spending some hours to remember where he left
the project).
- The project must be finished for the next month
- ....
The aim of this, is to show that a project management should be able, given
some data (resources, tasks and constrains) to produce a planning and let the
project manager how to reach the constrints associated to a project.
Hope this helps.
Regards,
Pierre-André
--
StarXpert - www.starxpert.fr
e-mail : [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
