(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?
(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.
I read many blog entries that are very derogatory about MS Project, but struggle to understand why there seems to be such negativity about this product. It may not be as feature dense as products such Primavera P6 but it is a powerful, effective, planning tool that can be used for monitoring and controlling both large and small projects. I regularly come across criticisms such as:
- It isn't a CPM tool
- You can't track progress or costs
- You can't calculate and report earned value
- The float calculations are inaccurate
- You can't store multiple baselines
- Its grouping and sorting capabilities are poor
- Progress sometimes ends up in the future (beyond the status date) after updates
- Unstarted work remains in the past (behind the status date) after updates
- You can't use it for large complex projects
All of these criticisms are in fact falsehoods and stem from the fact that the users in question do not understand the capabilities of the software, haven't configured it correctly or haven't entered data into it properly in the first place.
In fact it is perfectly possible to create MS Project schedules that accurately determine the critical path and calculate free and total float. All the normal cost control and earned value features that any project would need are catered for and can be reported; in many cases this requirement is irrelevant as cost and earned value calculations are often calculated outside of the scheduling software using Deltek Cobra or similar. Updating progress against tasks can be accurately and efficiently accomplished providing the appropriate calculation settings are selected and proper care is taken when updating the schedule; The key is understanding how the software works and then using it properly. Up to 10 baselines are available, more than enough for most people's needs in my opinion. Projects can be planned with up to 99,999 tasks per file; I don't believe too many people are managing scopes larger than that and schedules become very unwieldy when the file gets that big. If more than 99,999 tasks are needed then multiple files can be linked together. The grouping and sorting activities are now very flexible; using Outline Codes a myriad of different views and summary reports can be created.
I'd be interested in hearing what others think?
Printer-friendly version
Send to friend
Bernard,
Why not start a new discussion thread under your terms, starting with the issues of most interest to you. It does not matter if the same issues it will be easier to follow.
As an advance, on the initial issue presented on the old thread I found that using my software I get results as I expected though they might not be the expected by others. I got the following.
Phase 1 is in agreement with what I expect of CPM computations. Note that the aparent time gap between the finish of Task 2 at the end of the day and the continuation on next day in the activity calendar is zero or non existent on Task 2 work hours. Task 2 ended on 07/02/2012 at 4pm which is for purpose of the calendar equal to 8am next day. Because the link is a FF link you get the same start hour by subtracting the eight hours to 4pm on 07/02/2012 or by subtracting the eight hours to 8am next day. 100% of Task 2 minus an infinitesimal of % whose limit is zero can be performed before end of Task 1 and by this time after an infinitesimal of time Task 3 shall be able to start. This is something that can confuse some schedulers. If you need Task 3 to start after Task 1 then an additional link between Task 1 and Task 3 is needed as shown in Phase 3, even though some believe it would be a redundant link.
In this case it is more about understanding the math than an error; think of it as 8:60AM = 9:00AM .Best regards,
Rafael
I realize this thread is just over a year old, but it's the first time I've seen it. If Donald is still around, I suggest reviewing this even older discussion here on the PP:
I found very strange results with critical path calculations from various software packages without even complicating the issue with resource leveling. It's been a few years, so I don't know if the software I tested at the time still calculates the same, but it may give you some ideas about the potential quirks that can be found in CPM scheduling software.
Hi Rafael
It is high time you asked to be a moderator - at present it is only Gary and I that patrol the site and we welcome some help.
Best regards
Mike Testro
Arend,
I no longer have any access to P6 or MSP even when around here most Contractors have switched to Microsoft Project.
It would be senseless to have a discussion about P6 if I no longer have access to it and this is why lately I limit mostly to ask about how any software does it for the case I must transfer their files into Spider Project, to understand the differences and figure out how to make the Spider model mimic what the other software does to validate the conversion that once validated I usually modify to include expanded functionality of Spider Project.
Here it is considered of bad taste reviving old topics, perhaps because it is of no interest to those who years ago posted something, perhaps because new software upgrades makes the presentation obsolete, particularly on those software notorious for thousands of bugs.
http://www.planningplanet.com/forums/project-management-issues-discussio...
I only disable a thread if it is an old topic that some new member has resurrected or if it is developing into a "me too" feeding frenzy for copies of files.
Quite often it is both.
As you can see at times they are even locked by moderators who do not like this.
I no longer have interest in this topic at this time. The following issues presented long ago by Paul Harris might already been solved, it would be unfair to bring it now, a year latter.
As a matter of fact my perception about MSP has changed a lot, especially after looking at their latest version, I cannot say the same about P6.
Suggest you stop reviving old topics, especially on those I was involved, same as some moderators consider it inappropriate reviving old topics I do.
