Re: [glob2-devel] concept for building priorities

[Joe Wells]

"Kai Antweiler" <address@hidden> writes:

>> >> >> Probably the time the globs take should be adjusted based on their
>> >> >> speed to get instead the time the "average" glob _would_ have taken.
>> >> >
>> >> > But if you want to compute a time period, you have to normalize the
>> >> > value.  Otherwise you can not compare UT and HT.
>> >>
>> >> I don't understand the two sentences above.  Can you clarify?
>> >
>> > If you just multiply speed and time you will get the distance (or
>> > better the length of the path) between glob and Inn.
>>
>> I don't understand.  How is this related to my proposal?
>
> Oh, I see.  You meant something different.
> When you wrote:
> "Probably the time the globs take should be adjusted based on their
> speed to get instead the time the "average" glob _would_ have taken."
>
> Probably corresponding to my comment:
> "We can reduce this effect somewhat, if we (let's say) multiply the
> time by the speed of the globs and maybe divide by the average speed."
>
> Dividing by the speed of the glob would not achieve anything.  We would and
> up with the original time again.  So I assume you meant we could skip the
> normalizing, but keep the speed times time.
> Now I think: you want to skip both.

I'm not sure I'm following.  Anyway, this seems to be discussion of
very low level details that would no doubt need lots of tuning and
adjustment.  So I'm not sure it is worthwhile for me to try to
understand.  It would be easier to comment on a concrete
implementation.

>> Also, what about traffic congestion?
>
> I don't know.  I always assume that globs always move and never stand still
> if they have a spot they can go to.  So traffic congestions would just 
> increase
> the length of the path.
> But now I think that is wrong, because unemployed globs are said to be 
> consuming
> less food.

My point was that it doesn't help a globule to be faster if traffic
congestion causes it to spend lots of time moving back and forth
without making real progress.  So we would not want to try to adjust
the actual time too much based on speed differences, because that
would lose valuable information about traffic conditions.

-- 
Joe