Critial path : total float & longest path option

Khuong,

I’m interested in your thoughts on the multiple calendar idea and the resultant float, as I’m relativley new to Primavera (Going on 7 months now) I have been interested as to how the float is calculated, mostly because I get some weird answers that don’t make sense.



As I deal with hard mandatory constraints on a daily basis given the nature of our work and the unfortunate circumstance of the customer issuing the contracts way too late, and the now added complexity of using resource leveling, any converstation on this topic or the like is of great interest to me at this time.



Apart from Ron and Paul’s sites, does anyone else have any other information they could pass onto me?



[email protected]

Thanks

Andy



P.S. Paul Harris has just recently published a paper I wrote on resource leveling, and I can say that I’m seeing many benefits in my schedules as a result of this method, I’m currently working on a range of processes, including code structures and resource assignment protocals to further enhance these ideas of mine.

In my own thinking, a contract which imposes a milestone with mandatory constraint is considered bad contract....more worst thing if it falls along the critical path..

Hi All

My point of view,Suppose according to contract you should define some milestones which should be fixed by constraint on the schedule(contract obligations),during execution phase,all activities which havn’t met mentioned milestones and passed,have total float=0 or less than 0,but are not necessarily on the longest path,therefore we have some activities on the longest path that could have float and some other activities which have 0 or negative total float without being on the longest path.



kian

Enterprise will defer P3 in calendar algorithm, P3 uses calendar in predeccesor while P5,P6 use either succ or pred cal.

The presentation you mentioned only covers P3.



Does P5 discount hammocks and give a true longest path?



Oliver

Thanks Dieter for the paper. Thats Beautifully written.



Dont know, how everytime, we end up with a PAUL HARRIS paper to come to a conclusion.



Cheers,



Raviraj

And Khuong,



One more thing.....



If there is only one calender assigned for the project, then LONGEST PATH is equal to TF=0



Am i correct?

Hi Raviraj

Thanks to Khuong there is the answer.

There is another good presentation from our collegue Ron Winter which may help:

www.primavera.com/partners/files/Longest_Path.pdf



A GPS navigation system all short pieces of way wouldnt be a great help for you, but the shortest way - just inverse algorithm.



Regards

Dieter

Hi Khuong,



Just one more query. Logic is correct because LONGEST path is derived from activities having early finish equal to latest calculated early finish for the project, but....



is it because of the different calender assignments used.



But, then what u r using - LONGEST PATH or activities having TF=0.



For me, it shud be LONGEST PATH.



Any suggestion guys???

Yup Dieter,



I think its possible. C, its like this



All activities on LONGEST PATH will have TF = 0,

but activities having TF = 0, maynot be on the CRITICAL PATH



Does this sense correctly???



Cheers,



Raviraj

Hi Raviraj



in some cases, it is the same. In bigger projects it’s different but with some overlap.



Dieter

Hi Khuong,



Does, P3/P5 give any different critical path, in the above scenario (between TF=0 and for the LONGEST PATH)



Never tried...



Cheers,

Hi Khuong

I like this option. If a project plan, where many part project leaders are involved, has to be optimized because customer’s wish or the contract are not equal to the results of planning, it is more efficient. With the longest path there is a unique way through all subprojects. To optimize this path is much easier to understand than an activity here and an activity there.

Also, if managers ask for the critical path, its rather the longest path.

It’s just my experience.



Regards



Dieter

Oliver

It works with Level of Effort activities. For the rest, as far as I know, the algorithm of P3e/P5 is equivalent to P3.



Dieter