Providing support for all E. Crane Computing products, including POWERGEN, PBSERVICE, HARPB, VERSIONEDIT, PBLRESCUE, and PBLEXPLORER.
You are not logged in.
Pages: 1
Hi Nagi,
I'll look at this and get back to you later today or tomorrow at the latest.
Phil
Hi Nagi,
I'm still looking at this. I understand what you're trying to accomplish and determining the best way to get you the information.
I'll be back to you tomorrow.
Regards,
Phil
Nagi,
Here is a proposed solution that would be fairly "safe" to implement, if it suits your purpose.
In the initial regeneration, whenever an object is regenerated because one of its ancestors was modified, the output log includes a line like "Descendant of modified object 'object_name'. This would have to be an option because other users may be parsing the log based on the existing format.
Let me know if this will meet your requirement.
Regards,
Phil
Phil
Our parsing routine is coded well to take care of any changes to the Powergen log file schema.
Please let us know how we can make the "Descendant of modified object" section appear in the log f ile.
We need the section to appear in the log and we need the descendent regenrated objects of modified obejcts to appear in the section.
So to repeat my question please let us know, what option we need to check-off in Powergen for the "Descendant of modified object" section to show up in the log file. Thanks.
Thanks!
Nagi
Also we are using PowerGen Version 5.5.1.2.
There is no option in PowerGen 5.5.1.2 to enable the "Regen descendants of modified objects".
Could you please let us know. Thanks.
Nagi,
Sorry for the delay getting back to you.
The proposed solution that I described doesn't exist yet. I was just trying to get your agreement that it would meet your needs before we implemented it. As an example, with the change I'm proposing the output log would like this for:
Regenerating Objects in Inheritance Order
DataWindows
Structures
[None]
User Objects
Windows
w_center C:\Dev\PB10 RT\pbexbm.pbl 09/08/05 16:28:03
Descendant of w_center
w_benchmark C:\Dev\PB10 RT\pbexbm.pbl 09/08/05 16:28:03
Descendant of w_center
w_machine C:\Dev\PB10 RT\pbexbm.pbl 09/08/05 16:29:03
Descendant of w_center
w_report C:\Dev\PB10 RT\pbexbm.pbl 09/08/05 16:29:06
...
Note that the regeneration of descendants would be intermingled with the regeneration of the modified objects, as it is currently. To move the regeneration of descendants to somewhere else in the process would be risky.
Would this be acceptable to you?
Regards,
Phil
Hi,
We do our PowerBuilder builds with PowerGen. We have built an PowerBuilder utility which parses the PowerGen log file and provides us with important build summary results.
Some of build summary reports show the following data:
Which objects got regenerated because they were physically modified?
Which objects got regenerated because they are dependent objects of modified objects?
The way we categorize the objects as being compile dependent/ compile modified is based on where they appear in the PowerGen log file.
IF the object appears in the {Regenerating Objects in Inheritance Order} section we mark it as Modified.
IF the object appears in the {"Regenerating objects that use changed objects as variables", "Regenerating descendants of dependent objects", "Creating PBD's/DLL's"} section we mark it as Dependent.
The problem we are experiencing is as follows.
I will try to explain with three objects
uo_objectA , uo_objectB and uo_objectC .
uo_objectB is inherited from uo_objectA
uo_objectC has an instance variable of type uo_objectA
When ever we modified uo_objectA we expected our build summary report to show that
uo_objectA Compiled due to modification
uo_objectB , uo_objectC Compiled due to Dependency
In actual reality our build results are as follows:
uo_objectA , uo_objectB Compiled due to modification
uo_objectC Compiled due to Dependency
THIS IS THE PROBLEM ONLY ONE OBJECT uo_objectA HAS BEEN MODIIFED BUT BOTH OBJECTS SHOW AS BEING MODIFIED. THIS IS INCORRECT.
This is occurring because PowerGen log file has no difference between uo_objectA and uo_objectB . Both objects are listed in the {Regenerating Objects in Inheritance Order}
I FEEL THAT uo_objectB SHOULD BE IN AN NEW SECTION IN THE LOG FILE POSSIBILY Regenerating dependent Inherited objects
Please advise us as to how we can solve this problem. Please let us know if you will be changing the schema of the log file. Please advice. Thanks.
- Nagi
I didn't see anything in your reply. Would the solution I proposed be acceptable to you?
Regards,
Phil
The solution you proposed should work.
Please let me know in which version it will be implemented in.
- Nagi
Pages: 1