========================================================================
 Issue: 466682    04/23/07  	Status: Closed	04/27/07  
Severity : 4         	Est. release: tyresias
Label  : PDM documentation: refernced version should include those greater tha
	 n v8
Description :                             
	 *** Monday, April 23, 2007 12:00:13 PM *** ahill ***
In the PDMUG, the following text is listed under the Reference Properties 
section:

Implementing referential integrity
In PowerDesigner 6, the implementation of referential integrity was defined
as a generation option. However, in PowerDesigner 8, you indicate
the implementation of referential integrity using declarative referential
integrity, or by the use of triggers, from the Integrity tab of the reference
property sheet.


The customer feels that the text 'in PowerDesigner 8,' should read 'in 
PowerDesigner 8 or greater'. 
========================================================================
 Issue: 466614    04/20/07  	Status: Closed	04/24/07  
Severity : 2         	Est. release: tyresias
Label  : DB Gen: GTL - Case format option prevents Collection information bein
	 g generated
Description :                      
	 *** Friday, April 20, 2007 8:03:52 PM *** ahill *** 12.1.0 EBF9 & build 
2139
\\neo\cir\...\466614
The Version12 and Version11 zip files include the actual GTL syntax that the 
customer is using.


DB Gen: GTL - Case format option prevents Collection information being generated

The customer has added similar code to the Table\AfterCreate object in the DBMS 
file:

select 'in AfterCreate' from dual;

.foreach_item(Columns)
  select 'In foreach loop' from dual;
.next(\n)

select 'end of AfterCreate' from dual;

In v11, this generates the following code:

select 'in AfterCreate' from dual;
  select 'In foreach loop' from dual;
  select 'In foreach loop' from dual;
select 'end of AfterCreate' from dual;

In v12, the following is generated for the same table:

SELECT 'in AfterCreate' FROM DUAL;

SELECT 'end of AfterCreate' FROM DUAL;


In review of the customer's files, Marc determined that the problem is due to 
the generation character case. The customer has uppercase selected. If you 
select mixed case in v12, the collection is evaluated properly.

The workaround is to select 'mixed case' for the generation character case.

 
========================================================================
 Issue: 466284    04/18/07  	Status: Closed	04/27/07  
Severity : 2         	Est. release: tyresias
Label  : OOM:generation of xsd returns error: unresolved member targetNamespac
	 e
Description :                                
	 -----------------------
PSE NOTES:
-----------------------
Description of Problem:   PDes 12.1/OOM/XML Schema
Forward engineer returns namespacePrefixA translation error: unresolved member 
targetNamespace


OS:  w2000
Versions and Build #s Reported:  12.1 ebf 1 GA
Versions and Build #s Tested:  PD 12.1 ebf 9
Database/Driver File/Version:  Not db specific.
Comments:  
                                                                                                                                   
========================================================================
 Issue: 465897    04/13/07  	Status: Closed	04/25/07  
Severity : 2         	Est. release: tyresias
Label  : PDes 12/DB2MVS8/fails to add new column if tablespace has database as
	 signed
Description :                           
	 -----------------------
PSE NOTES:
-----------------------
Description of Problem: PDes 12./DB2MVS8
fails to add new column if tablespace has database assigned

OS:  w2000
Versions and Build #s Reported:  12.1 ebf 9, In PDes 11.1.0.1547 it works very 
well. Versions and Build #s Tested:  12.1 latest ebf.
Database/Driver File/Version:  Not db specific.
Comments:  In PDes 11.1.0.1547 it works very well, So the bug must have come 
into PDes between EBF01 and EBF10 for PDes 11.

                        
========================================================================
 Issue: 464435    03/30/07  	Status: Closed	04/20/07  
Severity : 2         	Est. release: tyresias
Label  : XML : Replicate object : Annotation object is no longer be synchroniz
	 ed
Description :                               
	 During replication of element, simpletype and complextype, Annotation 
object is not longer synchronized

Works fine with PD 11.1 broken from PD 12
reproduced in PD 12 GA, PD 12.1 ebf 8 and PD 12.5 beta 1

Cr 324875 seems to be back but only for Annotation object, the other properties 
described int he cr are OK

Steps to reproduce
New XML model
Add an element
Element properties --> Annotation
Add a documentation or an application information
fill in the ID
New xml, replicate the element just created int he first xml model
The replication of the attribute Annotation object wil be no longer be 
synchronized.
Note the ID has been replicated
Check the replication properties --> Attributes : Annotation object is unchecked 
========================================================================
 Issue: 464366    03/29/07  	Status: Closed	04/10/07  
Severity : 1         	Est. release: tyresias
Label  : application error when consolidating PDM
Description :                                                              
	 *** Thursday, March 29, 2007 10:49:23 PM *** ahill *** 12.1.0 EBF8

(I have been unable to test in 12.5.0.2121. I get application errors when trying 
to open the properties of a reference.) 

\\neo\cir\...\464366

Repository has been upgraded from v11.1 to 12.1.0

The customer has modified a PDM and attempts to consolidate it back into the 
repository. The consolidation fails with an application error:

Src\repcntxt.cpp (2666) Context applying action: Changing format of symbols 
(unknown)

Upon review of the customers model, there is a reference, 
TST_DSCLSR_STS_TYP_FK1, with the same name and code but a different parent/child 
combination for the join criteria. If I try modifying the name/code of one of 
these references, I get an application error.

I believe that the problem may be due to the stereotypes associated with several 
references. Via the list of shortcuts, there are several references that have 
stereotypes associated with them that do not exist in the repository. However, 
when comparing this model with an extracted copy, these references are not 
flagged as different.

To reproduce
Customers repository is installed on MS SQL Server 2000. For the repository 
user, I used ADMIN and blanked out the password via sql.

1) Open the customers PDM
2) Connect to the customers repository
3) Consolidate using integrate into the folder \Development\Multi-Program 
Registration and Reporting (MMR)\Development Data Models\ 
 
========================================================================
 Issue: 464022    03/27/07  	Status: Closed	04/10/07  
Severity : 3         	Est. release: tyresias
Label  : Classes  cannot be linked to sub-Activities as extended dependencies
Description :                                  
	 classes cannot be linked to Activities as extended dependencies if this 
activity is a sub-activity of another.
It is only possible to select in "add object" the class attributes but not the 
class itself

Steps to reproduce
New oom 
1 class diagram with one class few attributes
1 activity diagram with 1 activity
decompose the activity
Sub-activity : extended dependencies, add objects
Only the attributes of the class are listed not the class itself

Works fin in PD 12.5 #2117 and #2119              
========================================================================
 Issue: 463577    03/23/07  	Status: Closed	04/20/07  
Severity : 2         	Est. release: tyresias
Label  : Update repository : merge inconsistent
Description :                                                                
	 During a second update, the changes made by another user are checked by 
default in the merge even if the current user didn't modified the objects. Due 
to that during one update, an user can cancel all the work done by another one.
Reproduced with PD 12.1 ebf 7 and 8 and with PD 12.5 beta 1
Repository on ASA
Steps to reproduce

1 oom 
2 activity diagrams D1 with one activity A1 and  D2 with one activity A2
Consolidate
Extract oom to OOM1
Extract oom to OOM2

OOM1
Move symbol activity A1 into D1 diagram from 1 position to another one
consolidate with update extract checked (this one is very important to reproduce 
the problem)
in the merge, A1 is checked by default which is correct
Save

OOM2
Move symbol activity A2 into D2 diagram from 1 position to another one
consolidate with update extract checked
In the merge windows
A1 is unchecked by default and A2 checked ; Which is correct
Save

Back to OOM1
Move symbol activity A1 into D1 diagram from 1 position to another one
consolidate with update extract checked
in the merge A1 is checked A2 unchecked : Which is correct
Save

Back to OOM2
Move symbol activity A2 into D2 diagram from 1 position to another one
consolidate with update extract checked
in the merge window
A1 and A2 are checked : which is not correct

After each consolidation and extraction, I compared model in extracted with the 
one stored in the rep, each time,no change has been detected


This problem  is a huge problem for the customer as they are working with many 
diagrams with  many different users. 
 
========================================================================
 Issue: 463554    03/23/07  	Status: Closed	05/07/07  
Severity : 2         	Est. release: tyresias
Label  : Unexpected error while updated an OOM with a shortcut of the diagram
Description :                                  
	 When an oom is including a shortcut of another diagram, during the 
repository update : Unexpected error while updated an OOM with a shortcut of the 
diagram
The consolidation of the document has been cancelled

Steps to reproduce

New OOM with few classes : M1
Consolidate with extract
New PDM or new CDM or new PDM : M2
with few objects
Consolidate and extract

Activate M1 the OOM
From the workspace select the diagram and drag and drop it into the OOM
You should get a shortcut of the M2 diagram 
Consolidate the OOM with update
Unexpected error while executing the following action
Linking object Shortcut of Physical diagram/XML Model Diagram with object target 
model reference

The consolidation of the document has been cancelled

Could not reproduce this behaviour when using
CDM with shortcut OOM 
PDM with shortcut OOM 
XML with shortcut OOM 
OOM with shortcut OOM 
PDM with shortcut CDM
PDM with shortcut XML

reproduced also with 12.5 beta1
Could be related to this problem : when you delete such diagram shortcut in the 
OOM even if you select "delete objects" and not only "symbols", the symbol will 
be listed  in the Symbols dialog box 
========================================================================
 Issue: 463353    03/21/07  	Status: Closed	04/09/07  
Severity : 2         	Est. release: tyresias
Label  : Invalid drop table DDL generated for tbl using 'with restrict on drop
	  option
Description :                         
	 *** Wednesday, March 21, 2007 4:38:14 PM *** ahill *** 12.1.0 EBF7 & 
12.5.0.2114
Target DBMS: IBM DB2 UDB 7 OS/390

Customer has generated a table using the with restrict on drop physical 
option. When using PD via database generation or modify database to drop the 
table, the drop statement generated by PD fails when executed via ODBC with the 
following error:

[IBM] [CLI DRIVER][DB2] SQL0672N Operation DROP not allowed on table 
xxx.tabl_name. SQLSTATE = 55035

Per the DB2 documentation

55035 = the table cannot be dropped because it is protected.

   -672   OPERATION DROP NOT ALLOWED ON TABLE table_name
Explanation: The DROP operation failed for one of the following reasons: 
 The table being dropped has the RESTRICT ON DROP attribute. 
 The table space or database being dropped contains the specified table, which 
has the RESTRICT ON DROP attribute. 

System Action: The DROP statement cannot be executed. 

Programmer Response: Before dropping the table, alter the table, specifying DROP 
RESTRICT ON DROP. 


The following table has been generated to the DSNDB0X DB2 7.1 OS/390 server

In order to drop the table this alter statement must be executed before the drop 
table statement

alter table AHILL.TABLE_4
     drop restrict on drop;

< syntax generated by PD>
drop table AHILL.TABLE_4;


create table AHILL.TABLE_4 (
   F                    CHAR(5)                not null,
   A                    CHAR(5),
   G                    CHAR(5),
   H                    CHAR(5),
   constraint P_KEY_1 primary key (F)
)
in database DSNDB04
with restrict on drop
ccsid ebcdic;
 
========================================================================
 Issue: 463324    03/21/07  	Status: Closed	03/30/07  
Severity : 3         	Est. release: tyresias
Label  : Generate DB: 'Use database' statement no longer generated before drop
	 ping object
Description :                      
	 *** Wednesday, March 21, 2007 11:30:42 AM *** ahill *** 12.1.0 EBf8 & 
12.5.0.2114
Target DBMS: ASE family 

Generate DB: 'Use database' statement is no longer generated before dropping 
objects.

In previous releases, the 'use database' syntax was generated before the drop 
statements for objects. In 12.x, the 'use database' statement does not get 
generated until after the last drop statement is generated.

To reproduce
target DBMS: ASE 12.5.2 
1) add table with several columns
2) add a database object 
3) select generate database
4) for the database object select 'use database' for table object select all 
options
5) preview the ddl
Note: the drop statement for the table is generated before the 'use database' 
statement.

If the same is done in PD 11.x or earlier, the 'use database' syntax is 
generated before the drop statements.

workaround is to add the 'use database' statement into the script header item in 
the XDB file. 
========================================================================
 Issue: 462683    03/14/07  	Status: Closed	03/27/07  
Severity : 3         	Est. release: tyresias
Label  : DB Gen: Object filter omits alternate key if not defined in table at 
	 model level
Description :                      
	 *** Wednesday, March 14, 2007 10:03:41 PM *** ahill *** 12.1.0 EBF7 & 
12.5.0.2103
\\neo\cir\...\462683

Database generation:
Object filter incorrectly filters 'alternate key' if there are no alternate keys 
defined at the model level.

It appears that when an alternate does not exist at the root level of the model, 
PD incorrectly filters alternate keys from the Options tab of the Database 
Generation window. However, in the generated DDL, the alternate key is 
generated.

Workaround:
change the filter to view 'all objects'

1) Open sample PDM
    contains a table at the root and in a package.
    the table in the package has an alternate key defined
2) Select generate database
  Note: the alternate key option is not available for selection
3) Add an alternate key to the table at the root level
4) Select generate database
  Note: the alternate key option is now available  
========================================================================
 Issue: 461524    03/05/07  	Status: Open	
Severity : 2         	Est. release: tyresias
Label  : Procedure with very long lines full of space cannot be generated
Description :                                                  
	  *** lundi 5 mars 2007 10:03:31 *** vaudino ***12.0
open procedure property sheet.
call preview tab
You have an hour glass and PD does not come back                                                                                                                                                                                                                                                                                                                                                                          
========================================================================
 Issue: 461272    03/01/07  	Status: Closed	04/25/07  
Severity : 2         	Est. release: tyresias
Label  : model Update: failed to connect to the target model
Description :                                                   
	 It is not possible to update a model when the target model has not been 
created from a generation, the first update will be OK but as soon as the source 
model will be closed and reopened, it will be not possible to update the target.