We do not limit our discussions to a single software, we discuss software and compare functionalities, but avoid personal attacks. Blog entries might be derogatory about MS Project as well as about any other software but there shall be no derogatory comments against any member of this community. I consider your comments offensive and in the future will avoid any conversation with you.
Best regards,
Rafael
Dear Rafael,
maybe your settings for Primavera were not correct in your example for free float.
To use your own words: garbage in is garbage out..
Trying to hit on others work or prefered software will not convince, instead you could give a more positive attitude on your preferred spider project.
I thought this website was meant as an connecting point, a web of planners where you can help each other out.
Not a kind of football field where people try to tackle each other.
Hi Chris
Do the work in Powerprohect and then convert it to MSP.
Best regards
Mike Testro
one thing is, I cannot find planned % complete in MSP 2010;
and also I cannot just drag my mouse to define task relationships;
and what about constraints? I think MSP has some pre-defined constraints by itself (even if you have not defined anything).
My 2p on this feed....
I am not as experienced as some of the planners here, but I was trained is MSP, moved to the UK and finally got into Asta - Horray! I realised the potential of such a useful tool. Purely from an actual all day every day usage point of view, I am juggling many different projects and tenders on a day to day basis and swapping between projects gets very confusing to say the least! Particularly in interior construction the game changes constantly, the way you plan in the tender stage never turns out on site due to client changes, changes on site and most often the PM just wants to do it his way. I have found that Asta is much more friendly and adaptable to making all these changes, is easier to visualise where changes & adjustments can be made, especially when your programme extends over a few thousand lines with multiple areas of logic and SS relationships sqeezing timeframes and sometimes A doesnt always come before B. Tight timeframes, strict working conditions and many other things make for interesting planning and Asta simply made my job that much simpler and less stressful. Not being shallow it just looks better too.
Unfortunately, the company I work for uses MSP from now on in and I have only wanted to throw my laptop out the window twice. But there is not much we can do but understand the software and use it as best as you can. It does the job but anything mildly complicated, it spits the dummy.
Gimme Asta any day of the week
Paul,
For this need, you can use outline code funtion.
Please go to tools - customize - field, then you will have a custom fields dialog box where you can define code for both task and resource.
When your codes are set up, pls add column of what you built-up, then you can use filter funtion for your need.
Actually this code stuff is effective for simple cost control.
Good luck,
Matthew
Donald
I have used MSP for over 12 yrs, I have also used Primavera for 13 yrs, there is no comparision, yes you can use project for basic plans critical path etc once you understand how to manipulate MSP, but when it comes to the useful things like resource, costing, exporting etc it is pointless, I know you can buy the enterprise model to do more powerful resourcing etc, but why should you, and i dont want to go on about the basic principles of resource allocation in %'s that msp goes on about, anyway Primavera is a much stronger tool in every department apart from cost, Even using Suretrak is better than MSP
regards
Brad
(a bored planner)
Brad,
Have you used MS Project before? If so what have you done with it and what were the specific problems that you encountered? In my experience most of the issues stem from a lack of knowledge and training on the product rather than the program itself. Once you understand the product in terms of its capabilities and its limitations it is quite functional for most situations. The bottom line is that many companies, across many industry sectors, successfully manage critical projects using this product. These companies can't all be wrong? They presumably all evaluate the costs and benefits of their software purchases?
Regards,
Donald
Mike,
P6 and MSP both yield garbage, wrong math, wrong float values after resource leveling and at times very weird, cannot model true shift work, partial assignments functionality is very limited, resource leveling is very poor, work is defined using hours and this is inadequate for CPM scheduling especially to account for productivity, have no resource production so you can only model availability of resources fixed by date but not as the schedule move (the way contractors manage their resources), lag calendars are defined on a limited set not individually such as when lag is based on calendar days and successor and predecessor work on regular work week, and the list go on.
Be honest, sorted or unsorted garbage is garbage.
Best regards,
Rafael
lets be honest project is shite,
Regards
Hi Paul
Glad you got it to work.
The same principle applies with decimal points - so if you want to replace something between say 05 and 06 you can use 05.01 etc and it will drop into place.
Best regards
Mike Testro
Mike,
Yes, i was sorting by text. i have now prefaced all the code structure titles with 01, 02 etc etc and the layout is as expected. It's a bit clunky as if i need to make a change i have to manually renumber part of the structure but it's workable.
thanks
Paul
Hi Paul
I am no MSP expert - in fact I never use it if I can avoid it but your problem may be from a simple cause.
Are you sorting on text or numbers?
If you have numbers in text format then you must use leading zeros - 00123 - for the sorting to work correctly.
Best regards
Mike Testro
Donald,
Perhaps you can help me with something: I am trying to use the MSP grouping and sorting function. I want to replicate the functionality in P6.7 where I first group activities by a code and then sort by their start date.
