Comments on Current CPM Scheduling Practices
In an article in Engineering News Record (Critics Can't Find the Logic in Many of Today's CPM Schedules - ENR May 26, 2003) the CPM scheduling techniques used at that time were criticized. These techniques are still used today.  The main focus was on:
Multiple calendars

Project Services Company's position is that each of these topics can become a problem in the hands of operators who are not familiar with the consequences or who are intent on distorting or abusing the scheduling process. As has always been the case, there are both good and bad schedules out there. Project Services has no hard data on trends.  However, all of the features that have been criticized offer benefits in specific applications, but should be used wisely and, for the most part, sparingly.

The primary function of the CPM schedule is to assist in the planning and the successful execution of the project by, as accurately as possible, forecasting and prioritizing future work.  It is important to not loose sight of this primary function when the schedule becomes a tool for or against claims for more money or more time. The simpler (in the sense of less convoluted) the schedule the better it will serve both this primary function and also will more clearly demonstrate "who did what to whom" in any claim process.

Multiple calendars: Such as a 5 day work week for normal work and a 7 day work week for currently critical activities.  The software calculates the float differently for activities assigned a different calendar, hence activities in the same path can have different floats. This obviously is confusing and the use of multiple calendars should be discouraged. One possible exception could be in a situation where more than one trade is working weekends and they need to dovetail and to see the Saturday or Sunday start or finish dates on their detailed schedules.  A practical alternative in most cases can be to annotate the activity description with the shift requirements and just use a single calendar, and adjust the remaining duration if necessary.                                                 
   
Excessive leads and lags: This occurs when many activities overlap and are interrelated by their starts or by their finishes or by both. Although this happens in reality during project execution much more than in the planning phase, an excessive number of leads or lags in a CPM schedule masks the true critical path. Moreover the lag or lead duration tends to be assigned arbitrarily and in reality will most likely vary with the physical location of the lead activity.  Moreover continuity of work will, in practice, have an influence on the actual start of the lag activity.  The advantage of leads and lags is that they avoid the otherwise meaningless splitting of a continuous activity and help reduce resulting clutter in the schedule.  Project Services Company uses them judiciously.  
Assigned constraints:  This consists of placing early start or finish constraints at intermediate points within a project. These essentially destroy the integrity of the float calculation and should be avoided wherever possible.  However, some agencies mandate their use.  About the only occasions when these constraints can serve a useful purpose is when completion of the constrained activity has a major economic impact on the overall completed project and the remainder does not have same vital importance (one building is leased and the others are spec-built).  Another case would be where work becomes less hazardous after a certain fixed date (the date school is out) or vice versa. The effect of constrained activities is less confusing if the constrained activity is off the critical path. Project Services attempts to limit the use of assigned constraints.

Retained logic: When out-of-sequence progress occurs and retained logic is used, the calculated remaining durations maybe unreasonable. The ideal remedy, from a scheduling standpoint, is to modify the logic in accordance with the actual sequence that the activities were executed in. The main reason for not doing this is economic, the cost of the scheduler, as reworking the logic is time consuming. However with experience, it is fairly easy to spot the out-of-sequence activities on the critical path and on the near-critical paths. There is a fair chance that if the affected path is well off the critical path, as demonstrated by large float, the path will not be a problem in the future and can harmlessly remain without correction. The downside of making logic changes is that it complicates the development of delay claims and can be hard to explain, let alone convince, a jury or panel when the opposing side is screaming that the changes were purely self-serving.

Project Services attempts to have the initial schedule mimic, as close as possible, the job as planned and to select the activities based on the trade and the physical item of work. This helps to minimize the occurrences of out-of-sequence progress.  Project Services then uses retained logic when scheduling updates, looking out for clearly anomalous results and correcting them.

Another aspect that received criticism in the article was the current use of bar chart format for presenting the schedules. This was on the grounds that it does not exhibit the essential elements of CPM scheduling. As both the float bars and relationship connectors can be toggled on or off depending on the intended audience, Project Services uses the bar chart format in the belief that it is understandable to the widest audience and still can show the essential elements of CPM scheduling when required.
Excessive leads and lags

Assigned constraints

Retained logic