Steps to reproduce
1 new CDM 1 entity
1 new PDM 1 table
save both models
from CDM
generate with update the PDM
select the pdm created
save both models
Check target models
CDM : OK PDM is linked
PDM nothing
close the cdm
reopen it
from cdm generate with update the pdm
in the select dropdown, the pdm is selected
note : dbms  is empty
click ok
got message : the model ... is not the target model, are you sure you want to  
use it in place of the target model
yes
Error message : failed to connect to the target model
following by
a model must be selected
and it is not possible anymore to update the pdm


Same problem between OOM and PDM, after the first update if OOM is closed, it 
will not possible anymore to update the PDM 
========================================================================
 Issue: 461238    02/28/07  	Status: Closed	03/13/07  
Severity : 1         	Est. release: tyresias
Label  : Find Objects function can expose hidden Doucments name in Repository
Description :                                  
	 -----------------------
PSE NOTES:
-----------------------
Description of Problem:
Using the access mode "List" on a Document requires that the Parent Node 
(folder) must have
at least "Read" access permission. That means if Parent folder with "List" 
access mode will 
hide all the Documents inside the folder. However, it is reported that 
Repository function 
"Find Objects" does not care the access mode. Though the "Find Objects" function 
does not 
allow to search the Content of Documents, the function can expose the name of 
hidden 
documents of the folder.

OS:  Win XP
Versions and Build #s Reported:  PD 12.1 EBF 2
Versions and Build #s Tested:  PD 12.1 EBF 7
Database/Driver File/Version:  Tested with DataDirect OEM 5.1 SQL Server ODBC
Comments: 
--------------------------------------------
"Repository Browser View" need the Parent Node to have at least "Read Access 
Mode" to 
display the Child Content with "List Access Mode". e.g. Document with "List 
Access mode". 
Otherwise the Repository Browser View only Display the Parent Node and stop at 
the Parent 
Node.

On the other hand, for "List Access Mode" Document, the "Find Objects" 
repository function 
does not care whether the Parent Node is at least "Read Access Mode" or not. It 
will just 
search to Documents Name with "List Access Mode". (but not the content of Doc 
with List 
mode).

With this testing result, it seems that "Find Objects" function has a little 
conflict when 
compare with the "Repository Browser View".

Besides, after the execution of  "Find Objects", the "Repository Brower View" 
will be 
expanded to show more nested Folders and Document. i.e. The rule - "List Mode 
Document need
 the Parent Node to have at least 'Read Access Mode' "  is no longer exist.
-------------------------------------------- 
========================================================================
 Issue: 460689    02/23/07  	Status: Closed	04/26/07  
Severity : 2         	Est. release: tyresias
Label  : Option "Update" gets disabled when updating model into repository
Description :                                     
	 Tested and reproduced with PD 12.1 and PD 12.1 EBF#6

When updating a model after a merge, the option "Update" gets disabled when 
updating model into repository.                                                                                                                                                                                                                                                                                                                                                              
========================================================================
 Issue: 456058    02/13/07  	Status: Closed	03/01/07  
Severity : 3         	Est. release: tyresias
Label  : Test data : value 10 with number(1)
Description :                                                                   
	 Test data generates the value 10  even if length of the datatype is 1 
reproduce with Number, decimal and float

Steps to reproduce

1 table
C1 Number(1)
distinct value : 100%
Generate test data
No profile used
Number of Rows : 20
In the list of values generated we got Number(10)

Note :  we got only the value 10 and not higher                                                                                                                                                                           
========================================================================
 Issue: 455955    02/12/07  	Status: Closed	03/13/07  
Severity : 3         	Est. release: tyresias
Label  : Creation and Modification dates/users are set to Unkown in PDM to CDM
	  generation
Description :                      
	 Tested and reproduced with PD 12.1.0 and 12.1.0 EBF#6

When generating CDM from PDM, the Attribute Version Info properties display 
"Unkown" for Creation User, Creation Date and Modification User.                                                                                                                                                                                                                                                                                                                            
========================================================================
 Issue: 454996    02/02/07  	Status: Closed	03/01/07  
Severity : 3         	Est. release: tyresias
Label  : Shortcut objects in HTML report using SVG image are not hyperlinked (
	 navigable)
Description :                       
	 *** Friday, February 02, 2007 2:46:54 PM *** ahill ***  12.1.0 & 12.5.0 
build 2071
\\neo\cir\...\454996 (includes sample model and generated reports)

When generating HTML report using SVG as the image format, included shortcut 
objects do not have hyperlinks in the generated graphic.  However, generating 
the same report with a different image format, the shorcut objects do have 
hyperlinks.

To reproduce:

1) Open sample model
2) Open the report in the model and select 'png' as the image format 
3) Generate a HTML report
4) In the generated report,click on the package_2 diagram link
5) Click on the entity shortcut
Go back to PD
6) Select 'svg' as the image format 
7) Generate a HTML report
8) In the generated report,click on the package_2 diagram link
9) Click on the entity shortcut
NOTE: the entity shortcut is not hyperlinked 
========================================================================
 Issue: 454330    01/26/07  	Status: Closed	03/01/07  
Severity : 2         	Est. release: tyresias
Label  : HTML Report has broken link in "Next Button" for Table List
Description :                                           
	 Tested and reproduced with PD 12.1.0, 12.1.0 EBF#4, 12.1.0 EBF#5 and 
12.1.0 EBF#6

After generating the HTML Report, when navigating through it, inside Table List, 
the Next button leads to a "Page not found".
Similar CRs were fixed in previous builds of PD 12.1.0: CR#431348 and CR#426101                                                                                                                                                                                                                              
========================================================================
 Issue: 454206    01/25/07  	Status: Closed	02/15/07  
Severity : 3         	Est. release: tyresias
Label  : View "instead of" triggers fail to reverse engineer via script for MS
	  SQL SVR 2k
Description :                      
	 *** Thursday, January 25, 2007 3:59:36 PM *** ahill *** 12.1.0 EBF5
\\neo\cir\...\454206

Target DBMS: MS SQL Server 2000
View "instead of" triggers fail to reverse engineer via script.

Workaround: reverse engineer via ODBC

To reproduce:

1) Reverse engineer mytest.sql 
Note: Trigger SQL is skipped
2) Reverse engineer the same via ODBC, the trigger is reverse engineered.                                                                                                                              
========================================================================
 Issue: 452756    01/11/07  	Status: Closed	04/26/07  
Severity : 3         	Est. release: tyresias
Label  : After RE, a space is added  in partion by range column list
Description :                                           
	 A change has been detected in database synchronisation windows even if 
nothing has been  modified in the model.
It is due to the fact than PD adds a space in the table partition by range list 
columns after the reverse

Steps to reproduce with PD 12 and 12.1 ebf 5
RE into a new PDM the following script
create table T1  (
   C1             CHAR,
   C2             CHAR,
   C3             CHAR
)
partition by range
(C2)
(
partition Partition_1 values less than (0)
tablespace tspace);
Without doing nothing

Archive the model
Go to T1 properties --> Tab  partition
Have a look at Column list a space has been added after the column name
Click apply
Generate a script to modify database, 
In the synchronisation windows, a change is detected due to the space added in 
the column list.
Of course, if you archive the model again the problem will be solved

 
========================================================================
 Issue: 452615    01/10/07  	Status: Closed	01/25/07  
Severity : 2         	Est. release: tyresias
Label  : DBcreateAfter is not analysed during generation when objects are in p
	 ackages
Description :                          
	 cr 360644 is back in 12.1, has been fixed in 11 
Regression, working fine in 11, 11.1, 12
Broken in 12.1

DBcreateAfter is not analysed during generation when objects are in packages. 
Problem occurs for views and for procedures.
No problem without packages

Steps to reproduce 
New pdm with 2 packages 
Package_1 : View_aa 
Package_2 : View_bb 
Add to View_aa, an extended dependencies to View_bb and select stereotype 
DBCreateAfter 

Generate the script 
View_aa has been generated before view_bb 
 
========================================================================
 Issue: 452410    01/08/07  	Status: Closed	01/25/07  
Severity : 2         	Est. release: tyresias
Label  : XEM ediotr: Model Type for Transformation Profile mapped to incorrect
	  model type
Description :                      
	 *** Monday, January 08, 2007 11:05:57 AM *** ahill *** 12.1.0
**May be related to CR 319875**

When adding a transformation profile to an XEM, the Model Type on a profile maps 
to the wrong model type. The following mappings are incorrect:

'Conceptual Data Model' maps to 'Erwin 3.5.x model as CDM'
'Erwin 3.5.x model as PDM' maps to 'Physical Data Model'
'Object-Oriented Model' maps to 'Rational Rose Class Diagram'

To reproduce:
Create an XEM and add a Transformation Profile
Add a Profile
Click the Model type combo box and select 'Conceptual Data Model'
Note: The Model type immediately maps to 'Erwin 3.5.x model as CDM' 
========================================================================
 Issue: 452398    01/08/07  	Status: Closed	01/29/07  
Severity : 3         	Est. release: tyresias
Label  : Incorrect expression for view column when using distinct in SQL Query
	 
Description :                                 
	 When create a view with 'select distinct' or 'select all' in SQL query 
body , column expression is incorrect 

Steps to reproduce

New PDM
1 table T1(C1 char, C2 char, C3 char)
Create a view based on this table 
Go to check View column expression for C1 is T1.C1
Go to SQL Query and modify the statement
Add "Distinct" after the select
you should get
select Distinct
   T1.C1,
   T1.C2,
   T1.C3
from
   T1

Go to tab  view column
Have a look at the column  C1 
Name is T1.C1 and expression is distinct 
========================================================================
 Issue: 451940    12/28/06  	Status: Closed	04/09/07  
Severity : 3         	Est. release: tyresias
Label  : Some scripts not reversed ok for MS-SQL Server 2000 DBMS
Description :                                              
	 Tested and reproduced with PD 12.1 and PD 12.1 EBF#5.

Some scripts with ALTER COLUMN and REVOKE/GRANT are successfully reversed into 
existent PDM, but no expected changings are done to the objects inside PDM.                                                                                                                                                                                                                                                                                                              
========================================================================
 Issue: 451781    12/21/06  	Status: Closed	02/19/07  
Severity : 2         	Est. release: tyresias
Label  : Reverse  view with distinct  : no dependencies
Description :                                                        
	 Tested with 10, 11 but view columns dependencies didn't exist
Broken in 11.1, 12 and 12.1 ebf 5

When reverse a view with "select distinct" or "select all" in SQL query body , 
column expression is incorrect and dependencies for the first view column is 
lost.

Steps to reproduce
RE into a new pdm the following script
Create Table T1 (c1 char, c2 char));
Create view V1 as
Select 
c1,
c2
from 
T1;

View --> properties 
Dependencies : Table T1 OK
View columns --> Properties for C1 and C2
Dependencies c1 parent Table T1 OK

Modify the script by ading distinct, you should get
Create view V1 as
Select distinct
c1,
c2
from 
T1;

Re into a new pdm
the table and the view script 
go to check the view
View --> properties dependencies : t1 ok
column view c1 --> Porperties dependencies : empty
column view C2 --> properties dependencies : OK

Note : if Distinct is added in the SQL Query editor, no problem at all, problem 
occurs only during the reverse when the view query contains Distinct or All 
========================================================================
 Issue: 451779    12/21/06  	Status: Closed	01/29/07  
Severity : 3         	Est. release: tyresias
Label  : No datatype  when using select distinct in a View
Description :                                                     
	 Reproduced with PD 10.1, PD 11, PD 11.1 and PD 12.1 ebf 5

When create or reverse a view linked to a table, view column will  have the 
datatype defined to the table column. If distinct or All is added in the query 
view the first column view will not have anymore a datatype

Steps to reproduce

New PDM
1 table T1(C1 char, C2 char)
1 view
View Query
Select
C1,
C2
from
T1
 go to check column view :datatypes are visible

Modify the SQL Query by adding Distinct
Select distinct
C1,
C2
from
T1
apply,

No datatype anymore for column view c1.

If exporting data through automation the missing datatypes cause major problems. 
========================================================================
 Issue: 451749    12/20/06  	Status: Closed	02/19/07  
Severity : 1         	Est. release: tyresias
Label  : Fail to consolidate repository when Item sequence changed in XML mode
	 l.
Description :                               
	 Description of Problem: Item sequence of XML model with multiple Item 
types (e.g. Elements, Complex Types and Shortcuts of Complex Types) is NOT 
correctly updated to Repository.
OS:  Windows XP
Versions and Build #s Reported:  PowerDesigner 12.1.0.1948 EBF4
Versions and Build #s Tested:  PowerDesigner 12.1.0.1960 EBF5
Database/Driver File/Version:  Any
Comments:  
                                                                                                                                            
========================================================================
 Issue: 451671    12/20/06  	Status: Closed	04/13/07  
Severity : 3         	Est. release: tyresias
Label  : PowerDesigner Java Proxies : sample JAR  file doesn't return out para
	 meters to t
Description :                      
	 in the PD scripting API, the pd2com_12.1.0.jar file doesn't return out 
parameters to the caller. 
One of our consultant(Christian Kuhle)  did some modification  to this file 
File modified has been copied in
\\neo\CIR\CIRMails\45xxxx\451671

Eclipse SWT libraries only use datatype Variant as ByRef parameters, so they all 
have to be of type "COM.VT_VARIANT|COM.VT_BYREF". 
The parameter handling inside the IDispatchEx.invoke method was changed to pass 
ByRef parameters with correct datatype. Return parameter handling after method 
execution needed changes as well, e.g. parameter order was reversed according to 
Microsofts OLE convention. 
Maybe parameter handling needs further refinement, at least ByRef parameter 
values are returned to the caller with these changes. 

I checked the file pd2com_12.1.0.jar sent by him and modification have been done 
in
BaseObject.class
public void GetCollectionByKind(/* in */ int kind, /* out */ ObjectBag[] 
collection)
public void GetCollectionByKind(/* in */ int kind, /* out */ ObjectBag[] 
collection, /* in */ boolean compositionOnly)
public boolean CanSetAttribute(/* in */ java.lang.String member, /* in */ 
VariantEx value, /* out */ java.lang.String[] message) 
public boolean CanAddInColl(/* in */ java.lang.String member, /* in */ 
BaseObject object, /* out */ java.lang.String[] message)
public boolean CanRemoveInColl(/* in */ java.lang.String member, /* in */ 
BaseObject object, /* out */ java.lang.String[] message)
public boolean CanCreateObject(/* in */ int kind, /* out */ java.lang.String[] 
message) 
public boolean CanCreateObject(/* in */ int kind, /* out */ java.lang.String[] 
message, /* in */ java.lang.String parentCol)
public boolean CanCreateObject(/* in */ int kind, /* out */ java.lang.String[] 
message, /* in */ java.lang.String parentCol, /* in */ int pos)

CheckModelBaseItem.Class
public void Verify(/* in */ BaseObject obj, /* out */ BaseObject[] msg)

CheckModelControler.Class
public void Execute(/* out */ int[] err, /* out */ int[] wrn)

Changes in the code are marked with a comment line starting: 
   //C. Kuhle 

The modification done is not complete, he only changed the classes that he was 
using in his application
 
========================================================================
 Issue: 451335    12/15/06  	Status: Closed	01/29/07  
Severity : 3         	Est. release: tyresias
Label  : Teradata  unique index  with option generation inside :wrong syntax i
	 s generated
Description :                      
	 Result OK in 11 broken in 12 and 12.1

Syntax is wrong for primary index when the database generation option create 
Index inside is checked 
Steps to reproduce
New PDM Teradata
1 table with few columns and PK
Create a new index linked to the pk
Preview with the generation option Create Index Inside checked

create table T1 (
    C1                   CHAR                           not null,
    C2                   CHAR,
    C3                   CHAR
);

create unique index IDX_1 (
   C1
) on T1;

Instead of

create table T1 (
    C1                   CHAR                           not null,
    C2                   CHAR,
    C3                   CHAR
)
 unique primary index IDX_1 (
   C1
); 
========================================================================
 Issue: 450460    12/14/06  	Status: Closed	03/27/07  
Severity : 2         	Est. release: tyresias
Label  : Extended Attribute user method is not triggered by Column properties
Description :                                  
	 Tested and reproduced with PD 12.1 and PD 12.1 EBF#4.