In MSP this doesn't appear to be possible. I have managed to group by a series of outline codes which I have created but they do not appear on the screen in order in which they appear in the outline code structure I created. When I use sort to try to fix this problem the only options, on selecting the codes, are ascending or descending - this doesn't get me the result I need. This is particularly frustrating as one of the outline code structures is "phase" - not much help to have the phases displayed in the wrong order.
Also, when I sort by start date, MSP rearranges all my grouping - also not very helpful.
Any suggestions or is the functionality just not as comprehensive as P6.7?
Thanks
Paul B
Cost of MS Project $680, cost of P6 $3500+. That is why it is used more widely than any other option!
P6 is being picked up more in NZ, especially on Civils contracts and amongst other organisations who use it as a full on Enterprise software solution, but the lack of an established Planning culture in the general NZ Building/Construction market means that for scratching the surface and doing high level programming, MS Project is all the people use (or in some cases even aware of existing!) as it is a Microsoft piece of software, feels a bit like using Excel when you start playing with it and is a relatively inexpensive investment. For me, using it to a greater level of detail than most, it simply doesn't/can't do what I want it to do and the basic functionallity has only got worse since I got "upgraded" to 2010!
When I last worked in the UK (2007) most main contractors were using Asta or P6, smaller subs were generally the only people I dealt with who still used Project, but then the UK construction market is more advanced (and bigger!) than the much smaller NZ market.
Yes I have on many many occasions over the last 11 years, hence the negative comments in my first post. From personal experience I wouldn't trust it as far as I could throw it to accuratley do anything vaguely complicated. I have always had to use Project for certain clients and projects alongside both P6 and Asta, and working on multiple platforms at once shows up the limitations and problems that Project can have as a serious Planning tool.
Alas in NZ it is the pre-eminate planning software used amongst most contractors, project managers and consultants, probably due to it's relative inexpensiveness compared to P6 and similar (also Asta is not supported in NZ at present).
On a basic level it doesn't handle simple logic linking as well with as many options as Asta or P6, such as the ability to link the same 2 activities together in more than one way or identify a type of link (eg Asta you can identiy differnet types of link, such as resource or sequence/logic) or give me as many options for resource management and modelling that I require. I find the basic functionality poor at times, never mind talking about the more complex issues.
If I never had to use it again I would be extremely happy!
Iain,
Have you ever actually used MS Project to create and progress a resourced schedule? Judging from your comments I suspect that you know very little about it.
Donald
I think the general negativity towards Project is to do with it being an anti-intuitive, un-user friendly piece of nasty rubbish that Microsoft keep updating, instead of actually doing a proper full on review of the software, with each update making it even worse! It is also a very very dangerous piece of software, a little knowledge is a very dangerous thing with Project, a lot of users (generally project managers and supervisors) have a rough grasp of how it works which I have seen lead to problems.
Project 2010 is truely horrible, it's a toy, it should not be considered as usefull or accurate for anything more than a short sketch of a schedule or programme.
I'll stick to P6 or Asta Powerproject thank you!
Sometimes it is possible to believe that more detail means more control. That's true to a point, but eventually you can get lost in the small details. You can't manage a project if the schedule is so complex that it is impossible to understand.
I cannot agree more with you. Our schedules are for relatively small jobs, on average about 20 US million, our schedules on average are of about 500 to 700 activities because of specification requirements to include all procurement activities. At times we agree only to include long lead procurement. These jobs duration is about 20 months. If 200 activities are procurement activities then 500 are construction activities divided by 20 months yields about 25 activities per month including al subcontractors work. Because it is common for some activities to span more than a month our monthly updates for construction activities results in about 40 construction activities to be updated. The procurement activities are updated from the Submittal and Procurement log.
At the end of the month it takes the PM about an hour to check the monthly updates and e-mail me the data. After a couple of cycles in less than 2 hours the update is finished along with all schedule reports in PDF.
Sorry but here I have seen no schedule with over 1,000 activities by any scheduler. The issue is not with the activity count but with the poor resource loading functionality of many software packages that cannot even model shift work adequately, that cannot handle partial assignments the way the industry share resources among different activities. The issue is on how they handle resource details as many do not provide functionality for multi-resources making it very difficult to load and see the details of your crews as both views are needed.
By tracking crews you can see the macro, by tracking crew details you can see the micro, both are relevant. Peaks in resource demand are predicted by crew usage but if suddenly you need a couple of extra crews you need the composition details, you need to know what you have now and what you need in order to re-distribute personnel and ask the union for the trades you need. Here the schedule will tell you about the resource details available form other crews, and will tell you a lot about the projected demand and availability.
