PR 243658 True Position Basic Dimensions (nominals) not reflected in report

huangyhg

True Position Basic Dimensions (nominals) not reflected in report
<< Doug Schueneman  --  10/02/06  12:05:52>>
<<END>>
<< Doug Schueneman  --  10/02/06  12:05:37>>
I built a Dimension True Position and entered nominals (basic dimensions) on the advanced tab of the dialog.
In the status window, or the report the nominals are different than what I entered.
Steps to reproduce:
See attached .prg file FCFLOC2
<<END>>


<< Changes made by Doug Schueneman -- 12/30/08  16:33:20>>
Status:  RESOLVED to CLOSED
<<END>>
<< David Petrizze  --  11/14/06  10:09:45>>
Reviewed.
<<END>>
<< Changes made by David Petrizze -- 11/14/06  10:09:50>>
Action:  David Petrizze to Doug Schueneman, Status:  REVIEW to RESOLVED
<<END>>
<< Don Turcotte  --  11/07/06  12:00:23>>
The remaining three issues have been resolved.
Files inserted to server
------------------------
V42\DIMENS\FEATCTRLFRM.CPP
V42\INCLUDE\FEATCTRLFRM.H
V42\MENU\FEATCTRLFRMDLG.CPP
V42R\DIMENS\FEATCTRLFRM.CPP
V42R\INCLUDE\FEATCTRLFRM.H
V42R\MENU\FEATCTRLFRMDLG.CPP
V43B\DIMENS\FEATCTRLFRM.CPP
V43B\INCLUDE\FEATCTRLFRM.H
V43B\MENU\FEATCTRLFRMDLG.CPP