When coding an user method inside Extended Attributes and attempting to trigger 
it from within Attributes properties.                                                                                                                                                                                                                                                                                                                                                   
========================================================================
 Issue: 450329    12/13/06  	Status: Closed	01/25/07  
Severity : 4         	Est. release: tyresias
Label  : Refresh : Undo action restores View column deleted with incorrect nam
	 e
Description :                                
	 reproduce with 9.5, 9.5.1, 9.5.2, 10, 10.1, 11.1, 12.1 and with 12.5 2027
Refresh problem in the browser on view column when  view columns have been 
deleted and CRTL Z have been used.

Close the model and reopen it solves 
the========================================================================
 Issue

Steps to reproduce
New PDM ASA
1 view
SQL Query 
Select
A,
B,
C
from
Toto
From the browser, delete view columns A and after B
CTRL Z twice to cancel the 2 last actions
In the browser, the result is A, A, C (in the diagram, view columns are OK)
From the browser, Edit the properties of both A columns, Name/Code are correct
First A : Name/Code is A
Second A : Name/Code is B
Even if It is only a refresh problem in the brouwser, it can be very annoying
Note If you delete first column B and after column A, CTRL Z twice, in this case 
the result will be OK.

*** mardi 23 janvier 2007 16:58:04 *** vaudino ***
You should not be allowed to delete view column like this.
It is forbidden in Columns grid from View property sheet and should be the same 
in the browser 
========================================================================
 Issue: 450318    12/13/06  	Status: Closed	01/29/07  
Severity : 3         	Est. release: tyresias
Label  : Update repository will increase alll views version number
Description :                                             
	 Broken from V10, It was OK with previous version (tested with 9.5, 9.5.1, 
9.5.2)
Broken In V10, V10.1 V 11, V11.1 V12.1 ebf 4

If you update the repository even if views are not modified their version 
numbers are increased


New PDM ASA
2 views
Vieuw : V1
SQL Query
Select A
from Toto
View 2 : V2
SQL Query
Select B
From Tata
Consolidate with freeze after consolidation and extract
Via ISQl, execuet the following
select OBJT, VRSN, NAME from PMOBJT
result
OBJT 	VRSN	NAME
48	1	V1
49	1	A
50	1	V2
51	1	B

Return to the model and modify View V1 (only), modify sql query
Select 
A,
AA
from Toto

Consolidate with freeze
In the merge model, view column is going to be added and sql query is going to 
be updated.
Accept all the changes
Note V2 is not going to be modified
Check the reposiotry, all view versions have been increases
Via ISQl, execute the following
select OBJT, VRSN, NAME from PMOBJT
result
OBJT 	VRSN	NAME

49	1	A
51	1	B
55 	2	V1
57	2	A
59	2	V2
48	1	V1
50	1	V2
54	1	AA
58	2	B

delete the document in the rep and run the server with option -oc:asa9\test.sql 
in order 
to get the SQL executed during the update

Consolidate with freeze and extract
modify the PDM update view V1
Chech the trace got
insert into 
PMOBJT(OBJT,CLSS,POID,BRNC,LVRS,FRZN,PVOB,VRSN,OCUS,OCDT,OMUS,OMDT,GOID,NAME,COD
E,STRN,CMMT,DSCR,ANNT,HIST,OEXA,BCOL,CUSR,CDAT,MUSR,MDAT) select 59, 1211, POID, 
1, 1, 0, 50, '2', 200, ?, 200, ?, 
GOID,NAME,CODE,STRN,CMMT,DSCR,ANNT,HIST,OEXA,BCOL,CUSR,CDAT,MUSR,MDAT from 
PMOBJT where OBJT=50

So why this version has been updated for View V2


New test
New PDM
View V1
Select
A
from Toto
View V2
Select
B from Tata
Table T1 (C11 integer)
Table T2(C21 integer)
Consolidate with freeze in order to get version and extract
modify pdm
Table T1 add a new column consolidate with freeze
Check the rep
T1 version has been increased which is correct
T2 version has not been  increased which is correct
but both view got their version number increased so even when no view have  been  
modified, their version number are increased
 
========================================================================
 Issue: 450239    12/12/06  	Status: Closed	01/26/07  
Severity : 2         	Est. release: tyresias
Label  : Default Physical Options: Apply to list is empty for the Primary Key 
	 option
Description :                           
	 *** Tuesday, December 12, 2006 5:49:51 PM *** ahill *** 12.1.0 EBF4
\\neo\cir\...\450239

Target DBMS: Oracle and ASE family databases

The customer is attempting to set Default Physical Options for Primary keys in a 
PDM targeting Oracle 9i. After setting the physical options, they select the 
Apply To button, but the list of PK objects is empty.  There are over 200 PKs in 
the customer's model.  (This works find in PD 11.1.0 EBF10.)

To reproduce:
1) Open the customer's model or project.pdm (change target DB to Oracle or ASE 
12 or greater)
2) Open Database\Default Physical Options and select Primary Key tab
3) Select Apply to button
Note: There are no PK objects listed
 
========================================================================
 Issue: 449500    12/04/06  	Status: Closed	03/15/07  
Severity : 2         	Est. release: tyresias
Label  : Merge : Extended collection not updated for pdcommon metaclasses
Description :                                      
	 PD fails to update extended collection value when the metaclass is coming 
from  PDCommon

Steps to reproduce
Test 1
New PDM ASA
In the xdb, add metaclass Business rule 
Defined an extended collection target type : Table
Into the new PDM, 
2 tables T1, T2
add a business rule extended collection T1
From PDM1, generate PDM2
Return to PDM1 and business rule --> Properties --> collection add T2
Merge PDM 1 into PDM 2
In merge window, T2 is going to be added accept the change
Model success fully merged
Go to check in PDM 2, the business rule, the second table has not been added


Test 2

New PDM ASA
In the xdb, add metaclass Business rule 
Defined an extended collection target type : Table
Into the new PDM, 
2 tables T1, T2
add a business rule without extended collection 
From PDM1, generate PDM2
Return to PDM1 and business rule --> Properties --> collection add T1
Merge PDM 1 into PDM 2
In merge window, T1 is going to be added accept the change
Model successfully merged
Go to check in PDM 2, the business rule : no problem at all


So problem occurs only when the collection contains already an object
I reproduced using 
CDM with XEM using Business rule, 
Traitement Merise with XEM and Business rule
PDM  with XDB using  FileObject or ExtendedObject metaclass
Repository update

Note cutomer got the problem during a repository  update  with Traitement merise 
Model metaclass Business rules Extended Collection target type Data 
Due to this problem, it cannot update the repository

 
========================================================================
 Issue: 449389    12/01/06  	Status: Closed	04/08/07  
Severity : 3         	Est. release: tyresias
Label  : RE by ODBC IQ 12.6 Unable to list the columns
Description :                                                         
	 If database has the option quoted_identifier = 'OFF', when RE by ODBC the 
database, an error occurs with the following mesage

Unable to list the columns.
[Sybase][ODBC Driver][Adaptive Server Anywhere]Syntax error near 'default' on 
line 11
SQLSTATE = 37000

Due to this option, it will be not possible to list the domain
error for the domain will be
Unable to List the domains
Syntax error near 'Check' on line 6

Check xdb file and
SYASIQ1260::Script\Objects\Column\SqlListQuery
replace case(lower(c."default")) by case(lower(c.[default])) 
SYASIQ1260::Script\Objects\Domain\SqlListQuery
replace    d."check" by    d.[check]

Steps to reproduce

Start asiqdemo version 12.6
via ISQL, execute the following
set  option public.quoted_identifier ='OFF';
By default quoted_identifier ='ON'
From PDesigner
New PDM IQ 12.6
RE via ODBC the database using standard ODBC driver
Errors will occur
Note : [Check] and [Default] is working using  quoted_identifier ='OFF' or 'ON'
So could be betert o use this syntax

We could get similar problem with item
SYASIQ1260::Script\Objects\Key\SqlAkeyIndex
  i."unique" not in ('Y', 'N')
 
========================================================================
 Issue: 449099    11/28/06  	Status: Closed	01/25/07  
Severity : 1         	Est. release: tyresias
Label  : RQM is exported incorrectly to printable area of MS-Word
Description :                                              
	 PowerDesigner 12.1.0.1943 EBF 3
MS-Office 2000/2003/2007

When RQM is exported to MS-Word, exported Word file is not shown by MS-Word. But 
the correct details are shown in outline mode of MS-Word. Nothing is shown in 
printing display mode. 

Is that the details of RQM are shown only for outline display mode 
specification? 

TSE sent email to powerdesigner_team.
----- Forwarded by Masahiro Koyanagi/SYBASE on 2006/11/28 PM 02:02 -----
Masahiro Koyanagi/SYBASE 
2006/11/27 PM 05:32	
To powerdesigner@sybase.com, powerdesigner_team@sybase.com
cc Masahiro Koyanagi/SYBASE
Subject Exported Word document from RQM is not printable

Hi all,

I exported a RQM to MS-Word 2000.
The RQM was exported to the outline area for MS-Word correctly.
However it was not printable.

How can I print the exported document ?
Are there any options to do so ?

I tested with PowerDesigner12.1.0.1943 EBF 3, and put the test result in bellow;

- RQM for test:
   http://www-
dse.sybase.com/cases/11301000s/11301623/tse_info/RequirementsModel_1.rqm
- The exported Word document:
   http://www-
dse.sybase.com/cases/11301000s/11301623/tse_info/RequirementsModel_1.doc

Thank you,
Masahiro Koyanagi
TSE Japan

----- Forwarded by Masahiro Koyanagi/SYBASE on 2006/11/28 PM 02:02 -----
Masahiro Koyanagi/SYBASE 
2006/11/23 AM 09:41	
To powerdesigner@sybase.com, powerdesigner_team@sybase.com
cc Masahiro Koyanagi/SYBASE
Subject Export as word documents in Requirement Model

Hi all,

My customer run "Export as word documents" for Requirement Model.
Then the model was exported to the outline area of MS-Word.
However he expected to export to the printable area.

Is this behavior normal ?
Are there any methods to export the model to the printable area of MS-Word ?

Any suggestions or comments are appreciated.

Thank you,
Masahiro Koyanagi
TSE Japan
 
========================================================================
 Issue: 448504    11/21/06  	Status: Closed	02/15/07  
Severity : 2         	Est. release: tyresias
Label  : Modify Database: table rename cause ODBC error due to existing synony
	 m
Description :                                
	 *** Tuesday, November 21, 2006 11:46:31 AM *** ahill *** 12.1.0 EBF4
\\neo\cir\...\448504

Target DBMS: DB2 7 OS/390

Customer has tables with synonyms associated with them. They modify the table 
with a change that forces the table to be dropped and recreated. The PD 
generated modify script does not contain syntax to drop the synonymn before 
renaming the table. As a result, the rename of the table fails with an ODBC 
error:

[DataDirect][ODBC DB2 Wire Protocol driver][UDB DB2 for OS/390 and z/OS]Unknown 
error: SQLCODE -148 SQLSTATE = S1000

If the DDL is modified to drop the synonym associate with the table, then the 
RENAME statement succeeds.

Per the DB2 documentation, the RENAME statement can't be used if a synonym is 
associate with the table.

To reproduce:

1) Open the syn_test1.apm file and execute generate database (may need to change 
table owner)
2) Open syn_test1.pdm and execute modify database and reference syn_test1.apm
3) Execute the modify script via ODBC. 
========================================================================
 Issue: 448230    11/17/06  	Status: Closed	01/29/07  
Severity : 3         	Est. release: tyresias
Label  : BUG: The MMR page setups' margins and orientation values are not save
	 d
Description :                                
	 *** Friday, November 17, 2006 3:19:17 PM *** ahill *** 12.1.0 thru EBF4

The customer has created a multi-model report, MMR. In a section, they change 
the page setup orientation option from portrait to landscape. After saving the 
MMR, closing it and reopening it, the orientation option has reverted back to 
portrait and all the margins are now 1 inch. Perviously, they ranged from .315 - 
.433 inches.

To reproduce:
1) Add several models to workspace
2) Create a MMR
3) Select Page Setup for Section_1
4) Change the Orientation to landscape 
Note: The values for the Margins
5) Save the MMR and close it.
6) Reopen the MMR and review the Page Setup options
Note: The orientation is portrait and the margins are set to 1 inch
 
========================================================================
 Issue: 447940    11/15/06  	Status: Closed	03/12/07  
Severity : 3         	Est. release: tyresias
Label  : Generate DB: Create rule syntax generate twice based on selected char
	 acter case
Description :                       
	 *** Wednesday, November 15, 2006 2:24:27 PM *** ahill *** 12.1.0 EBF2 
\\neo\cir\...\447940
This========================================================================
 Issue may be related to CR 291531.

Target DBMS: ASE 12.5

Generate DB: Create rule syntax is generated twice based on selected character 
case.

PD is generating an extra business rule when the EnableBindRule option is set to 
yes for the column object in the DBMS and the character case is uppercase or 
lowercase under the database generation options.

To reproduce:

DBMS modified to set EnableBindRule to yes for the column object

Create a validation rule
Create a table with a column
Add the business rule to the column
Under Model Properties select Preview
Select Database Generation Options and select the Format tab.  Set the Character 
Case to uppercase or lowercase.
Note: Two create rule statements are generated (for the business rule and for 
the column).