Yes in complex projects you need the details of actual and forecasted resource demand and availability. Using the appropriate tool makes it easier for you to solve the issues, the computer alone will not do it for you but it is of much help if the tool is good, otherwise it will make not much of a difference and the PM as well as field personnel will not look at the CPM but will manage the job the old fashion way, using their own improvised Gantt Charts.
Unless the job is delayed by external factors, the delays by contractor are mostly due to poor handling of resources, very rarely the logic is wrong as construction jobs are no rocket science. If for some reason the construction sequence must deviate from the standard practice then the construction documents must be issued with specific instructions.
Best regards,
Rafael
Donald,
I do not expect that the Spider schedule is always optimal but am sure that it is at least close to optimal. It is just impossible to verify it manually in any case. Spider schedules are always shorter than the schedules created for the same projects by other packages and for smaller projects it is clear that Spider finds the best solutions.
What do you mean by limiting resources? The schedule model shall reflect the reality and is used for scheduling, budgeting, resource management. We suggest to limit project materials including those that have low cost and are always available in Other category. But we do not exclude any renewable resource. Don't forget that the schedule is also used for budgeting and resource planning.
Constrained resources are usually machines. Leveling manpower is not enough. Adjusting logic means that you add artificial constraints that actually do not exist. It means that you level manpower manually. When something will change you shall reconsider all these artificial constraints. Don't forget to apply the same process to all other resources, that also may be overloaded. You will not be able to verify that your schedule is good and what if analysis is too hard. I don't think that it helps to advise the project manager of the potential logistical and resource constraint risks that lie ahead. Using automated leveling you can play many what if scenarios and find better options. And I don't believe that you repeat this process removing artificial links and adding new ones each time when enter some changes.
Best Regards,
Vladimir
Rafael,
In general I use a resources breakdown that might look something like: Supervision & Management, QA\QC, Structural, Piping Fab, Piping Install, Piping Test, E&I, Mechanical, Scaffolding, Painting, Piping & Mech Commissioning, E&I Commissioning etc.
Using an appropriate WBS combined with Activity Codes allows me to sort and analyse progress by Phase, Area, System, Sub System (part product) and Resource\Discipline. That way I can quickly identify and rectify problem areas.
I think that it is important to get the right balance between a meaningful schedule (adequate granularity) with something that is manageable and easy to analyze. I've seen project schedules that were massive, literally hundreds of pages of A3 that were incredibly time consuming to maintain and update and very difficult to analyse. In a very large project handling such a large volume of information might be inevitable, but this should be managed by dividing the project into smaller stages. Sometimes it is possible to believe that more detail means more control. That's true to a point, but eventually you can get lost in the small details. You can't manage a project if the schedule is so complex that it is impossible to understand.
Donald
Donald,
Applying dozens of resource descriptions across thousands of activities adds complexity that is simply not necessary nor beneficial as it does not improve control over the project.
To me it is kind of primitive when it is enough to create your crews only once, with their appropriate composition using the required combination of individual resources, skills, individual resources productivity, partial workload assignments and the like. For this we use the concept of multi-resource. Yes it is insane, it is is sick, it is primitive and even very stupid. What do you use?
How do you schedule for a single crew that on the same day performs work at several activities? And what about a single resource working the same day on several activities? Do you cut into pieces the resource, where does the head goes, the hands and the behind? I guess all parts must follow their respective behind.
Best regards,
Rafael
Vladimir,
The best approach when developing any large construction schedule is to try to limit the resource descriptions to relatively few disciplines, trades or even crews\teams in order to control the administrative effort necessary to maintain and update the schedule. Applying dozens of resource descriptions across thousands of activities adds complexity that is simply not necessary nor beneficial as it does not improve control over the project. Given that, producing a leveled workable schedule is simply accomplished by monitoring manpower histograms to identify resource overloads as you are building the schedule and making adjustments to durations and logic to resolve them. I acknowledge that this is quite time consuming but it ensures that the scheduler has a much clearer picture of all of the resource constraints in the project and is far better equipped to advise the project manager of the potential logistical and resource constraint risks that lie ahead. Simply pressing the leveling button might be quicker but certainly won't allow such a strong appreciation of the program pinch points to be attained.
I'd like to pose a question to you now. How do you know that the resource leveling algorithm in Spider always produces an optimal and reliable result in a construction schedule with thousands of activities? What have you done to verify that?
Donald
Donald,
I suppose that you have a good knowledge of planning and understand MS Project.
Then please explain how to schedule large projects with limited resources applying manual levelling. It is easy to find a solution in the schedule consisting of 5 activities. But most construction schedules consist of thousands activities and use at least several dozens resources. Do you use some special technique?
Best Regards,
Vladimir
Abdullah,
if you assigned higher priority (manually) to activity 3 it will be scheduled first.
If you set activity order where activity 3 has lowest row number then using "ID only" leveling option it will be scheduled first.
