Re: [Planner Dev] Prop. for new feat. - first patch

[Matteo Nastasi <nastasi alternativeoutput it>]

On Thu, Jan 19, 2006 at 09:47:58AM +0100, Richard Hult wrote:
> Hi,
> 
> Matteo Nastasi wrote:
> > On Wed, Jan 18, 2006 at 08:08:44PM +0100, Richard Hult wrote:
> >>I realize that I should clairify what I meant here. I mean that it would 
> >>be nice to get the same kind of delaying of lower prioritized tasks no 
> >>matter if the (so-called vampire) tasks are fixed in time or not, and 
> >>just let the priority decide which task gets the resource usage first.
> >>
> >>This would improve the scheduling a lot and we would get rid of the very 
> >>annoying synthetic dependencies or fixed dates that you are forced to 
> >>use now to get tasks scheduled properly.
> >>
> >>I haven't looked closely at the patch yet to see whether that would be 
> >>possible though. What do you say Matteo, does it make sense to you?
> > 
> > Vampire task are the first step to go in this direction.
> > 
> > In my mind vampire tasks are a special (and simple) case of much more
> > complex scenario of full priorities management; vampire is a "preview"
> > of it without full implementing it and solves the common case (for me)
> > of temporary re-allocation of some resource to a not planned task (and
> > not necessarly project related).
> > 
> > To talk about priorities (IMHO) we must open a discussion to identify
> > very well what are independent variables, what policies need to apply them,
> > if the data behind planner is adeguate to our needs, a set of sufficient
> > use cases, if there are some cases where there are not univocal
> > policies and so on.
> > 
> > In addition to this I don't know if it's necessary some form of "real
> > resources usage tracker" where eventually store info about the real flow
> > of the project.
> > 
> > I agree the opinion that see vampire only as a particular case of high
> > priority task, in fact the graphical code just don't use this property
> > except to magnify it with red dashed border.
> 
> OK, that sounds good enough to me. If I understand the solution
> correctly, there will not need to be any changes to the file format for
> example, since only the already existant fields will be used (fixed date
> + priority).
Yes, it is correct.

> By the way, would it be possible to split the patch into the drawing
> parts for non-working time and the rest? It's going to be a pretty big
> patch so that would help a lot.
I think that it is possible, but I worning that the 2 patches are very 
unbalanced, vampire are maybe the 20% or less of the total.

It's possible to disable the vampire feature with #ifdef instead of
split the patch ? 

If I will see that the 2 patches are more balanced I will review
my proposal.

> /Richard

Regards, Matteo.

> -- 
> Imendio AB, http://www.imendio.com/
> _______________________________________________
> Planner-dev mailing list
> Planner-dev lists imendio com
> http://lists.imendio.com/mailman/listinfo/planner-dev

-- 
 Matteo Nastasi  -  Milano - Italy  | HomePage: www.alternativeoutput.it
 Sostenere e supportare GNU/Linux ! | IRC:    #linux-mi irc freenode net   
 Milano Linux                       | E-Mail:   matteo nastasi milug org 
        Users Group  www.milug.org  |       nastasi alternativeoutput it