<<END>>
<< Changes made by Don Turcotte -- 11/07/06  12:00:29>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  11/06/06  17:36:45>>
I have fixed the problem reported here by Doug.  Additional changes made last week had broken the setting of nominals.  I found three additional problems, though.
1.  The fix for PR243702 is resulting in the deviation in the report of my test program being incorrect when the program is first loaded.  I have to F9 on the FCF and press ok to get the correct report.
2.  There is a problem when first loading Doug's test program which I have traced to the datum constraining when the datums are PLN, PLN, PLN.  I have duplicated this problem with my test program.  The issue is the probe comp'ing that is done by Best_pla.cpp.  This depends on the active tip.  When the program is first loaded, the FCF must convert the axes values to cad absolute (since the cad absolute values were not serialized - only the values in the reporting coord sys were seriailized).  Since the internal tp command does not save its datum reference frame, we must call do_math() on the tp command to re-create it.  This results in the datum constraining being called at a time when there is no active tip, so the probe comp'ing on the dependent datum planes is not done.  As a result the report shows incorrect deviations (off a probe radius) until you F9 on the FCF.  I am looking at a way to delay this conversion of axes nominals to cad absolute until there is an active tip.
3.  If you F9 on Doug's FCF, the feature diameter on the FCF tab of the dialog shows the feature nominal and not the dimension nominal.  If you go to the advanced tab, the dimension nominal diameter is correct.  Returning to the FCF tab of the dialog now shows the dimension nominal diameter.  This appears to be a display update problem only, and not a problem in the math since the dimension nominal diameter is actually set.
<<END>>
<< Changes made by Bill Wilcox -- 11/06/06  10:17:48>>
Action:  David Petrizze to Don Turcotte, Assigned:  Tim Wernicke to Don Turcotte
<<END>>
<< Doug Schueneman  --  11/03/06  16:31:22>>
I have attached a program called GD&TProblem.prg
Open the GD&T FCFLOC1
Go to the second page - Advanced
Click on CIR1
Enter -80 for X Nominal
Enter 154 for Y Nominal
OK and close the window.
Look at the FCFLOC1 in the status window or in the report.
The nominal X is not -80 and the nominal Y is not 154
<<END>>
<< Changes made by Doug Schueneman -- 11/03/06  16:31:25>>
Assigned:  Don Turcotte to Tim Wernicke, Status:  REVIEW to OPEN
<<END>>
<< Don Turcotte  --  10/25/06  14:01:01>>
Fixed and uploaded to V42, V42R, V43B.
Files inserted to server
------------------------
V42\DIMENS\DATUMREF.CPP
V42\DIMENS\FEATCTRLFRM.CPP
V42\DIMENS\TP_START.CPP
V42\EDITORS\EDITWIN.CPP
V42\INCLUDE\FEATCTRLFRM.H
V42\INCLUDE\FEATCTRLFRMDLG.H
V42\MENU\FEATCTRLFRMDLG.CPP
V42\MENU\FEATCTRLFRMOUTPUTDLG.CPP
V42R\DIMENS\DATUMREF.CPP
V42R\DIMENS\FEATCTRLFRM.CPP
V42R\DIMENS\TP_START.CPP
V42R\EDITORS\EDITWIN.CPP
V42R\INCLUDE\FEATCTRLFRM.H
V42R\INCLUDE\FEATCTRLFRMDLG.H
V42R\MENU\FEATCTRLFRMDLG.CPP
V42R\MENU\FEATCTRLFRMOUTPUTDLG.CPP
V43B\DIMENS\DATUMREF.CPP
V43B\DIMENS\FEATCTRLFRM.CPP
V43B\DIMENS\TP_START.CPP
V43B\EDITORS\EDITWIN.CPP
V43B\INCLUDE\FEATCTRLFRM.H
V43B\INCLUDE\FEATCTRLFRMDLG.H
V43B\MENU\FEATCTRLFRMDLG.CPP
V43B\MENU\FEATCTRLFRMOUTPUTDLG.CPP
<<END>>
<< Changes made by Don Turcotte -- 10/25/06  14:03:48>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  10/24/06  17:47:36>>
I have fixed the problem with text mode not showing dimension nominals.  I also fixed a problem with changing DF axis nominals which caused other dimension nominals to be lost.  The problem with compound datums existed in V42R before any of my changes so I have submitted a separate PR regarding this issue (PR244380).  Everything else seems to be working.  I will complete testing tomorrow and upload my changes.
<<END>>
<< Don Turcotte  --  10/23/06  17:31:04>>
I've fixed the problems noted previously, but I have found two other issues :
  1) the text mode report doesn't show the dimension nominals (it's showing feature nominals).
  2) Creating a true position with a compound datum has bizarre behavior. Report looks ok at first. Then F9 the FCF and change the output alignment, then press ok.  Refresh the report and the report for this FCF is missing.  F9 the FCF again and go to the advanced tab and the true position group box is greyed out.  Can't select output alignment or even see axes values.  This occurs only with a compound datum.  The FCF internal commands list has only the constructed set command and nothing else at this point (the tp commands have been lost).
<<END>>
<< Don Turcotte  --  10/20/06  17:34:24>>
I've added all the necessary code so that the math is performed correctly and the report is generated correctly when the user changes dimension nominals without "updating the feature".  I am still testing. Generally everything works correctly, but I have found a case where adding a feature to a program causes the editwin to call update_internal_commands(...) on an existing feature control frame which causes the datum bonus to be calculated incorrectly (not based on the dimension nominal diameter for the datum feature).  I am still looking at this problem...
<<END>>
<< Don Turcotte  --  10/19/06  17:41:05>>
I have this working now so that when you switch between "Datum Reference Alignment" and "Current Alignment" the axis nominals update correctly and are also correct in the report.  I am now working on changes to use the dimension nominals correctly when the true position is evaluated.  Currently it's possible if the user has changed the dimension nominals but not checked "update feature nominals" that the feature nominals will be used in the math and not the dimension nominals.
<<END>>
<< Changes made by Tim Wernicke -- 10/18/06  10:24:26>>
Priority:  Critical to Stop Rel.
<<END>>
<< Don Turcotte  --  10/05/06  17:09:37>>
The issue with the size tolerance getting lost and becoming +infinity has been fixed by PR243666.  I am continuing to look at the other issues here related to how we handle dimension nominals.
<<END>>
<< Don Turcotte  --  10/03/06  17:43:14>>
I am able to duplicate this problem.  This issue is related to how the axes values for nominals on the Advanced tab are applied.  In particular, if the user has not checked "Update feature", then the nominals in the axes list can be lost, especially when changing from "Drf alignment" to "current alignment".  In addition, the size tolerance on the considered feature can be lost and become a huge value (+infinity) when switching the Alignment type on the Advanced page.  This is because the internal FCF flag m_bLine1Format is never set properly.  In Debug mode this will default to non-zero, but in Release mode this will default to 0, so the size tols can then be lost.
<<END>>
<< Changes made by Tim Wernicke -- 10/02/06  14:30:09>>
Category:  Report Labels to GD&T, Action:  Tim Wernicke to Don Turcotte, Assigned:   to Don Turcotte, Priority:  to Critical
<<END>>