All these options are manual. I don't know MSP automatic settings that produce good schedule.
Best Regards,
Vladimir
Donald,
But you are still avoiding the question about shift modeling, just prove MSP can handle the issue in a practical manner. Prove you can get good shift modeling results with MSP, prove MSP can handle such small job, a single activity job.
It is so simple you shall be able to do it with MSP with your eyes closed, just enlighten us.
http://www.youtube.com/watch?v=1_qqDYbdq3w
By the way Donald, welcome to PP you will be surprised how good for your understanding of MSP this site is, even those of use who make occasional use of MSP will try to make positive contributions to other MSP users. Still we believe disclosing the tricks, traps and limits of all software is good for the community. Here you will find people who know hot to take out the most of MSP. Consider the debate as the fun part.
Best regards,
Rafael
Abdullah,
You are of course correct. Once the software and calculation methods are understood and the scheduler has a good knowledge of planning then good quailty results can be created.
Regards,
Donald
Rafael,
You have provided a lot of input to this thread that has created considerable debate and that is good. However, I don't think you have demonstrated that MS Project is an unreliable and unproductive scheduling tool. Most of the arguments that have been put up are pretty unconvincing and reinforce my opinion that many schedulers simply don't understand MS Project's capabilities. Poor results will always be achieved when tools are used incorrectly by people that don't understand them? As I stated in my original post MS Project does not have the same feature set as some of the other software available. Clearly Spider has some advantages when it comes to resource levelling.
However, at the end of the day, in competent hands, used the correct way, MS Project is a very power scheduling tool that is capable of managing all sorts of projects large and small.
Abdulah,
From: http://www.tech-archive.net/Archive/Project/microsoft.public.project/200...
"To the best of my knowledge, Microsoft has never published the exact algorithm that the software uses to level over allocated resources. Essentially, the software uses a combination of five factors in the Standard order to determine which tasks to split or delay: Predecessor, Amount of Total Slack, Start Date, Priority, and Constraints. After 7 years with this software, I'm still not sure which of these five factors is considered first when using the Standard order, and like you, I can find counter examples. If you set the leveling order to Priority, Standard then the Priority number factor is considered first. If none of the factors determine which tasks to move or delay, then the software uses the ID number to make its determination."
Look for MSP Leveling options. 
If you compare it to Spider Project Options you will still see more options. Some options like standard are similar to the options in other software, options a bit less "un-advanced" than MSP options. These even let you choose the field values that will fix the priorities. 
As I said before, instead of manuall looking for the combination that will yield shortest job duration I use the optimization option, an option that will look for the best prioritization automatically, therefore under Optimization the Priority box will be dimmed. Note that in Spider even under Optimization you have some control of priorities, either by individual activities, by WBS or by both.
No matter what prioritization you use different software/algoritms might yield different results. There is no way under complex jobs to make sure the algorithm yields optimum results, all you can expect is it will yield near optimum results. There are methods to study and compare the results, one of them is by running the algorithm in a set of jobs for which optimum is known and compare the results. It is the scientific way, it has been performed by some European University but to my knowledge the available results are limited to a few software and to older versions.
Seems for the moment they are lagging, the best we can do if our software is not reported yet or if our software version has been upgraded is to compare results ourselves or with the help of others. Because I have a very limited access to a P6 license in order to compare Spider to P6 is with the help of others in Planning Planet I can make such comparison.
Best Regards,
Rafael
Vladimir,
This looks intersting, all i did is moved the task 3 up the order before i assigned any predesessor or levelled the schedule. The result is shown below, it is the 21 day duration you were looking.
Both resources A & B have been levelled too. Do u think this because MS Project automatically schedules the top most activity at an earlier date because bot activities 1 abd 3 have the same predesessor 2 and have no priority preference.
Could it be because all the activities are scheduled "As soon as possible" and hence the top most activity get the first come first serve treatment.
I get your point here, but then what you stated is defied by the below result. It means the software has its algorith right somewhere may be we have to understand it a bit further.Sorry for the poor image quality but i could email it you if you so wish.
Regards,
Abdullah
Donald,
You are a 100% right. icould not understand the login of the activites in the Spider Project. The MSP shows a FS between the activities allocated with resource A and then the leveling changes the logic.
I tried a similar 4 activity plan with same prioritization and a single resource allocated to every activity with resouce and activity on the same calender and then i leveled the plan and i didnt find anything wrong with the result. I am dead sure that the same algorith goes for even the more complex leveling iterations too.
But there is one strange doubt that i have. It is a little offthe main topic, when i go to select the field "Planned Cost"
it does not show up in the list of pre defined field?
Regards,
Abdullah.
