Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - Paolo F Cantoni

Pages: 1 [2] 3 4 ... 80
16
Bugs and Issues / Bizarre bug in QuickLinker!
« on: September 13, 2018, 01:52:22 pm »
In investigating the use of @Element in the QuickLinker, we accidentally discovered (after many hours tracking) a bizarre bug!

You can comment out a line in the QuickLinker by the use of a leading //

So far so good!  We generate the VAST bulk of our QuickLinker file via an automated process and we sprinkle some comments to help break the extremely large file up into sections - such as:
//,//,·,//,·,·,·,·,·,·,·,·,·,·,·,·,·,·,,,,,

//,//XSD Simple Type,·,//XSD Simple Type,·,·,·,·,·,·,·,·,·,·,·,·,·,·,,,,,


Notice the use of multiple // in the same line!  We've had NO problems with these since we started automating.

We had an entry in the manually managed section of the QuickLinker file concerning @Element:
//@Element,,@Element,,,Class,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,True,True,(Nesting),0,,,,,
 //Doesn't Work!   Source Element type needs to be concrete!
We placed this at the start of the section where we were manually managing the concrete source types such as Activity,,@Element,,,@Element,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,True,True,Nesting,0,,,,,

Class,,@Element,,,@Element,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,True,True,Nesting,0,,,,,

To remind us of the issue.

We had the concrete sources working, and then added the comment.  Sometime later, we found the sources weren't working anymore - although AS USUAL there was NO indication that there was a problem.
After many hours of debugging we tracked it down to:
//@Element,,@Element,,,Class,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,True,True,(Nesting),0,,,,,
 //Doesn't Work!   Source Element type needs to be concrete!
(on the same line)
Will KILL any following lines in the QuickLinker file, whereas
//@Element,,@Element,,,Class,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,True,True,(Nesting),0,,,,,

//Doesn't Work!   Source Element type needs to be concrete!
(on separate lines)

Doesn't!  How Bizarre!!!

Reported,
Paolo

17
Bugs and Issues / Use of @Element in QuickLinker
« on: September 10, 2018, 05:06:54 pm »
We can use the @Element (for example) in a QuickLinker entry to allow more generic relationships specification.  Here's one entry that works...
Class,,@Element,,,Class,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,,True,(Nesting),0,,,,,

Any Class can be nested within any Element.
Since any element can be nested within another element (or package - which in this context appears to be an element) we thought we'd generalize it to:
@Element,,@Element,,,Class,,Nesting,Nstng,to,nesting¯,nesting¯,True,True,,True,(Nesting),0,,,,,


However, this doesn't work.  Should it?  If not, why not?

Reported,
Paolo

18
Is there any way to force the inclusion/exclusion of specific fields in the query result of the Model Search?
For example, a lot of our searches benefit from including the Version field in the output.  However, it's not there by default and we have to add it via the Field Chooser.

Must I throw away the Query Builder and do everything via direct SQL queries to get the specific fields and sort orders I need?

TIA,
Paolo

19
Our diagrammer Add-In was created a LONG time ago when it wasn't possible to create attached diagrams through the API.  We've left it alone, until now, when we've discovered that when it creates the attached diagram for an InteractionFragment our existing process will set the type of the fragment to "strict", regardless of what it started as.  This is because the NType field is incorrectly overloaded by the architecture.

Notice I said "attached diagram" above.  In our repository for every modelling vertex, we create a Neighborhood diagram.  We store it as a child of the vertex.  It's not strictly a Composite Diagram, but we want the UI effect of double-clicking on the vertex in a diagram opening up the attached Neighborhood diagram.  It's our understanding that the only way to attach such a diagram to the vertex is to make the vertex composite and the set the composite diagram.  Can anyone confirm that this is the correct understanding of how to attach such a diagram to a vertex?

As I read the current Help System, I have to make the element composite by setting the IsComposite property and then attaching the diagram via the SetCompositeDiagram method.  I can use the CompositeDiagram property to determine if the Composite vertex already has an attached diagram.

Do I understand it correctly?

TIA,
Paolo

20
Bugs and Issues / Sequence Diagram connectors and t_diagramlinks
« on: September 06, 2018, 12:41:02 pm »
One of our Integration Architects has started using Sequence Diagrams to specify details to our developers.  One of our internal integrity checks is to ensure there is only 1 copy of a semantically specific relationship between two elements.  For example, you can't can't have more than one Association with the same name (or Null) between the same two objects.  If they are representing two different semantics then they have to be different in some way (by name).

So, we started getting "duplicate" connectors of the type "Sequence".  We have now established that these are legitimate and will adjust our tests accordingly.

However, in investigating the "goings on", I wrote a query that said: for such Sequence Connectors, show me which diagrams they appear in?  Imagine my surprise, when NO entries appeared in t_diagramlinks!  Should this be the case?  It's a long time since I, personally, used Sequence Diagrams and at the time I certainly didn't look inside the repository.  So, is it the case that arcs of the type "Sequence" won't appear in t_diagramlinks?