Select Database Generation Options and select the Format tab.  Set the Character 
Case to mixed case.
Note: Only one create rule statement is generated. 
========================================================================
 Issue: 447878    11/15/06  	Status: Closed	01/25/07  
Severity : 2         	Est. release: tyresias
Label  : RQM:Generat RTF report file is incorrect when Japanese Font is set in
	  templete.
Description :                       
	 OS: Windows XP SP2
PD: 11.1 EBF#10, 12.0 EBF#4

When Japanese fonts are set as template of RQM, PD generates wrong RTF. 
Could not open the generated file by MS-WORD.
This========================================================================
 Issue occured when he set MS P Gothic in TOC Title for Report Templete.
This issus also occurs with MS Mincho or others.

When it sets up except Japanese fonts, PD generates the correct tag. 

{\f23\fswiss Tahoma;}
{\f24\fmodern MS UI Gothic;}
}
{\colortbl
;

But when Japanese fonts are set up, PD generates wrong tag. 

{\f23\fswiss Tahoma;}
{\f24\fmodern MS UI Gothic;}
{\f25\fmodern lr oSVbN{\colortbl    <---- "}" is insufficient. 
;

When it changes as follows, RTF can be opened by MS-WORD. 

{\f23\fswiss Tahoma;}
{\f24\fmodern MS UI Gothic;}
{\f25\fmodern lr oSVbN;}
}
{\colortbl
;

This is 
fatal========================================================================
 Issue when using PD in Japanese environment.  
========================================================================
 Issue: 447797    11/14/06  	Status: Closed	04/18/07  
Severity : 1         	Est. release: tyresias
Label  : Proxy server: PDM with views fails to extract from migrated repositor
	 y
Description :                                
	 *** Tuesday, November 14, 2006 4:55:14 PM *** ahill *** 12.1.0 EBF2

This========================================================================
 Issue appears to be related to CR 396183.
Repository server : MS SQL Server 2000
Target DBMS: DB2 8 common server

Proxy server: PDM with views fails to extract from migrated repository

Customer has upgraded from PD 11.x to PD 12.1.0 EBF2. Via a direct connection to 
the repository, they can consolidate and extract all models. However, via a 
proxy server connection, they get the following error message when extracting a 
PDM:

PowerDesigner  Error
A command sent to the repository proxy failed.
22005 (SQLSTATE [Microsoft] [ODBC SQL Server Driver] Invalid character value for 
cast specification)

When reviewing the proxy server log, the following warning is documented after 
this ODBC error:

ExtractModel: Failed when extracting OBJT = 20466

Upon querying the PMOBJT table, this object id belongs to a view

To reproduce:
1) Create a repository in v10.1.x and consolidate a PDM that contains a table 
and a view
2) Upgrade the repository to v11.1.x and add a second table and view to the PDM 
and consolidate it
3) Upgrade the repository to v12.1
4) Via a direct connection, extract the PDM
NOTE: extraction of the odel succeeds
5) Via a proxy server connection, extract the PDM
NOTE: The extraction fails with the stated error.

A repository using these repro steps exists on the TBONE server. 
Database login: 		pd_eng
Database password: 		pd_eng1
Server:				tbone
Database: 			ts_data

Repository login:		ADMIN
Repository password		<BLANK>

Proxy server:			tbone
Port:				30000
Instance:			tsdata
 
========================================================================
 Issue: 447181    11/08/06  	Status: Closed	04/18/07  
Severity : 3         	Est. release: tyresias
Label  : Informix : Indexes based on function not RE by ODBC
Description :                                                   
	 Indexes based on function are not  RE by ODBC.
Such indexes are stored in the system table sysindices

Steps to reproduce
Against a server informix , execute the following script
create table T_Index  (
  C1             integer                         not null,
  C2            integer,
  C3            integer,
 C4             char(5),
primary key (C1)
      constraint pk_T_Index
)
;

create unique cluster index my_i1 on T_Index (
        C2             ASC,
        C3             ASC
);


create function my_lower (input_str char(5)) returning char(5)
with (not variant);
Define Result Char(5);
Let Result = LOWER(input_str);
Return Result;
end function;
/


create index my_i2 on T_Index (
       my_lower(C4)             ASC
);


Re into a new PDM informix
Index my_i2 has not been RE

Checked the informix system tables
Indexes are stored in the view sysindexes which is based on the system table 
sysindices
Indexes function are stored in the table sysindices in the column indexkeys 
which has the datatype indexkeyarray
This column stored the information like the following
(function id, Col1,...,coln).
The fields within the indexkeys columns have the following significance:
  The function id appears only if the index is on return values of a
function that is defined over the columns of the table. (That is, if the
index is a functional index.) Here the function id is the same as the
procid value for the function in the sysprocedures table.
  The list of the columns (col1, ... , coln) in the second field identifies the
columns over which the index is defined.
  The operator class id shows the secondary access method that is used
to build and to search the index. Here the operator class id is the same
as the opclassid of the access method in the sysopclasses table.

To access to this information, you have to use something similar to 

Select indexkeyarray_out(indexkeys) as my_info from sysindices where idxname = 
'my_i2'
The result will be similar to <177>(4)[1]
where 177 is the id of the function and 4 the column number


Select p.procname ||'(' ||c.colname||')'
From 
   informix.sysprocedures p, 
   informix.sysindices i,
   informix.systables t,
   informix.syscolumns c
Where
   t.tabid=i.tabid
   and c.tabid = t.tabid
   and c.colno = substr(indexkeyarray_out(i.indexkeys),7,1)   
   and p.procid = substr(indexkeyarray_out(i.indexkeys),2,3)
   and i.idxname = 'my_i2'

will return my_lower(c4)
 
========================================================================
 Issue: 446514    11/01/06  	Status: Closed	02/08/07  
Severity : 2         	Est. release: tyresias
Label  : BPEL4WS 1.1 : translation error unresolved member: isCompensate
Description :                                       
	 Steps to reproduce
New BPM target BPEL4WS 1.1
Add a process P1, decompose it
into the new sub-diagram, add a new process P11, a synchronization and a flow 
between the process and the synchronisation 
Flow --> properties --> Tab condition
Event : create a new one and properties in order to select a stereotype : Fault 
or compensation
Note : no problem when the stereotype is timer
Preview P11
result OK
Decompose P11
Preview
BPEL4WS 1.1::Profile\Flow\Criteria\isCatch\Templates\definition(3) (Flow 
"Flow_1"):
   [translation error] unresolved member: isCompensate

Check the xpl file and
BPEL4WS 1.1::Profile\Flow\Criteria\isCatch\Templates\definition
 .if (%Destination.isCompensate%)
is calling a template which doesn't exist when the destination is a 
synchronization
Under
BPEL4WS 1.1::Profile\Synchronization\Templates, there are only isJoin or isSplit

The problem will not occur when
P11 is not decomposed
or the flow is not a Fault or compensation
or the destination is not a synchronisation
 
========================================================================
 Issue: 446474    10/31/06  	Status: Closed	03/13/07  
Severity : 1         	Est. release: tyresias
Label  : Model Dependencies not listed after changing target model of a Data S
	 ource
Description :                            
	 Description of Problem: After saving AS NEW MODEL, and changing target 
model of a Data Source OOM Dependencies not listed 
OS:  Windows
Versions and Build #s Reported:  PowerDesigner 12.1.0.1924 EBF1
Versions and Build #s Tested:  PowerDesigner 12.1.0.1924 EBF2
Database/Driver File/Version:  
Comments:  
                                                                                                                                                                                                       
========================================================================
 Issue: 446401    10/31/06  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : FK column migrated from a shortcut badly regenerated
Description :                                                  
	 Tested and reproduced with PD 12.1 EBF#2

When applying vertical partitioning in a PDM table, eliminate the references and 
then generate it again from CDM with "preserve modifications" set, the 
references do not exist but their respective columns are improperly generated.

*** 09 November 2006 16:14:16 *** vaudino ***
change title for a more accurate one for future reference
new title: FK column migrated from a shortcut badly regenerated
old title: PDM Columns improperly generated after applying vertical partitioning 
into model 
========================================================================
 Issue: 446380    10/31/06  	Status: Closed	03/21/07  
Severity : 2         	Est. release: tyresias
Label  : Triggers are not previewed in Preview tab of table properties
Description :                                         
	 Tested and reproduced with PD 12.1 EBF#2

The trigger scripts for a table are not previewed in the Preview Tab.                                                                                                                                                                                                                                                                                                                                                                                                                
========================================================================
 Issue: 445864    10/25/06  	Status: Closed	02/08/07  
Severity : 3         	Est. release: tyresias
Label  : XML Model Mapping Editor: Unable to map table columns to XML elements
	 
Description :                                 
	 *** Wednesday, October 25, 2006 2:33:52 PM *** ahill *** 11.1.0 & 12.1.x

XML Model Mapping Editor: Unable to map table columns to XML elements

To reproduce:
1 Create PDM with a table and several columns

2 Generate from PDM an XML model using the XML generation option of "elements" 
in the Detail Tab

3 In the XML model bring up the Mapping Editor.

4 Set data source to your PDM source gen

5 Try to map table column to XML elements

The XML model only displays the 'database' element but no nested elements to map 
to. 
If I change the embedded type on the database element from 'complex' to 'none', 
then all the elements are listed in the mapping editor.


*** Wednesday, November 15, 2006 6:33:52 PM *** vaudino *** 
"complex" should not be added on database element during model generation
 
========================================================================
 Issue: 445008    10/18/06  	Status: Closed	04/08/07  
Severity : 3         	Est. release: tyresias
Label  : Informix : cluster designation  is not RE by ODBC
Description :                                                     
	 'clustered' designation for index is not  reversing by ODBC

Steps to reproduce

Generate the following script against an Informix 9 server


create table my_dbo.my_t  (
  column_1             integer                         not null,
  column_4             integer,
  column_2             integer,
  column_3             integer,
primary key (column_1)
      constraint pk_my_t
)
;

--==============================================================
-- Index: my_i
--==============================================================
create unique cluster index my_dbo.my_i on my_dbo.my_t (
        column_2             ASC,
        column_3             ASC
);


RE into a new PDM
Cluster information is not RE 
========================================================================
 Issue: 444259    10/11/06  	Status: Closed	04/27/07  
Severity : 2         	Est. release: tyresias
Label  : Erwin 4.x Import: Identity column datatype not imported for DB2 datab
	 ase model
Description :                        
	 *** Wednesday, October 11, 2006 12:00:36 PM *** ahill *** 11.1.0 & 12.1.0

\\neo\cir\...\444259

When importing an ERWin model with DB2 UDB 7 as the target DBMS, the identity 
columns are not imported properly. The column is imported, but is missing the 
identity option and the physical options, start with and increment by, values.

To reproduce:
Import Tony_test.XML or do the following

Create an ERWin 4.x model with target DBMS DB2 UDB 7 or 8.
Add a couple of tables with several columns.
On table_1, add an identity column with a start and increment value
Create a domain with a datatype that supports an identity column
On table_2, associate this domain to a column and make the column an identity 
column with a start and increment value.
save the model and import it into PD.


NOTE: The 'identity' information associated with the column is not imported.

In the customer's model, the column is named "WORKFORCE MEMBER IDENTIFIER" in 
the table 'LOADING WORKFORCE MEMBER PROPERTY'
 
========================================================================
 Issue: 443058    09/27/06  	Status: Closed	04/18/07  
Severity : 2         	Est. release: tyresias
Label  : PhysOpt-DB2 Index option Include Columns incorrectly list existing in
	 dex columns
Description :                      
	 *** Wednesday, September 27, 2006 3:00:21 PM *** ahill *** 12.1.0

Target DBMS: DB2 8 Common Server

PhysOpt-DB2-Index option Include Columns incorrectly list existing index 
columns. The columns not currently included in the index should be listed.

On Physical Options (Common) 
For option Include Column : clicking on button for list of columns lists the 
existing index columns. If these columns are selected, PD generates invalid DDL. 
When the DDL is executed, the following ODBC error occurs:

"column_xxx' is duplicate column name



This was implemented in CR 403752 PhysOpt-DB2-Index-Include Columns-Clicking on 
button does nothing 
========================================================================
 Issue: 442904    09/26/06  	Status: Closed	02/05/07  
Severity : 3         	Est. release: tyresias
Label  : Add support for DB2 9, UDB and OS/390
Description :                                                                 
	 *** Tuesday, September 26, 2006 12:17:03 PM *** ahill *** 12.1.0

Add support for DB2 9, UDB and OS/390                                                                                                                                                                                                                                                                                                                                                                                                                        
========================================================================
 Issue: 442622    09/22/06  	Status: Closed	03/21/07  
Severity : 3         	Est. release: tyresias
Label  : keeps creating temporary tables instead ofgenerating sp_rename
Description :                                        
	 Modify Database after changing the mandatory property of a column 
generates a temp table followed insert into statements instead of simple 
sp_rename although option 'Use alter statements when possible' has been enable

OS: Windows XP
Versions and Build #s Reported: PDes 12.0 EBF3 MSS 2000
Versions and Build #s Tested: PDes 12.1 EBF1 MSS 2000
Database/Driver File/Version: MSS 2000
Comments:
Step-By-Step Instructions:

1* Open sdehandler pdm_v7.pdm
 2* Go to Database > Modify Database...
 3* Select archive model named sdehandler pdm_v7.apm
 4* Check 'One file only' and select 'Use alter statements when possible' 
option. Click OK
 >> It detects no changes at all so the generated file is empty. You can use the 
customized XDB on ectsss20 or change the pk/fk name filter to get rid of those 
changes.
 5* Open the property sheet of Table e_event_document and make column id_lndr 
(#6) mandatory.
 6* Go to Database > Modify Database...
 >> It generates

create table tmp_e_event_document (
   ...)
go

set identity_insert tmp_e_event_document on
go

insert into tmp_e_event_document (...)
select ...
from e_event_document
go

set identity_insert tmp_e_event_document off
go

Changing the mandatory property column is neverthless not complex enough to 
force the creation of a temporary table.
If I start from scratch using a new blank model, I nor cust do not reproduce 
anymore with or without using internal shortcuts.
I simply get:
execute sp_rename TABLE_1, tmp_TABLE_1
go


The temporary table is caused by using domains. However done a couple of tests 
showing that sp_rename is accepted by MSS server with domains, so PDes should 
manage that syntax instead of CREATE and INSERT INTO statements. The following 
script does return no error at all.

execute sp_addtype Domain_1, 'integer'
go
execute sp_addtype Domain_2, 'integer'
go


create table Table_11281061 (
   COL_1            Domain_1       not null,
   COL_2            Domain_2       null,
   COL_3		 		 int		 null,
   constraint PK_Table_11281061 primary key (COL_1)
)
go

alter table Table_11281061
   drop constraint PK_Table_11281061
go

execute sp_rename Table_11281061, tmp_Table_11281061
go

create table Table_11281061 (
   COL_1             Domain_1                  not null,
   COL_2             Domain_2                  not null,
   COL_3             int                  not null,
   constraint PK_Table_11281061 primary key (COL_1)
)
go

insert into Table_11281061 (COL_1, COL_2, COL_3)
select COL_1, COL_2, COL_3
from tmp_Table_11281061
go

drop table tmp_Table_11281061
go

Test Case (zip file) Name: \\Neo\cir\CIRMails\44xxxx\442622\442622.zip
 
========================================================================
 Issue: 442017    09/15/06  	Status: Closed	10/30/06  
Severity : 2         	Est. release: tyresias
Label  : IBM DB2 MVS7 : fails to reverse piecesize value via ODBC
Description :                                              
	 IBM DB2 MVS7 : fails to reverse piecesize value via ODBC
Please update the XDB files 

Reverse a schema which has a table index with piecseize option set to a numeric 
value followed by " K"
For instance,

create type 2 unique index INDEX_1 on TABLE_1 (
   COL_1                 ASC,
   COL_2               ASC
)
piecesize 2097152 K;
" K" has been lost at reverse via ODBC.