Se de Leon,
Donald said ' It may not be as feature dense as products such Primavera P6 but it is a powerful, effective, planning tool that can be used for monitoring and controlling both large and small projects. ", therefore your statement about "Posters should argue against your original post instead of introducing unrelated issue to your original post (you never mentioned resource loaded CPM in your original post)" is kind of out of order, speaks a lot about your character and suggest some interest on your part in hiding the reasons why many of us consider MSP dysfunctional, but this is the central question.
I do not see any request by Donald for us to interpret his question in a limited way, I believe it is an honest request to hear any and all reasons others do have to make derogatory comments about MSP. For some it is high opinion for others it is low. Mine is clear, I have like many in PP a low opinion about MSP and believe it is the idea of the debate for all of us to express our opinion.
In order to control large and small projects the need to deal with resource leveling is implicit, but because as you say anyone can get the same results I would appreciate everyone let Donald tell us what prioritization is he to use as for get the same 21 days results as I tried with whatever was the default in MS Project and got about 50% more duration with MSP default priority, of course using activity specific prioritization after knowing the solution it would be dishonest, would be cheating, same as creating a "soft link" to force resource leveling. I do not even know where in MSP you set the prioritization, is something I do not need in Spider Project as it looks for optimal or near optimal resource leveling under the "optimization" setting. Less than optimal are available for comparison purposes and for other strategies when you do not want to disclose optimal or near optima solution to others.
This is an extremely simple schedule David shall be able to figure out about prioritization rules alone by himself, even if by trial and error, then as I mentioned we can latter try more complex problem an you will be invited to figure out on the spot a better solution with MSP but this time it will be you who shall disclose first what is the best solution you can get with MSP.
I would like to see your solution to the single activity shift problem I presented before, is just a single activity you shall be abler to model in MSP with your eyes closed, for you shall be very easy, just enlighten us about the capabilities of MSP.
David asks for others who do not like MSP to express his opinion, he is not asking those who believe to be good to express why they believe it to be so good, in any case these are intended to be left out. But no, I believe all issues shall be debated by anyone interested and believe David is interested in all point of views.
Best regards,
Rafael
Donald,
Your absolutely spot on in your first post. Posters should argue against your original post instead of introducing unrelated issue to your original post (you never mentioned resource loaded CPM in your original post)
Nonetheless, I tried resource levelling of msp using the example and it yields the same result as Spider. As you may well know, in MSP, if you want to try different levelling scenarios, one way to do it is by changing the priority numbers. Who cares if other softwares are doing it in a different way, as long as you get the same results.
Cheers
Donald,
I expect that the software shall level the project, not the scheduler.
You suggested to level resources manually introducing links that do not exist.
So you are happy with MSP schedule and critical activity with 9 days free float and other activities with floats that cannot be used without delaying project finish. I am not sure that other schedulers will agree with you.
In my example the logic is correct and good PM software shall calculate resource critical path without introducing artificial restrictions. You shall learn other tools before talking about "any" project management software.
Best Regards,
Vladimir
Donald,
In Vladimir sample job there is no link between activities 2 and 4 (ID codes 4 and 6 respectively), the resource leveling algorithm creates the resource dependencies, but MSP resource leveling algorithm is poor.
Before resource leveling resources A and B are over allocated. After resource leveling MSP delays some activities until there are enough resources.
The following screen views show before and after resource leveling results by MSP and Spider Projects.
MSP before leveling: 
MSP after leveling, using default prioritization rules: 
Spider before leveling: 
Spider after leveling: 
As you can see Spider resource leveling yields better results after resource leveling without the need to perform multiple runs using several combinations of prioritization in the hope you will get a better result. Elementary resource leveling in this case but not so elementary when hundreds of activities and combining skills, multi-resources, partial assignments and shift work.
Some MSP enthusiast before said that if you know MSP it should be a piece of cake to figure out the combination of prioritization that will yield optimum or near optimum resource allocation, guess you will be able to tell me on the spot what prioritization settings are required for MSP yield 21 days after resource leveling of such job. Thereafter we can talk about other resource leveling jobs, a bit more difficult but still no big deal.
Best regards,
Rafael
Vladimir,
I'm not really sure what your example proves? You have a leveling delay (lag) introduced following leveling between activities 3 & 4. Given the scheduling logic the critical path is correct. The only way any scheduling software can create an accurate resource constrained critical path is if the logic is correct. In your example that can only be accomplished by introducing FS logic between activities 2 and 4. All of the comments I've seen so far reinforce my first statement that the fault lies with the schedulers not the software.
Donald
Donald,
Now we all know MSP cannot display correct values of float after resource leveling, lets try with another issue.
The following is a sample job of a single activity as I could not find a suitable sample of less activities. Try it with MSP and let us know, is a shif model as simple as it can be.