TIA,
Paolo

21
Bugs and Issues / Printed text from Shapescript not copied from diagram
« on: August 31, 2018, 09:54:05 am »
We implemented a Datastore element and wanted to be able to handle multiple types of datastores: RDBMS, DocumentStore, Cache, CSV etc.  So we created a tagged value DatastoreType to manage this.
Normally we create a conditional shapescript which uses the tagged value to change the shape and or icon of the element controlled by the tagged value.  Time and resources didn't permit (at this time) so we jury-rigged an interim solution where we just print the tagged value below the SW corner of the item with the following code.
Code: [Select]
setorigin("SW",20,0);`addsubshape("datastoretype", "SW");
shape datastoretype
{
v_align="top";
h_align="center";
print("(#TAG:DatastoreType#)");
}
Works fine on the diagram, but when I select the item on the diagram and paste it into (say) an email, the datastoretype is NOT copied.

Reported,
Paolo

22
Bugs and Issues / MDG Load Error Logs Needed
« on: August 30, 2018, 12:32:33 pm »
The "silent" failure of MDG loads is REALLY starting to bug us.

We need an Error log of errors encountered during the loading of the MDG so we can debug the MDG.

Reported,
Paolo

23
Bugs and Issues / MDG Enabling checkbox malfunction
« on: August 30, 2018, 12:28:50 pm »
In the past, if you loaded an MDG and the checkbox in the MDG Technologies dialog was unmarked, then it meant there was an issue with the MDG. 
In v14.1 that's not necessarily the case.  We're finding that if we open EA the checkbox may be unmarked, but a substantial portion of the MDG is operational but other parts (particularly Searches and Package Models) are not.
 However, if we exit then open v13.5 and enable the MDG (so that the checkbox is marked), if we reopen v14.1 the checkbox remains marked and the MDG operates properly (as far as we can tell).

Reported,
Paolo

PS: a log of the MDG load errors would be VERY handy!

24
Bugs and Issues / Spurious Tagged Value types (non-enabled MDGs)
« on: August 23, 2018, 05:14:22 pm »
I've suddenly (b1425) got Tagged Value types from non-enabled MDGs appearing in my Tagged Value drop-down - for example, DMN1.1::*.  First time I've added some tagged values for a while.

They're not in t_propertytypes (not unexpectedly) and the applicable MDGs are disabled in the Specialize dialog.

How can I get rid of them?  Is it a bug that they are appearing?

Paolo

25
I've just got a new Surface Book 2 (under Windows 10) and use Sparx with 3 screens, two Desktop screens @ 1680x1050 and the 13.5" Surface book 2 @ 3000x2000 (with scaling set at the recommended 200%).

Most apps handle this pretty OK.  Moving windows between screens adjusts the size of the window accordingly.

However, EA (as usual) does something unexpected.  If I move the Element Browser window (for example) from the Desktop screen to the Surface Book Screen, the physical size reduces dramatically (in order to maintain the pixel size) and a "ghost" window is also created - slightly lower and to the right at what looks like the correct size if the window was being adjusted for the scaling.  The ghost window stops me positioning the window in the correct place and the real window is too small to read without the magnifier!

Anyone else seeing this at high resolutions with multiple screens?

TIA,
Paolo

26
Automation Interface, Add-Ins and Tools / Last Editor - functionality
« on: June 29, 2018, 11:30:37 am »
Since it doesn't look like Sparx is going to implement it soon, we've decided to implement "Last Editor" functionality.   In an enterprise-wide environment (as I've mentioned before) the original author of an item in the repository is of only passing (and usually historical) interest.  However, the last person to edit the item s is much more important!

Since EA has only one author field, to add Last Editor functionality, we need to either
  • Hijack the Author Field
  • Add the Last Editor property as a Tagged Value
If we hijacked the Author field, we'd need to add the Original Author property as a tagged value.
We'd also probably add some timepoint properties to record the event.

I guess we'd hook into the OnPostNewXXXX and OnNotifyConttextItemModified events to support this.

Thoughts?
TIA,
Paolo

27
The Help says:
On a diagram to a file
To create a hyperlink on a diagram to an external file, simply click on the file in a file list (such as Windows Explorer) or on your Desktop and drag it onto the diagram.
A short context menu displays with two options - 'Hyperlink' and 'Artifact'. Click on the 'Hyperlink' option to create the hyperlink on the diagram.
The link is effective immediately, and you can right-click on it to add or change properties as necessary.
Files of most types - including .sql and .ddl - are opened within the appropriate Enterprise Architect code editor.


I tried to drag a number of different types of files onto a diagram.  All I got was the "No go" glyph.  What am I doing wrong?

TIA,
Paolo

28
We're looking at possibly intercepting EA when a user attempts to visually embed one item "into" another.  However, looking at the list of Broadcast Events, I can't see anything I can leverage to detect an attempt at visual embedding.

Can anyone suggest a mechanism that we could use (even outside EA) to detect the "dropping" of one item into another or the removal of a visually embedded item out of the encompassing item?

TIA,
Paolo

29
We have a number of relationships defined in our QuickLinker between relationship metatypes and various element metatypes.  Under v13.5 the Quicklinker displays them as appropriate.  Under v14 with the SAME MDG file, they are missing.

Reported,
Paolo

30
Automation Interface, Add-Ins and Tools / Oracle Reverse Engineer
« on: May 28, 2018, 02:55:51 pm »
One of our users is trying to reverse engineer an Oracle DB.

We're usually a SQL Server shop, so I haven't done one of these for ages.  Using b1351, he seems to be doing all the right things on the Import from ODBC dialog.  Are there any "gotchas", "secret sauces" that we should be using?

The result is that after asking it to reverse engineer everything, for the tables, we get the table objects, no features or relationships - just empty boxes...

TIA,
Paolo

Pages: 1 [2] 3 4 ... 80