I am able to reproduce 
the========================================================================
 Issue using the PB DataDirect Wire Protocol.
Removing the extra space char between the piecesize value and K

Steps to fix it:
 1* Go to Database > Edit Current DBMS...
 2* Select DB2MVS7::Script\\Objects\\Index\\SqlOptsQuery
 3* Remove the extra space char between the piecesize value and K
Previous value
   concat  case i.piecesize when 0 then '' else ' piecesize ' concat 
varchar(i.piecesize) concat ' K' end
New value
   concat  case i.piecesize when 0 then '' else ' piecesize ' concat 
varchar(i.piecesize) concat 'K' end


 
========================================================================
 Issue: 441880    09/13/06  	Status: Closed	01/29/07  
Severity : 4         	Est. release: tyresias
Label  : Change text of PD confirmation message for backwards compatibility
Description :                                    
	 *** Wednesday, September 13, 2006 4:09:45 PM *** ahill *** 12.0.0

Customer has both PD 11 and PD 12 installed.  They made the mistake of 
associating a v12 resource file with a v11 model.  In PD 11 when they open the 
model, a PD message is generated stating:

PowerDesigner - Confirmation:
   The model you wish to open has been saved with a more recent version 
(12.0.0.1700) than the one you are using (11.1.0.1604).
    It is strongly recommended you update your version of the product.
    If you choose to open the model with this version, the integrity of the 
model is not guaranteed in the case that is opens successfully.
    In addition, the model will be opened in read-only mode.


Seeing that the version of the model is v11, they customer feels that this 
dialogue is incorrect and misleading.  They suggest the following:

<customer comments>
"Please create a different dialog box if only a support file has been updated.  
Do NOT say that the model itself is updated.  Treat the situation similar to if 
the XDB file is missing. Offer the user the chance to select an alternate XDB.  
Be specific about which file is actually updated."
 
========================================================================
 Issue: 441849    09/13/06  	Status: Closed	04/08/07  
Severity : 3         	Est. release: tyresias
Label  : DB2 5 XDB: Index SqlAttrQuery where clause includes extra 'where' tex
	 t
Description :                                
	 *** Wednesday, September 13, 2006 10:51:24 AM *** ahill *** 9.5.1, 10.1, 
11.1, 12.1

Target DBMS:
DB2 5 OS/390

In the XDB, the SqlAttrQuery for the Index script object contains an extra 
'where' in the where clause.

<snip>

from
   sysibm.sysindexes 
where where uniquerule not in ('P', 'U')

<snip>                                                                                                                                                                                                        
========================================================================
 Issue: 441836    09/13/06  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : two tables with same names and distinct users are not reversed in cor
	 rect folder
Description :                      
	 
 *** mercredi 13 septembre 2006 11:39:16 *** mledier ***
Reverse the following script in PowerDesigner (I used ASA xdb)
<<
create table mledier.t(a integer)
;
create table psitja.t(a integer)
;
>>
In model, create two packages, one for each user and move the corresponding 
table in them.
Reverse again the script, the merge window shows move action, it should not.
OK the merge window, both tables are in the same package.
If the tables had distinct names, it would work correctly...                 
========================================================================
 Issue: 441799    09/12/06  	Status: Closed	11/16/06  
Severity : 3         	Est. release: tyresias
Label  : "Table Trigger Code Preview" node ignores 'generation settings' at pa
	 ckage level
Description :                      
	 *** Tuesday, September 12, 2006 4:49:29 PM *** ahill *** 12.1.0
\\neo\cir\...\441799

In a PDM with packages, the customer has created a generation setting to exclude 
the 'drop statement' for triggers.  In a report, the customer is including the 
'Table Trigger Code Preview' node at the model and package levels.  When they 
preview the report, the triggers at the model level correctly exclude the 'drop 
trigger' statement.  However, the triggers at the package level incorrectly have 
the 'drop trigger' statement included.

To reproduce:
1) open sample model
- includes 2 tables with references at the root and 2 tables with references in 
a package
- along with a report that includes the 'Table Trigger Code Preview' node at the 
model and package levels
2) generate or preview the report

 
========================================================================
 Issue: 441760    09/12/06  	Status: Closed	09/28/06  
Severity : 0         	Est. release: tyresias
Label  : PD12.1/OOM: attribute values can not be set for an object of a class.
	 
Description :                                 
	 In a object diagram, attribute values can not be set for an object of a 
class. In Object properties, Add Attributes Values option doesn't retrieve any 
attributes from the Class.

Failed in PD 12.1.1924
Okay in PD 11.x

To reproduce:
---------------------------------
-create a new oomodel 
-create a new class diagram 
-create a new class,  "customer", with "first name" and "last name" as 
attributes 
-create a new object diagram 
-place an object onto the diagram, edit properties: classifier "customer" -> 
Attributes Values tab -> Add Attribute Values -> empty list, nothing to add

issue is also reproduced in a Sequence diagram (follow the same steps), no 
attributes values can be assigned to an object with a classifier defined.



 
========================================================================
 Issue: 441657    09/11/06  	Status: Closed	11/24/06  
Severity : 2         	Est. release: tyresias
Label  : Custom form : text field do not word wrap
Description :                                                             
	  *** lundi 11 septembre 2006 12:01:20 *** vaudino ***12.1 GA
create a free model with a text extended attribute in a form
type a long text in it
the text does not word wrap
                                                                                                                                                                                                                                                                                                                                                
========================================================================
 Issue: 441309    09/06/06  	Status: Closed	10/26/06  
Severity : 2         	Est. release: tyresias
Label  : PD 12/Informix 9:mandatory option(if present) is removed using Modify
	  Database
Description :                        
	 PD12/Informix 9:property mandatory of a column is removed during Modify 
Database

PD12.1/Informix: If a column has a default value, and the table is reversed 
engineering, the default value is reversed. But in the next Modify Database, if 
there's an alter table,  it will be ignored and the script generated will remove 
it.

E: PD 12.1.0.1913

To reproduce:
--------------------------
1. Execute this script in an Informix database
create table test (a integer, b char(5) not null)
2. Reverse Engineering in a new pdm
3. The mandatory column has been RE correctly
4. In the pdm, add a default value for b
5. Modify Database -> option "Use Alter statements when possible" should be 
selected
6. Script generated is 
alter table test
   modify b char(5) default "a";

7. Run the script in the database, column is now a null column, not null 
property is lost.
 
========================================================================
 Issue: 441022    08/31/06  	Status: Closed	02/12/07  
Severity : 2         	Est. release: tyresias
Label  : Teradata reverse engineer via script fails for tables with the 'no lo
	 g' option
Description :                        
	 *** Thursday, August 31, 2006 4:40:15 PM *** ahill *** 12.1.0

PD is able to generate a table with the 'no log' option specified.  However, 
when you try to reverse engineer the table via script, all the attributes of the 
table are ignored.  If the 'no log' syntax is removed from the create table 
syntax, the table reverses correctly.

To Reproduce: 

1. Reverse Engineer the following script:

create GLOBAL TEMPORARY table DFLT_CAPN_GTT, no log (
    ETL_ROW_ID           DECIMAL(18)                    not null,
    MBR_KEY              VARCHAR(70),
    PKG_NBR              VARCHAR(6)
)
unique primary index (
   ETL_ROW_ID
)
on commit PRESERVE rows;

 
========================================================================
 Issue: 441017    08/31/06  	Status: Closed	02/12/07  
Severity : 3         	Est. release: tyresias
Label  : DB2 UDB 8.x Common Server xdb: Key\SqlListQuery is missing the 'order
	  by' clause
Description :                      
	 *** Thursday, August 31, 2006 4:00:03 PM *** ahill *** 12.1.0

Target DBMS: IBM DB2 UDB 8.x Common Server

When reverse engineering a database via ODBC, the customer is experiencing 
situations where the key column order is incorrect.

They are requesting that an 'order by' statement be added to the XDB file.
adding this to the "DB2UDB8::Script\Objects\Key\SqlListQuery" entry would 
guarentee the records for the Key column are reversed in the correct column 
order.

Customer comments:

If a key has more than one column, the SqlListQuery could return them in any 
order.  The MVS version of this query includes the line "order by 1, 2, 3, 4, 
c.colseq".  The colseq will force the key columns into the proper order.  By 
adding this line to the UDB query, it works, so there is a work-around.
Please update the shipping product to order the key columns.

If the columns come back in the right order, it is by coincidence.  In our 
database, they aren't coming back in order, because there is no order by 
statement.  You know databases, if you don't specify an order by, the records 
come back, but the order returned is not guaranteed.  Every query statement used 
by PowerDesigner MUST contain an order by clause.  
 
========================================================================
 Issue: 440999    08/31/06  	Status: Closed	03/31/07  
Severity : 2         	Est. release: tyresias
Label  : Rev ODBC ASE12: Dialog Error selecting Domain Tab due to varchar data
	 type size
Description :                        
	 *** Thursday, August 31, 2006 2:49:20 PM *** ahill *** 12.0.0 & 12.1.0

Target DBMS: ASE 12.5.x

When reverse engineering an ASE 12.5.3 or 12.5.4 database, selecting the Domains 
tab results in the following error:
'Unable to List the Domains: <ODBC Driver reference> Arithmetic overflow 
occurred.
sqlstate = 22003'

When the Domain\SqlListQuery is executed via Isql, the following error is 
returned:

'Arithmetic overflow during implicit conversion of INT value '1048' to a TINYINT 
field.'

In the database, there are two domains one varchar(1048) and the other 
varchar(2048).  The presence of these domains are causing the error to occur.

To resolve 
the========================================================================
 Issue, make the following change to the Domain\SqlListQuery:

Workaround: 
Edit Current DBMS for ASE 12.5.x:
Script > Objects > Domain
Edit the SqlListQuery
Add the Following line to the Case Statement
     when 39 then d.length

The Case Statement Now looks like the Following:

case(t.type)
      when 135 then d.length/@@unicharsize
      when 155 then d.length/@@unicharsize
      when 39 then d.length
   else isnull(d.prec, d.length)
   end,
 
========================================================================
 Issue: 440816    08/29/06  	Status: Closed	08/31/06  
Severity : 2         	Est. release: tyresias
Label  : DB & Modify generation: Synonyms created before Base Object
Description :                                           
	 *** Tuesday, August 29, 2006 4:48:35 PM *** ahill *** 11.1.0 & 12.1.0
\\neo\cir\...\440816

This========================================================================
 Issue appears to be related to CR 400470.

Target DBMS: DB2 7 OS/390 and DB8 OS/390

In v11.1, Modify Database generates the create synonym DDL before the base 
object, in this case tables. In v12, DB generation and modify DB generates the 
create synonym DDL before the base object, in this case tables.

In testing, the problem appears to be due to the dependency between tablespaces 
and databases within the model. On the tablespaces, the database physical option 
is specified.  On each table, the tablespace physical option is specified. This 
results in the tablespace physical otpion being generated as 
<database>.<tablespace> on the table physical option.  When this occurs, the 
synonym referencing the table as a base object is created before the table.


To reproduce:

1)Customer's .pdm and .apm pair can be used to repro in v11

or 

2) use syn_before_base_obj.pdm to see the problem in v12

 
========================================================================
 Issue: 440798    08/29/06  	Status: Closed	08/31/06  
Severity : 2         	Est. release: tyresias
Label  : Selecting Force Column List option for view does not generate column 
	 list
Description :                             
	 *** Tuesday, August 29, 2006 3:16:08 PM *** ahill *** 12.0.0 EBF4 & 
12.1.0 GA

Selecting the "Force column list" in the database generation for views doesn't 
generate the columns for the view.  This worked in v11.1.0.

To reproduce

1) Create a view with several columns  (tested using ASA 10, Oracle 9i, Oracle 
10g as target DBMS)
2) Select the 'Force Column List' generation option 
3) Preview the view                                                                                                     
========================================================================
 Issue: 440792    08/29/06  	Status: Closed	08/31/06  
Severity : 1         	Est. release: tyresias
Label  : Modify Database: selecting the 'Options' tab closes the PD applicatio
	 n
Description :                                
	 *** Tuesday, August 29, 2006 2:27:00 PM *** ahill *** 12.1.0.1913
\\neo\cir\...\440792

Target DBMS: Informix 9

PowerDesigner crashes when the 'Options' tab is selected on the Modify Database 
property window.

To reproduce:

Using the customer's model, if you select Modify Database and then select the 
'Options' tab, PD immediately shuts down. (Customer's model was migrated from 
12.0.0 EBF4.)

Using a sample model (repro.pdm), select Modify Database and the select the 
'Options' tab.  On the Options tab, click the 'show all objects' tool. PD will 
then shut down.

repro steps for sample model:
1) Create a new PDM ( Informix 9  & SQL server 2005 were selected as the target 
DBMS)
2) Add a table with several columns
3) Select Modify Database
4) Select the Options tab
5) On the Options tab, click the 'show all objects' tool. 
Note: PD will then shut down.