Activity 1
500cm rock excavation
Resource 1 production 10cm/hr and works Monday, Tuesday and Wednesday 10hrs/day
Resource 2 production 15cm/hour and works on Wednesday, Thursday and Friday 10 hrs/day
If activity starts on Monday:
Monday =>> Resource 1 will produce 100 cm
Tuesday =>> Resource 1 will produce 100 cm
Wednesday =>> Resource 1 will produce 100 cm
Wednesday =>> Resource 2 will produce 150 cm
Thursday =>> Resource 2 will produce 50 cm in about 3 hours
Activity will take 3 days 3 hours.
If activity starts on Wednesday :
Wednesday =>> Resource 1 will produce 100 cm
Wednesday =>> Resource 2 will produce 150 cm
Thursday =>> Resource 2 will produce 150 cm
Friday =>> Resource 2 will produce 100 cm in about 7 hours
Activity will take 2 days 7 hours.
If your software is not capable of modeling the above, simple shift work on a single activity, then you are using the wrong tool. Every time the activity is delayed, the distribution of work is shifted, when you have many such activities and work on different hour shifts, different days it can become quite complicated. Distributing work by hand is nuts, would not be true modeling of shifts.
Latter on we can discuss resource leveling using Vladimir sample MSP job where resource leveling by MSP yields longer duration than using other software.
Yes the list is big very BIG.
Regards,
Rafael
Of course even a kindergarten toddler knows 1+1=2 no much brain needed, same for ancient CPM with no resource loading.
If MSP do not want to combine resource loading then it shall not provide functionality for it. But it does, so it should be implemented correctly with mathematically correct float values, not as erroneous as it does.
I would have no issues with MSP if it does not provide functionality for resource loading, then it will not be misleading people that resource load their jobs.
Is a matter of mathematical correctness, is a matter of Bill Gates showing he understand true float values or at least to require his people to do the right thing. Aim for the truth, do not aim for mediocrity. It was ok for the first years to MS Project be so wrong, but after so many years there is simply no excuse, there is no depth of thinking behind it, nor even honesty about the software limitations.
Best regards,
Rafael
Let's be clear about this. Please don't combine the topic of task only CPM with resource loaded CPM. Results will never be the same. I tried the samples above without resources, MSP is showing correct results.
Abdullah,
Your comments are welcomed, this is a place to debate oposing views. The fact that you have no need to resource load your schedule avoid many errors within MSP. The relevant issue is that we all know the limitations of MSP and P6 with regard to disclosing true float values after resource leveling. I guess if you have no need for serious resource leveling MSP and P6 can be an alternative.
You are correct, resource leveling exponentially complicates things. Unfortunately for many of us in the construction industry we cannot schedule our work without consideration on limitations of resources. This is precisely why I believe MSP and P6 are no good for the Construction Industry, perhaps for the IT industry where software and IT people are cheap and available without limits. Where they do not produce any physical work so they do not understand the concept for work to be defined by some physical amount but by hours, even if no single line of code is done.
Best regards,
Rafael
Hi Donald,
I could not agree more with you on this. I beleive that at times Planners get overpossesed about creating a resource loaded schedule even if the situation does not actually require one. Just to prove their planning skills, they miss out on defining the calculation methods, and instead spend more time trying to make the schedule as complex as possible.
Somewhere down the road they realize the mess they are in, and then they come up writing these derogatory remarks about a simple to use efficient scheduling software.
I was previously using and P6, and have been using MS Project 2007 with my new company for the past 9 months now. I feel once you clearly define and understand how the calculations are to be set with your schedules, you should find that MS Project 2007 is not bad at all.
Regards,
Abdullah.
Vladimir,
Thanks for the clarification, in such case I will re-define again P6 and not MSP to be the Mother of all Garbage.
Just look at the following simplification of my sample job, this time without open ends as to satisfy the urge of some to not leave open ends. MSP still yields wrong float values.
Yes it is bad, but P6 I believe it to be worst, with more than 30 MGB of code they shall be able to deliver something better, and they are way over 40 MGB when you add the external database engine, a huge database system not appropriate for the purpose, perhaps good for the IRS keeping track of all individuals in the USA, corporations, societies and the like that are to pay their taxes, but for a few CPM schedules say 100 each with 100k activities. Still an overkill, especially when such overkill does not prevent P6 to be in error with regard to some float values, not to mention the incapacity to model shift work when resource productivity drive the duration of activities. To me this is even more insane than the sample job MS Project could not solve.
Perhaps after discussing the issues on how MSP handles work and productivity (if it can handle productivity in a meaningful way) we shall discuss the issues on shift modeling by MSP.
Best Regards,
Rafael
By the way I believe handling of huge Portfolios using the level of detail generated from original job schedules can become impractical, therefore I am still waiting to see the new concept in the working by Spider Project to generate Summary Schedules at a higher WBS level but keeping some relationships between WBS activities. This is what I always believed to be appropriate, although until now it requires some manual scheduling, automatic generation of such summary schedules will be great, especially when double dependencies will be generated, to my knowledge a feature not avilable in any other software. Yes for these purposes less will be more, otherwise you might go to sea and see no water.