*** Tuesday, August 29, 2006 2:40:53 PM *** ahill ***
Customer is requesting an EBF for 
this========================================================================
 Issue. 
========================================================================
 Issue: 440776    08/29/06  	Status: Closed	09/11/06  
Severity : 2         	Est. release: tyresias
Label  : Print preview in "generate report" dialog delete the report
Description :                                           
	  *** mardi 29 aot 2006 17:35:40 *** vaudino ***12.1 GA
in any model
create a report
launch Report > Generate report
select the report and select Print preview
close the print preview
the report has been deleted

                                                                                                                                                                                                                                                                                                    
========================================================================
 Issue: 440647    08/28/06  	Status: Closed	09/07/06  
Severity : 2         	Est. release: tyresias
Label  : VSNET-Close ModelExplorer-Close solution-Model Explorer brings up its
	 elf
Description :                              
	 *** Monday, August 28, 2006 14:11:58 *** jnsane ***1913
Object : PD Visual Studio - Model Explorer - Closing Solution

1-Open VS2005 within PD plug-in
2-Create a new project e.g Visual C"->Windows Application
3-Display Model Explorer View
4-Close Model Explorer View
5-Run "File->>Close Solution"
-->Get that the solution is closed
-->The Model explorer is displayed : it should not.

PS : works if we :
    close solution, the model explorer, close visual studio, reopen VS2005.                      
========================================================================
 Issue: 440440    08/24/06  	Status: Closed	04/08/07  
Severity : 3         	Est. release: tyresias
Label  : Informix : functions are not RE by ODBC
Description :                                                               
	 DBMS : Informix
Functions are not RE by ODBC, only stored proc are RE

Steps to reproduce

Against informix execute the following script
CREATE FUNCTION update_tdefault ( pct INT, pid CHAR(10))
RETURNING INT
RETURN pct;
END FUNCTION

RE into PD using ODBC connection
Select the function listed in the odbc windows
Inside PD, check th functionjsut RE, Defintion has not been RE

Checked Informix doc
IBM Informix Guide to SQL Reference
System catalog SYSPROCEDURE
Page 1-59
Parameter isproc char(1) Whether routine is a procedure or a function
t = procedure
f = function

 
========================================================================
 Issue: 440335    08/23/06  	Status: Closed	09/01/06  
Severity : 2         	Est. release: tyresias
Label  : Unwanted Ellipses displayed in entity compartent in IDEFIX mode
Description :                                       
	  *** mercredi 23 aot 2006 17:50:29 *** vaudino ***12.1 GA
create a CDM/PDM with IDEF1X notation
create an enioty with PK attributes, not PK attributes and identifier
you have "..." at the end of each comparment that were not here before
                                                                                                                                                                                                                                                                               
========================================================================
 Issue: 440301    08/23/06  	Status: Closed	04/08/07  
Severity : 3         	Est. release: tyresias
Label  : Informix : Datatype CHAR(X) is RE by ODBC without parenthesis
Description :                                         
	 DBMS : Informix 
Parenthesis is missed when RE by ODBC datatype Char(value)

Steps to reproduce

Against Informix database execute the following script
Create Table T1(C1 CHAR(5));

RE by ODBC into a new PDM using PD Datadirect wire protocol
result is CHAR5
                                                                                                                                                                                                                                                     
========================================================================
 Issue: 440212    08/22/06  	Status: Closed	11/16/06  
Severity : 3         	Est. release: tyresias
Label  : Informix modify database : column check constraints are not created
Description :                                   
	 DBMS : Informix
When generate a script to modify database with option Backup table and "Use 
alter statement when  possible" checked, Column check constraints are not re-
created.

Steps to reproduce
New PDM Informix 9
New table T1, C1 PK C2
C2 properties --> Check constraint
Min : 0
Mx : 10
Archive the model
Modify the mdel in order to have a drop/create table in the modify database 
script
Add a new cluster unique index based on C2
Generate a script to modify database
Backup table checked
Use Alter statement when possible : checked
Got the following script
alter table T1
   drop constraint PK_T1;

alter table T1 drop constraint CKC_C2_T1;

drop table tmp_T1;

rename table T1 to tmp_T1;

--==============================================================
-- Table : T1
--==============================================================

create table T1  (
  C1                   INT                             not null,
  C2                   INT,
primary key (C1)
      constraint PK_T1
);

insert into T1 (C1, C2)
select C1, C2
from tmp_T1;

alter table T1
   modify C2 INT;

--==============================================================
-- Index : INDEX_1
--==============================================================
create unique cluster index INDEX_1 on T1 (
        C2                   ASC
);

The constraint has be dropped but again here following cr  440186 done for ASA, 
ModifyColnChck is applied instead to AddColnChck with this time a syntax error , 
have a look to the Alter table
alter table T1
   modify C2 INT;

I checked the informix 9 doc
IBM Informix Guide to SQL Syntax 
Alter table
Modify clause page 2-65
MODIFY Clause
Use the MODIFY clause to change the data type, length, or default value of a
column, or to allow or disallow NULL values in a column.
Is it allowed to modify a column check constraint ?
 
========================================================================
 Issue: 440186    08/22/06  	Status: Closed	09/01/06  
Severity : 2         	Est. release: tyresias
Label  : Modify database : modify constraint is used instead of add constraint
	 
Description :                                 
	 DBMS : ASA
When modify database script without any modification on the column constraint ; 
modify column constraint syntax is used instead of add column constraint.

Steps to reproduce
Modify the xdb file following the cr 440185
New PDM ASA
1 Table T1, C1 PK, C2
C2 check constraint Min 1, max 10
Archive the model
Modifiy the model in order to get a create table
Insert a new column between C1 and C2
Modify database script
Backup tables : checked
Use ALter table statements when possible
result
alter table T1
   delete primary key;

alter table T1
   drop constraint CKC_C2_T1;

if exists(
   select 1 from sys.systable 
   where table_name='tmp_T1'
     and table_type in ('BASE', 'GBL TEMP')
) then
    drop table tmp_T1
end if;

alter table T1 rename tmp_T1;

/*==============================================================*/
/* Table: T1                                                    */
/*==============================================================*/
create table T1 
(
    C1                   char                           not null,
    C3                   char,
    C2                   numeric(10)                    default autoincrement,
    constraint PK_T1 primary key clustered (C1)
);

insert into T1 (C1, C2)
select C1, C2
from tmp_T1;

alter table T1
   modify C2 constraint CKC_C2_T1 check (C2 is null or (C2 between 1 and 10));

Have a look at the last Alter table, we got here a modify C2 instead of  Alter 
C2 add Constraint ...
So  even if Sybase AS Anywhere 9::Script\Objects\Column\AddColnChck has been 
modified,  PD used Sybase AS Anywhere 9::Script\Objects\Column\ModifyColnChck 
which is wrong.
ModifyColnChck should be only used when the constraint has been modified 
========================================================================
 Issue: 440185    08/22/06  	Status: Closed	04/09/07  
Severity : 3         	Est. release: tyresias
Label  : Gene script Columns with Constraint outside table, wrong syntax with 
	 DBMS  ASA
Description :                        
	 Wrong syntax is generated in database generate database for Column with 
constraint outside table
Steps to reproduce
New PDM : ASA
1 table T1, C1 PK C2 
C2 properties check constraint
Min :1 , max 10, autoincrement

Preview with option column constraint outside
create table T1 
(
    C1                   char                           not null,
    C2                   numeric(10),
    constraint PK_T1 primary key clustered (C1)
);

alter table T1
   modify C2 constraint CKC_C2_T1 check (C2 is null or (C2 between 1 and 10));


Which is wrong, as the check constraint doesn't exist, an error will occur if 
you execute this script

syntax should be
create table T1 
(
    C1                   char                           not null,
    C3                   char,
    C2                   numeric(10)                    default autoincrement,
    constraint PK_T1 primary key clustered (C1)
);

alter table T1
   alter  C2  add constraint CKC_C2_T1 check (C2 is null or (C2 between 1 and 
10));

ASA documentation
ALTER TABLE [ owner.]table-name
{ add-clause | modify-clause | drop-clause | rename-clause }

add-clause :
   ADD { column-definition | table-constraint }
| { ADD PCTFREE integer | PCTFREE DEFAULT }

modify-clause :
  MODIFY column-definition
| MODIFY column-name { DEFAULT default-value
   | [ NOT ] NULL
   | [ CONSTRAINT constraint-name ]
         CHECK { NULL |  ( new-condition ) } }
| ALTER column-name  column-modification
| ALTER constraint-name CHECK ( new-condition )

with
column-modification :
  SET DEFAULT default-value
| DROP DEFAULT
| ADD [ CONSTRAINT column-constraint-name ] CHECK ( condition )
| { DELETE | DROP } CONSTRAINT column-constraint-name
| { DELETE | DROP } CHECK
| SET COMPUTE ( expression )
| DROP COMPUTE
 
========================================================================
 Issue: 440177    08/21/06  	Status: Closed	10/12/06  
Severity : 1         	Est. release: tyresias
Label  : Replica object name problem when there is  colons in CDM Code
Description :                                         
	 Description of Problem: Replica object full name cut off at the colon 
when the data item code contains colon.
OS:  Windows
Versions and Build #s Reported: 12.0.0.1725 EBF 2  
Versions and Build #s Tested:  12.1.0.1913
Database/Driver File/Version:  
Comments:  
                                                                                                                                                                                                                                                     
========================================================================
 Issue: 440172    08/21/06  	Status: Closed	09/01/06  
Severity : 1         	Est. release: tyresias
Label  : Select object doesnt show complex type either from same model or ano
	 ther model
Description :                       
	 Description of Problem: 
Models with complex type, simple Content, and restriction in the Derivation 
field. 
1.Restriction properties base type drop down isnt showing othe models simple 
type nor complex type
2. Restriction properties select object only shows simple type of the models.

OS:  Windows XP
Versions and Build #s Reported:  12.1.0.1913
Versions and Build #s Tested:  12.1.0.1913
Database/Driver File/Version:  
Comments:  Related to CR 436213
                                          
========================================================================
 Issue: 440009    08/18/06  	Status: Closed	11/16/06  
Severity : 3         	Est. release: tyresias
Label  : VBScript : GenerationCharacterCase  Possible values don't follow the 
	 docume
Description :                           
	 According to the VBScript help, syntax for GenerationCharacterCase is 
=====================
Format option that specifies the case of objects in the script. Possible values 
are : lower case(0), upper case(1), mixed case(2).
======================

Those values are wrong from 12.1 version
In PD 12
Lower : 0
Upper : 1
Mixed : 2
As described in the documentation
From 12.1, values are
Mixed : 0
Upper : 1
Lower : 2                                                                                      
========================================================================
 Issue: 438927    08/04/06  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : Format of constraint expression for columns against informix is wrong
	 
Description :                                 
	 -----------------------  
PSE NOTES:  
-----------------------  

PDes 12.0/Informix 9.4
be able to customize format of constraint expression for columns
For instance, PDes generates: ( col1 is null or col1 in ('A', 'B') )
Cust needs ((col1 IS NULL) or (col1 in ('A', 'B'))

When At generation, I obtain:
     (COL_1 is null or (COL_1 in ("A","B")))

Doing Edit the current DBMS and select
INFORMIX9::Script\\Objects\\Column\\DefineColnChck

the value is
[check (%.A:CONSTRAINT%)
      [constraint [%R%?[%QUALIFIER%]]%CONSTNAME%]]

Unfortunately, %CONSTRAINT% is not customizable because hardcoded.
So we would a need a new feature.

But there is another 
related========================================================================
 Issue here regarding the script at reverse...
Informix stores it as :
     ((COL1 IS NULL) or (COL1 in ('A','B')))
It puts these extra parenthesis around the first check block.
When the database is reverse engineered, these constrainsts are not properly 
parsed so it is not displayed at the right place in PDes interface.
The preview looks fine but the 2nd check argument has been moved to the 
Additional Checks tab as a server expression:
(COL_1 is null) or (COL_1 in ("A","B")) and %MINMAX% and %LISTVAL% and %UPPER% 
and %LOWER% and %RULES%

OS:  XP  
Versions and Build #s Reported:  PD12 
Versions and Build #s Tested: PD 12, 12.1 
Database/Driver File/Version:  Informix.  
Comments:    
 
 


 
========================================================================
 Issue: 438809    08/03/06  	Status: Closed	08/30/06  
Severity : 2         	Est. release: tyresias
Label  : Cannot modify database because the generation is aborted
Description :                                              
	 ----------------------- 
PSE NOTES: 
----------------------- 
Description of Problem:    
Every time the physical model is generated from the conceptual model, the check 
model delete the tables, why this is happening?
Index seems to be corrupted, unchecking the Index Inclusion item in Check Model 
option, it works
Ct wants to know also why his model became corrupted
Merging the generated physical model is not an option, since the first time we 
did it that worked, and after a while, model had problems again. HE doesn't want 
to work with an unstable model. Ct wants that we investigate further the problem 
to be sure that the problem is the model and not PD 12.
Could the chkmodel.xem be involved ? It had been modified for a previous case, 
related to unique contraint names. (see case 11203245).

The log shows:
0 errors, several warnings
Error: Generation aborted due to errors detected during the verification of the 
model.

OS:  XP 
Versions and Build #s Reported:  PD12
Versions and Build #s Tested: PD 12, 12.1
Database/Driver File/Version:  Not db specific. 
Comments:   
 
========================================================================
 Issue: 438742    08/02/06  	Status: Closed	08/30/06  
Severity : 2         	Est. release: tyresias
Label  : PK constraint name truncated to 8 characters during consolidation
Description :                                     
	 *** Wednesday, August 02, 2006 3:28:37 PM *** ahill *** 10.1.0 EBF12 & 
12.1.0
\\neo\cir\...\438742

target DBMS = IBM UDB 7 OS/390

Primary key constraint name truncated to 8 characters during consolidation into 
an existing model.

When a model is consolidated for the first time, the PK constraint column is 
consolidated correctly. After extracting and making a change to the PK 
contstraint, subsequent consolidations truncates the constraint to 8 characters. 
However, the maximum length set in the DBMS file for a constraint is 18.

To reproduce:
target DBMS: IBM UDB 7 OS/390
repro.pdm can be used for 
this========================================================================
 Issue.

1) Add a table, human resources, with several colums
2) Create a PK with the name 'HmnRsrcs ID'
3) Consolidate and extract the model 
4) Modify the PK constraint name changing it to 'P_HMNRSRCS_ID12345'
       4a.[execute check model. note: no errors are generated]
5) Consolidate and extract the model using the update option
Note: the results in the General tab

    - Updating the repository document...
	Modify Key "HmnRsrcs ID" in the Table "Human Resources" (Constraint Name)
	    "P_HMNRSRCS_ID12345" has been renamed to "P_H12345"

The work around for 
this========================================================================
 Issue is to copy the DBMS into the model instead of using a shared DBMS file.

 
========================================================================
 Issue: 438704    08/02/06  	Status: Closed	08/30/06  
Severity : 1         	Est. release: tyresias
Label  : Java sample not working when executed from Eclipse 3.1.1
Description :                                              
	 ----------------------- 
PSE NOTES: 
----------------------- 
Description of Problem:    
Java sample CreateOOM.java ends up with error when executed from Eclipse 3.1.1
The debugger shows that it stops at CreateOOM.CreateAndDisplayOOM() at line 47.
Error in getInstance() has raised:
"Cannot create PowerDesigner application object. Please verify that 
PowerDesigner 12.1 or higher is installed."

OS:  XP 
Versions and Build #s Reported:  PD12
Versions and Build #s Tested: PD 12, 12.1
Database/Driver File/Version:  Not db specific. 
Comments:   
 
========================================================================
 Issue: 438493    07/31/06  	Status: Closed	03/12/07  
Severity : 1         	Est. release: tyresias
Label  : Repository update from 11 to 12 fails with error when installed on AS
	 E
Description :                                
	 *** Monday, July 31, 2006 10:21:14 AM *** ahill *** 12.0.0 & 12.1.0
\\neo\cir\...\438493

Repository update from 11 to 12 fails with error on Sybase ASE
The customer is upgrading from PD 11.0.0 <no ebfs> to 12.0.0.
During the upgrade, they get the following error message:

Msg 207, Level 16, State 4 Server 'SYBDEV32', Line 1 Invalid column name 'OTYP'. 

In testing, 
this========================================================================
 Issue has been reproduced when the repository is installed on ASE.
For my tests, I used ASE 12.5.4.

To reproduce:
Repository DBMS = ASE 12.5.4

1) Create a repository on PD 10.1.0 EBF12
2) Updated the repository to 11.0.0 GA
3) Updated the repository to 12.1.0 GA
4) In the generation log, search for ODBC, it will bring you to the error.

*** Executing statement 2429:
update PMPDMVIEW set TTYP=(select pt.TDAT from PMTEXT pt where pt.TXID=OTYP)

[DataDirect][ODBC Sybase Wire Protocol driver][SQL Server]Invalid column name 
'OTYP'.

SQLSTATE = S0022


NOTE:
In testing, the same update statement is generated for a repository on ASA.
During the update of the repository, this error is not generated.  However, if 
the select statement is run via isql, an error is generated stating that the 
column 'OTYP' does not exist.

 
========================================================================
 Issue: 438237    07/27/06  	Status: Closed	04/10/07  
Severity : 2         	Est. release: tyresias
Label  : Column check constraints reversed incorrectly for Informix by ODBC
Description :                                    
	 PD 12 and PD 12.1
Informix 9
I know than cr 273671 already exists and has been closed as not a bug but this 
new customer gave us some SQL to RE contraintes on the columns.

Steps to reproduce
Against a server informix 9, execute the following script
create table T1_C  (
  TC11                 CHAR                            not null,
  TC12                 CHAR,
  TC13                 CHAR
);

alter table T1_C
   add constraint check (TC12 in ("J","N"))
    Constraint CKC_TC12_T1_C;

alter table T1_C
   add constraint  check (TC13 in ("J","N"))
   constraint CKC_TC13_T1_C;

create unique  index TABLE_1_PK on T1_C (
        TC11                 ASC
);

alter table T1_C
   add constraint primary key (TC11)
      constraint PK_T1_C;



For sure the constraint look like it was constraints on  tables.
From PD, RE by ODBC
Constraint will be RE for table and not anymore for columns

Against a server informix 9, execute the following SQL

Select c.constrname, t.tabname
from
   "informix".sysconstraints c,
   "informix".systables t
where 
   c.tabid = t.tabid
   and c.constrname = "ckc_tc12_t1_c"

Got the table name

Execute now the following sql (sent by ct)
select  t.tabname, c.colname, s.constrname, d.constrid
from 
syscolumns c, 
systables t, 
sysconstraints s, 
syscoldepend d 
where s.constrid = d.constrid
 and  d.colno = c.colno
 and  d.tabid = c.tabid
 and  c.tabid = t.tabid
 and s.constrtype = 'C'  
 and 1 = (select count(*) from syscoldepend d2 where s.constrid =
d2.constrid)

Here we got the table name, the column name and the constraint name

So according to ct
This one gets all simple one column check constraints:

select  t.tabname, c.colname, s.constrname, d.constrid
from 
syscolumns c, 
systables t, 
sysconstraints s, 
syscoldepend d 
where s.constrid = d.constrid
 and  d.colno = c.colno
 and  d.tabid = c.tabid
 and  c.tabid = t.tabid
 and s.constrtype = 'C'  
 and 1 = (select count(*) from syscoldepend d2 where s.constrid =
d2.constrid)

This one would get multi-column-check-constraints:

select  t.tabname, c.colname, s.constrname, d.constrid
from 
syscolumns c, 
systables t, 
sysconstraints s, 
syscoldepend d 
where s.constrid = d.constrid
 and  d.colno = c.colno
 and  d.tabid = c.tabid
 and  c.tabid = t.tabid
 and s.constrtype = 'C'
 and 1 < (select count(*) from syscoldepend d2 where s.constrid =
d2.constrid)

This one would get table constraints:

select  t.tabname, s.constrname
from 
systables t, 
sysconstraints s 
where s.tabid = t.tabid
 and s.constrtype = 'T'


Due to this error, when doing a modify database, changes have been detected on 
the constraints and we should not get some changes
Note : When doing column --> Properties --> Standard check, add some values
Generation options, column --> Constraints outside
preview the script

create table TABLE_1  (
  COLUMN_1             CHAR,
  COLUMN_2             CHAR
);

alter table TABLE_1
   add COLUMN_1 CHAR check (COLUMN_1 is null or (COLUMN_1 in ("J","N")))
   constraint CKC_COLUMN_1_TABLE;

Columns are created twice, once in the create table and once in the alter table 
========================================================================
 Issue: 438229    07/27/06  	Status: Closed	02/14/07  
Severity : 3         	Est. release: tyresias
Label  : Deferred option of FK constraint is not RE for some references
Description :                                        
	 PD 12 and PD 12.1
XDB : Oracle 10G

RE into a new PDM the following script (Note the brackets for each constraint.)
create table test1 (
	a varchar(20) not null,
	b varchar(20),
	  CONSTRAINT test1_pk PRIMARY KEY (a))
/

create table test2 (
	c varchar(20) not null,
	a varchar(20) not null,
	  CONSTRAINT test2_pk PRIMARY KEY (c))
/
create table test3 (
	d varchar(20) not null,
	a varchar(20),
	  CONSTRAINT test3_pk PRIMARY KEY (d))
/

CREATE INDEX test2_IDX02 ON test2
(a)
LOGGING
NOPARALLEL
/

CREATE INDEX test3_IDX02 ON test3
(a)
LOGGING
NOPARALLEL
/

	
alter table test2 add (CONSTRAINT TEST2_TEST1_FK FOREIGN KEY (A) 
    REFERENCES TEST1 (A)
    initially immediate deferrable)
/

alter table test3 add (CONSTRAINT TEST3_TEST1_FK FOREIGN KEY (A) 
    REFERENCES TEST1 (A)
    ON DELETE CASCADE initially immediate deferrable)
/




Go to check each reference RE

Reference properties --> Tab Oracle
only one has Defered option completed

 
========================================================================
 Issue: 437524    07/20/06  	Status: Closed	03/12/07  
Severity : 3         	Est. release: tyresias
Label  : can not open model when user has inappropriate right on attached XEM.
	 
Description :                                 
	 PD 11.1, PD 12 and PD 12.1 beta  broken

When user has inappropriate right on attached XEM, instead to have no access to 
the xem and get a warning message, you can't open any models even if the user 
has full control on other xem and after that you can not create any new model
In Debug mode, I got
Src\sbbrowser.cpp : Browser view Double click and nothing else.
In output
Error: Could not find <o:FormTargetItem> in DTD or matching class, the contents 
of all <o:FormTargetItem> markups will be ignored!


Steps to reproduce
You should be logged as Administrator or your machine
Create a new folder : My_folder
New Model, add 2 xems, xem_1 and xem_2
save model and xems into my_folder
exit PD
Via Tools --> Ressources, modify xem path in order to  go by default on 
my_folder
On your machine, create a new user
From explorer, 
Right click XEM_1 --> properties --> Security tab --> Button advanced
Unchecked inherit from parent the permission  and click remove
Click the button Add
Click the button Location and select your machine
In Enter the object name location, type my_user and click check name (to be 
sure)
Select full control for this user and nothing else
You has to be sure than only this user can access to this xem
XEM_2, do the same thing but this time, only the admin will have the full 
control on the xem
To resume
XEM_1 : full control for my_user
XEM_2 : full control for admin
be sure that "everyone" user has no permisssion
You should be logged as admin on your machine
Run PD, open the pdm saved before

In debug mode, got error message
Src\sbbrowser.cpp : Browser view Double click and nothing else.
In output
Error: Could not find <o:FormTargetItem> in DTD or matching class, the contents 
of all <o:FormTargetItem> markups will be ignored!
The model is not opened, you can not create new model
Modify the path for the xem has fixed the problem
 
========================================================================
 Issue: 436952    07/13/06  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : Physical Options: unable to delete options from the Items tab
Description :                                         
	 *** Thursday, July 13, 2006 11:36:25 PM *** ahill *** 12.1.0.1908

Physical Options: Unable to delete options from the Items tab

To reproduce:
Target DBMS: MS SQL Server 2005

1) Add a table with several columns and create an index
2) Goto the Physical Options tab of the index
3) Add the 'on' option branch to the Items tab
4) In the Items tab, expand the options and remove the 'partition scheme' 
option.  
5) Apply the changes and close the table properties window.
6) Open the Index properties and go to physical options
Note: The 'partition scheme' option is still listed on the Items tab.

If you delete the column option and then delete the partition scheme option, the 
options are removed, but '?' place holders for the options still appear in the 
SQL.
To remove them, you must define the filegroup option a second time.  
========================================================================
 Issue: 436915    07/13/06  	Status: Closed	11/16/06  
Severity : 3         	Est. release: tyresias
Label  : Physical Options: From List of... unable to select/deselect options w
	 /check box
Description :                       
	 *** Thursday, July 13, 2006 5:08:36 PM *** ahill *** 12.1.0.1908
\\neo\cir\...\436915

Physical Options: From List of... unable to select/deselect options with check 
box.  Options with ddlb or sle are able to be set.

Via the List of Indexes, the customer is trying to modify physical options.
Specifically, they were trying to change pad_index, fillfactor, and filegroup. 
They are able to change fillfactor and filegroup, but when trying to select 
pad_index or any option with a check box, the change is not allowed.

To reproduce:
1) Open sample model
2) Select List of Indexes
3) Try to select any physical option that has a check box 
========================================================================
 Issue: 436700    07/12/06  	Status: Closed	09/01/06  
Severity : 2         	Est. release: tyresias
Label  : Add column comment syntax fails with error for MS SQL Server 2005
Description :                                     
	 *** Wednesday, July 12, 2006 11:56:06 AM *** ahill *** 12.1.0.1907
\\neo\cir\...\436694

Add column comment syntax fails with error for MS SQL Server 2005

*** Executing statement 15:
execute sp_addextendedproperty 'MS_Description', 
   'Use this field for storing RangeOfValue, Multiple Choice, and PriorityList 
answers.',
   'user', '', 'table', 'ANSWER', 'column', 'UserAnswerText'

*** Execution of statement failed:
[Microsoft][ODBC SQL Server Driver][SQL Server]Object is invalid. Extended 
properties are not permitted on 'ANSWER.UserAnswerText', or the object does not 
exist.
SQLSTATE = 37000


To reproduce:
Generate customer's model to the database.
Comments are located on the 'Answer' and 'ObjectClass' tables 
========================================================================
 Issue: 436694    07/12/06  	Status: Closed	09/21/06  
Severity : 2         	Est. release: tyresias
Label  : Create Index syntax w/options invalid for MS SQL Svr 2005
Description :                                             
	 *** Wednesday, July 12, 2006 11:43:02 AM *** ahill *** 12.1.0 build 1907

\\neo\cir\...\436694

Syntax generated for 'create index' is invalid for MS SQL Server 2005 when 
physical options are included.

PD generates the following syntax:

create index REFERENCE_2_FK on ANSWEROBJECTCLASS (
AnswerID ASC
)
with
    fillfactor= 80,
    pad_index= on
on "DEFAULT"

When this is executed via ODBC, the following error occurs:

[Microsoft][ODBC SQL Server Driver][SQL Server]Statement(s) could not be 
prepared.
SQLSTATE = 37000

To correct 
this========================================================================
 Issue, the physical options must be placed in parenthesis


create index REFERENCE_2_FK on ANSWEROBJECTCLASS (
AnswerID ASC
)
with
    (fillfactor= 80,
    pad_index= on)
on "DEFAULT"

Comment from customer:

In addition, there's a note in the SQL Server 2005 Documentation about backward 
compatibility in the relational index, as follows:   
"Important   The backward compatible relational index syntax structure will be 
removed in a future version of SQL Server. Avoid using this syntax structure in 
new development work, and plan to modify applications that currently use the 
feature. Use the syntax structure specified in <relational_index_option> 
instead."  
========================================================================
 Issue: 435073    06/26/06  	Status: Closed	08/29/06  
Severity : 4         	Est. release: tyresias
Label  : Eng Consult: set NO4 datatype translation to smallint identity specif
	 ically
Description :                           
	 Tested and reproduced in PD 12.0

We have a customer that migrated from PD 6.x to PD 12.0 EBF#2 and is using MS 
SQL Server 2000 as DBMS.

In old PD 6.0, when customer wanted to tell PowerDesigner CDMs that NO4 
datatype, which is serial(4), should be converted into smallint identity when 
creating a PDM from this CDM, he entered the DBMS .def file and added the 
following line under the data type translation table: 


NO4  = smallint  <---- added this line
NO      = numeric
NO%n    = numeric(%n)

After that, when creating PDM from CDM, automatically PD transformed specifics 
NO4 to smallint identity. 