Rafael,
the schedule that is shown in my post uses only two resources A and B (one unit each).
I do not use MSP and did not test its capabilities in schedules that are not resource loaded. But I expect that without resources everything will be fine because CPM is easy to model. I don't like MSP approach to scheduling of ALAP activities (using total float except free float) but this is MSP feature, not the error.
Best Regards,
Vladimir
Vladimir,
It is hard to believe it can be so grossly wrong.
Is your example resource loaded?
Is it possible for MS project even without resource constainig to yield wrong float values?
If this is so then MS Project should be called the mother of all garbage.
Best regards,
Rafael
Donald,
How bad MS Project and Primavera P6 are is not limited to the fact they are incapable of consistently produce correct values of float. Is also because of many other wrong concepts. for example their concept of work.
Of course the list is quite big, is HUGE.
Best regard,
Rafael
Donald,
you did not like Rafael example because activities were not linked. In real projects it is not rare that tasks belonging to different paths compete for the same resource.
But look at another example where all activities are linked:
The Leveled schedule is 30 days long (21 is enough),
Activity 3 is critical and has 0 free float that is wrong (9 days delay is available),
Activities 1 and 2 are actually critical in this schedule (their delay delays project finish date) but MS Project shows that they have 10 days total floats,
Free float of activity 2 is unusable.
The schedule of this example is the same in MSP 2003, 2007, 2010.
Rafael,
For MS Project , we may think it in this way,
- In Project1, we may split Task A into 3/4,3/7,3/8,3/9 and 3/10; Task B as 3/11,3/14,3/15,3/16,3/17, and so on for Task C and Task D, in mind.
- Task A float is 15 days in Project1 as its last piece of work 3/10 can delay 15 days without delaying piece 3/31 completion (last finish).
- In Project2, we delay piece 3/10 completion to March 25 (last finish) for a test. We can see Task A floats come down from 15 days to 4 days.
- Like building block, we can arrange 25 pieces of resource driven tasks here according to the resource schedule in anyway, and free floats of Task A,B,C,D always determined by its last finish
Anyhow, this is for pure resource driven only. For construction task, we had to logic link activities as process sequences do exist. Early/late dates are driven by resource, logic and duration.
Your statement, "In fact it is perfectly possible to create MS Project schedules that accurately determine the critical path and calculate free and total float", is in error, perhaps before resource leveling it can yield correct values of float but after resource leveling it is in such error you never know.
Just create a 4 activities schedule with a duration of 5 days each, all loaded with a single resource A, of which there is only one available. Do not link the activities.
After resource leveling it will display wrong float values. Try the same with P6 it will show wrong free float values. Therefore I consider both software pure garbage. How in the world you pretend to call your software true CPM software if the basic computations for float are wrong?
Take the challenge and prepare yourself the sample job as for you to convince there are no hidden trick to make either MS Project or Primavera P6 to look as rubbish. Comparing MS Project to P6 is comparing garbage to garbage.
If you know the very basics about total and free float you shall know on the spot that the total and free float values are all 0. If still you do not believe it, if still too much then try moving a single day any activity and see what happens to the project completion date and to the start of successor activity.
Only Spider Project is displaying 100% correct values of TF and FF for all four activities, 0.0000000000000000000000 In this example P6 got correct values of total float while it misses correct FF values in 3 activities. I know of other sample jobs where P6 misses both Total and Free Float values on some activities after resource leveling, so do not assume their BUG is only with Free Float computations. Someone before suggested that in order to get true float values in P6 you shall uncheck Preserve scheduled early and late dates but the naive did not do his homework as disclosed in the provided sample schedule where the suggestion was applied, still wrong float values.
It is unbelievable the most basic computations for float are wrong in most software and there is a whole industry that does not realize it or do not care or perhaps there are many self proclaimed experts out there unable of realizing it, not to mention to even be honest and recognize it for the good of the hole industry.
I honestly believe if you are a true professional and know the software is in error you have the moral obligation of disclosing the error as wrong values of float can result in misled use of the software with potential negative time and cost implications.
Thorough the years I have become skeptical about software developers and about the honesty of the so called "experts" some call "gurus" perhaps in reference to "prima dona".
Best regards,
Rafael
Hi Donald,
I don't know if the latest version of MSP run from the Data Date when scheduling? Or is there a "Data Date"?
Schedulers usually define logic first and run the schedule.
I also remember that MSP has only "predecessor" and you cannot define "successor" task. Is it 'til now?
cheers!
Email: PPadmin@planningplanet.com
Replies