We tried to do the same in PD 12.0 but apparently we couldn't get the same 
behavior. We opened PD 12.0 and edited MS SQL Server DBMS at 
MSSQLSRV2000::Script\DataType\AmcdDataType

Then. we inserted a NO4 and smallint in respective columns. But when generating 
PDM, PD still assumes numeric(%n) by reading the NO%n column. As a result a 
numeric(%n) identity column is created in PDM.

Is there a respective way to have PD 12.0 do the same as PD 6.0? Or is there 
other ways to tell PD 12.0 to create "smallint identity" specifically when 
encountering NO4 from CDM to PDM translation?

Thanks in advance for your attention. 
========================================================================
 Issue: 433194    06/06/06  	Status: Closed	01/18/07  
Severity : 3         	Est. release: tyresias
Label  : Add import support for the new file format from AllFusion ERwin Data 
	 Modeler r7
Description :                       
	 *** Tuesday, June 06, 2006 3:19:58 PM *** ahill *** 12.0.0 EBF3

Add import support for the new file format from AllFusion ERwin Data Modeler r7.                                                                                                                                                                                                                                                                                                                                                                              
========================================================================
 Issue: 430400    05/09/06  	Status: Closed	02/09/07  
Severity : 3         	Est. release: tyresias
Label  : Support of Eclipse 3.2
Description :                                                                                
	 Customer is requesting to have an Eclipse 3.2 plugin from end 2006                                                                                                                                                                                                                                                                                                                                                                                                                                                               
========================================================================
 Issue: 429882    05/02/06  	Status: Closed	04/25/07  
Severity : 3         	Est. release: tyresias
Label  : Changes to "Generate" option on references ignored during PDM to PDM 
	 generation
Description :                       
	 *** Tuesday, May 02, 2006 3:43:59 PM *** ahill *** 10.1.0 EBF12, 11.1.0 
EBf7, 12.0.0, 12.1.0 1841

Changes to the 'Generate' option is ignored on references when generating PDM to 
PDM using the Preserve Modification option. The customer reproted the problem 
<logical> to Oracle, but the DBMS type does not seem to matter.

customer description of the problem:

Between the <Logical> and a generated DBMS-specific PDM, if the only change you 
make is to the "Generate" checkbox on a reference, PD just doesn't see that 
difference. If you change other things in a reference, those things are seen, 
but "Generate" isn't in the Merge. You have to do extreme things like delete the 
reference in the target model to force the "Generate" checkbox to be migrated.
This bug exists in all versions from V10.

how to reproduce:
create a logical PDM with two tables with a primary key for each.
create a reference between the two tables.
generate a PDM for oracle.
the PDM is generated.

now go back to the logical PDM
open the reference properties panel.
uncheck the "generate" checkbox in the general tab.
merge the change to oracle

result:
the merge window does not show any change
and then nothing is merged.

now delete the reference from the oracle PDM
go back to the logical PDM and do a merge.
check in the merge window that the reference should be added

result:
the reference is back and the "generate" checkbox is unchecked!
 
========================================================================
 Issue: 422303    02/26/06  	Status: Closed	11/16/06  
Severity : 1         	Est. release: tyresias
Label  : PD12 XML Model "shortcut" element refer namespace's element miss Mapp
	 ing Tab
Description :                          
	 -----------------------
PSE NOTES:
-----------------------
Description of Problem:   
In XML model of PD10, "shortcut" items (element, attribute), which reference
a Namespace's item , is still allow to map to OOM or PDM item.
However, in XML model of PD12, the "Mapping" Tab of items will disappear as
long as it become a "shortcut" by defining the "reference" property.

OS:  Win XP
Versions and Build #s Reported:  PowerDesigner 12.0 GA bld 1700
Versions and Build #s Tested:  PowerDesigner 12.0 GA bld 1700
Database/Driver File/Version:  not related.
 
========================================================================
 Issue: 422053    02/23/06  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : Error message : Column cannot be null when using  union all
Description :                                           
	 PD 11.1 and PD 12
DBMS : DB2 CS V8

When you delete a view column into SQL query Statement which contains an Union 
All, the view columns name is well deleted but not the row.


Steps to reproduce
DBMS : DB2 CS V8
New view
View --> Properties --> Sql Query
type here
Select C1 A , C2 B , C3 C , C4 D
From
T1
Apply
Go tot View columns : result OK
Return to SQL Query and select all the statement and delete it,  Apply
return to View columns : Result is OK

New view
View --> Properties --> Sql Query
type here
Select C1 A , C2 B , C3 C , C4 D
From
T1
Union All
Select C1 A , C2 B , C3 C , C4 D
From
T2
Apply
Check View columns Result OK
Return to SQL Query and select all the statement and delete it,  Apply
return to View Columns 
There is no name anymore but have a look we still have 4 rows in the list.
Click  SQl Query tab
error message : Name cannot be null.

This problem occurs also when you delete only one columns in the SQL Query 
statement 
========================================================================
 Issue: 420773    02/09/06  	Status: Closed	04/09/07  
Severity : 3         	Est. release: tyresias
Label  : DB2 OS/390- decimal / dec data type flagged as different due to PhysO
	 dbcDataType
Description :                      
	 *** Thursday, February 09, 2006 2:33:37 PM *** ahill *** 11.1.0 & 12.0.0

As part of CR 268368, a space was inserted before the '%p' variable for the 
Decimal data type.
Decimal(%s,%p) was changed to Decimal(%s, %p).

As a result of this change, when the customer reverse engineers via ODBC into an 
existing model, Decimal data types are being flagged as different when a Dec 
data type exists in the model. 

To correct 
this========================================================================
 Issue:

1) Create a copy of the default DB2 7 OS/390 DBMS file
2) Change line 4 of the script\data type\PhysOdbcDataType category
 DEC(%s,%p) = DECIMAL(%s,%p)  becomes DEC(%s, %p) = DECIMAL(%s, %p)
(A space is inserted before '%p' for both items.) 


 
========================================================================
 Issue: 419266    01/24/06  	Status: Closed	09/07/06  
Severity : 2         	Est. release: tyresias
Label  : List report : mapping is missing
Description :                                                                      
	  *** mardi 24 janvier 2006 13:41:38 *** vaudino ***12.0 GA
in list report wizard or list report prop sheet
on PDM, in object combo, mapping object is missing
no way to create a list report on mapping although a list is available in report 
items                                                                                                                                                                                                                                                                          
========================================================================
 Issue: 408977    11/04/05  	Status: Closed	04/17/07  
Severity : 3         	Est. release: tyresias
Label  : The case on subobjects is not updated properly when using copy/paste
Description :                                  
	 *** Friday, November 04, 2005 4:29:04 PM *** ahill *** 11.1.0 EBF6

When using copy/paste between models with different cases set for the naming 
conventions, the case is not updated for subobjects (columns, attributes...)


I have the following scenario:

Two PDMs with codes being displayed by default

In PDM_1, the case for columns is set to lower-case
In PDM_2, the case for columns is set to UPPERCASE
Create a table_1 in PDM_1 with several columns
Copy & paste table_1 into PDM_2

Note: the columns remain lower-case

Modify a column
Note: the column code is now changed to UPPERCASE

*** Friday, November 04, 2005 4:29:05 PM *** mledier ***
This is an 
old========================================================================
 Issue (not a beta bug) for which you can enter a new CR.
The problem only occurs with subobjects (columns, attributes...)

A really simple vbscript that fix it would be
<<
dim t, c
for each t in ActiveModel.Tables
   for each c in t.Columns
      c.SetNameAndCode c.Name, c.Code
   next
next 
>>

 
========================================================================
 Issue: 379761    02/10/05  	Status: Closed	03/27/07  
Severity : 3         	Est. release: tyresias
Label  : "Compare Text Property Values" window does not show carriage return
Description :                                   
	 Reproduced also with procedures
Was working fine with 9.5.1 version
New behaviour from 10 version

*** 12 December 2005 10:21:54 *** mtinevez ***
The "Compare Text property value" window does not show carriage returns in the 
body of trigger syntax.
Tested and reproduced with PD 11.0_1363

Customer would like to be able to see the carriage returns correctly in "Compare 
Text Property Values" window. Today, the entire trigger syntax shows everything 
in a single line.                                     
========================================================================
 Issue: 369569    10/19/04  	Status: Closed	03/27/07  
Severity : 3         	Est. release: tyresias
Label  : Reverse trigger : a difference has been detected in the merge window
Description :                                  
	 PD version : PD 10, 10.1 and 11 build 1355
DBMS : reproduced with Oracle, ASE


RE a trigger inside an existing PDM, without doing nothing a difference is 
detected for each trigger Text value
Problem occurs during a RE by script and by ODBC,

Steps to reproduce
New PDM
2 tables T1 and T2
Few column in each table with PK
Reference between the 2 tables with integrity implementation sets to Trigger
Rebuilt triggers
Generate a trigger script
Without doing nothing
Re the trigger script
In Merge window, a difference has been detected on each trigger

Workaround
Menu database edit current DBMS
Tab Trigger templates
For each template, delete the first line : the comment line
Note
If the first line is replaced by an empty line, a difference will be detected on 
the text property but if you check the compare text property values, 
you will get no difference !!!!
 
========================================================================
 Issue: 365841    09/02/04  	Status: Closed	05/02/07  
Severity : 4         	Est. release: tyresias
Label  : PDM Model options, trigger section text should be Automatically rebui
	 ld triggers
Description :                      
	 *** Thursday, September 02, 2004 3:47:46 PM *** ahill *** build 1317

In the PDM Model options, text for the trigger section under Table & View 
reads...

Rebuild automatically triggers

This should read 'Automatically rebuild triggers'                                                                                                                                                                                                                                                                                
========================================================================
 Issue: 365394    08/28/04  	Status: Closed	05/02/07  
Severity : 4         	Est. release: tyresias
Label  : CDM/PDM generation extended depency link moves to forefront of attach
	 ed objects
Description :                       
	 *** Saturday, August 28, 2004 7:48:50 AM *** ahill *** 10.1.0

When generating from a CDM to a PDM, the link line of an extended dependency 
moves to the forefront of the attached objects in the PDM.

To reproduce:
1) Add an entity with an attribute
2) Add a Note with some text
3) Add an extended dependency drawn from the note to the entity
Note: Default is to have link in sent to back
4) Generate the PDM
Note: The link has now brought to the front of the attached objects
                          
========================================================================
 Issue: 347274    02/23/04  	Status: Closed	11/16/06  
Severity : 2         	Est. release: tyresias
Label  : when using drag/drop in ordered lists, the incorrect rows are moved
Description :                                   
	 *** Monday, February 23, 2004 10:31:23 PM *** ahill *** 10.0.0 EBF1 & 
9.5.2 EBF6

Arranging items on an ordered list using drag n' drop doesn't move some of the 
selected items (columns in a table). When multiple columns are moved, the last 
listed column is not moved.  The column preceding the first listed column is 
moved instead.  Also, for each column selected, the graphic shows an additional 
column being moved until you close the property table property window.

To reproduce:
1) Create a table with 9 columns, 1-9
2) Select columns 8 & 9 and drag them to the first column position

Note: In the graphical display, it shows that columns 5, 6, 7, and 8 have been 
moved.

3) Click OK
Note: Columns 7 and 8 have been moved eventhough columns 8 and 9 were originally 
selected.

(If more columns are selected, the behavior becomes more erratic.)

Workaround is to use the navigation arrows to move the columns. 
========================================================================
 Issue: 317414    04/28/03  	Status: Closed	05/02/07  
Severity : 4         	Est. release: tyresias
Label  : Add option to Find Data item in Browser from the entity attribute pro
	 perties
Description :                          
	 *** Monday, April 28, 2003 10:42:20 AM *** ahill *** 9.5.2.873

Add a new option to "Find the Data item in the Browser" from the Entity 
Attribute property window.  Currently, the entity attribute can be found using 
the "Find in Browser" (Ctrl+Shift+B) option.

The customer is requesting this functionality because extended attributes can 
only be associated through the Data Item metaclass.  If they desire to review 
this information, you must either manually find the Data Item in the browser or 
in the list of Data Items and then open the properties of the Data Item.
 
========================================================================
 Issue: 288037    06/19/02  	Status: Closed	05/06/07  
Severity : 3         	Est. release: tyresias
Label  : Generate Database: Foreign keys dropped without checking if they exis
	 t
Description :                                
	 *** 6/19/2002 AHILL***  9.0.0 EBF3 and Atlas beta
MS SQL SERVER family of DBMS
When executing generate database with the drop foreign key option selected, 
PowerDesigner generates DDL to drop the foreign keys without first checking to 
see if the foreign key constraint already exists in the database.  As a result, 
many warning/error messages are generated by SQL Server when the DDL is 
executed.

As a work-around, I added the following section from the drop table statement to 
the drop reference section:

if exists (select 1
            from  sysobjects
           where  id = object_id('[%QUALIFIER%]%TABLE%')
            and   type = 'U')
 
========================================================================
 Issue: 256751    08/09/01  	Status: Closed	05/02/07  
Severity : 4         	Est. release: tyresias
Label  : When zoomed in all the way the screen no longer scrolls or eventually
	  stops scro
Description :                      
	 In model graphics, Zoomed in all the way (F6 key) the screen no longer 
scrolls or eventually stops scrolling using Arrow keys on Keyboard as well as 
using mouse wheel.  You must back off/zoom out (F7 key) at least one degree to 
get scrolling to work again.


**** Thu Aug 09 14:17:58 EDT 2001 *** mbartrum **** 
tested in 7.5.0.826;  8.0.1.275 ; 9.0.0.399 BETA1  with the same results...

**** Thu Sep 26 09:33:01 EDT 2002 *** mbartrum **** 
Tested in v. 9.5.0.655 (ebf1) with the same results.



               
========================================================================
 Issue: 230876    01/09/01  	Status: Closed	05/02/07  
Severity : 4         	Est. release: tyresias
Label  : OOM generation fails if the directory does not already exist.
Description :                                         
	 
**** Tue Jan 09 14:59:44 EST 2001 *** ahill **** build 169 & 7.5.0.826
When trying to generate an OOM into a directory that does not exist, fails with 
an error message.  The customer reported this with VB 6 as the language, but the 
same behavior exists when Java is the target language.

To reproduce:
1) Create an OOM that contains Class 1, and Class 2.  Add several attributes to 
each Class and draw an association from Class 1 to Class 2
2) Select ..Language>Generate Java Code
3) Type in a directory name that does not already exist
4) Click OK
Note:  The generation process ends with the following error:  "Disk full or read 
only!  Generation aborted"

This is similar to CR# 195341
 

(109 rows affected)
