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 ... 74 75 [76] 77 78 ... 80
Over the decades, because EAUI places the labels in different positions depending on whether the relationship is reflexive (points to self) or not, I've tried to figure out how to be able to tell if a particular instance youare "shaping" is reflexive or not - all to no avail.

Is it possible?  If not, then surely a new connector property is needed.


I found some posts dating back to 2006-2008 on trying to import from Altova UModel.

I guess nobody's had any success?

The differences in xmi look pretty daunting!

Let me know if you've managed to do it...


Automation Interface, Add-Ins and Tools / XML Importer/Synchroniser
« on: May 25, 2015, 11:15:29 am »
Just as Objects are instances of their Classes, and XML file  is an instance of its XSD (if it exists).

I had a quick search to see if there's one yet.  It would appear (still not) not.  I built one (at least an importer) a few years ago - but before I re-invent that wheel, does anybody have a reasonably generic XML importer?


Automation Interface, Add-Ins and Tools / QuickLinker Groups
« on: May 19, 2014, 11:13:26 am »

I have got (at least some of) the hang of the QuickLinker.  I can get it to create menu groups when creating a new object by dragging from an existing object.  However, I can't get it to create the same groups when dragging between existing objects.  Am I missing something?


I think this questionhas been asked before but I couldn't find an answer using search.

I need to determine if the BPMN Add-In has been loaded or not; since the BPMN object will render differently if it is loaded than when it is not.

That's reasonable...  However, since the effect on diagram rendering can be quite dramatic, I need to be able to tell if the Add-In is CURRENTLY loaded (remember since v9.3 Add-ins can be dynamically loaded).


Anybody know how to create a stand-alone AssociationClass in an MDG Technology?

All my attempts have so far failed.  If I extend a Class metaclass, all works fine, If I extend the AssociationClass metaclass, it creates the Association arc AND related AssociationClass vertex (which is to be expected).

I just want to create a stand-alone AssociationClass (because I want to manage N-ary Associations).

I think I'd like a metaattribute (something like: _standalone) that would create the AssociationClass vertex only - without the accompanying arc.


Automation Interface, Add-Ins and Tools / PDF Reports - allowing hyperlinks
« on: February 14, 2012, 06:17:52 pm »
We can now generate PDF version of our diagrams directly from EA.

However, in the process we lose the ability to follow hyperlinks within a PDF reader.

in a related topic: HTML Reports - How to maintain links?  AndyJ and I discuss how it might be possible to retain diagram linkage information from generation to generation by means of using GUIDs as folder and file names.

In trying to combat Visio (which can produce PDFs of diagrams with executable hyperlinks), it would be neat if we could do the equivalent of creating a HTML Report (Web site) but with PDFs so that non-technical punters could just "follow the bouncing ball".

The GUID process is just one solution to the problem of target consistency.

What do others think?

If we want to get modelling accepted into the wider world, we need to dramatically lower the friction involved for non-technical people can more easily "climb on board".


My client has hundreds of business process diagrams created by external consultants in Visio using the Visio BPMN 1.2 stencils.

I had a go at importing them with the existing Visio importer and there were the obvious problems caused by the importer not understanding BPMN.

In the interim, because the diagrams were fairly simple, they came over reasonably well and I can see how to convert them via DB queries from what was imported to (more like) what they need to be.

I can get the shapes looking correct enough, but I can’t seem to convert them to EA BPMN v2.0 elements from the existing imported elements.  I thought all I would need to do was to fix up the Stereotypes record in t_xref and add an appropriate CustomProperties record in t_xref so that they looked like normal EA BPMN v2.0 elements.  However this doesn’t seem to be enough, EA won’t recognise them as BPMN elements and (consequently) won’t show the BPMN2.0 properties in the Dialog when the element is opened.

What else do I need to do to make EA recognise these imported elements as BPMN v2.0?


Automation Interface, Add-Ins and Tools / Getting consistency in MDG "Types"
« on: September 15, 2011, 06:31:13 pm »
We created a repository and added thousands of objects.   We LATER created an MDG technology and continued to add objects.

Now, in the Model Search window, when we do a search for objects we get inconsistent results in the Type column.  Suppose stereotype "S1" is a stereotype NOW associated with MDG technology T1.  The name of the type thus created in T1 is (say) T1_S1.  Some objects in the Model search window have Type = T1_S1 and stereotype = S1 and others have Type = Class and stereotype = S1 - as far as we can tell these are the objects created before the T1 MDGT was created and used in the toolbox.
ALL objects have been "FIXED" to ensure the t_xref stereotype entries have the value associated with the MDG T1 (that is, they have an FQName entry that references T1::S1) and consequently will ALL show T1_S1 in the Type slot of properties window for the object.
If one accesses the properties of a non-T1_S1 object and saves it, the Model Search Type changes from Class to T1_S1 - however this DOES NOT persist across EA invocations. >:(

So what I want to know is where is the "secret sauce" that will make all my objects show T1_S1 in the Model Search Type (or is it just another EAUI bug - among SOOOO many). ::)

We have literally thousands of pobjects to fix and manually is not an option.


The full text of the message is:
Selection information is unavailable on hidden diagrams

Anyone know what causes it (or even what it exactly means)?  

I can't yet reliably reproduce it and it stops my Add-In dead in its tracks when I'm trying to process a number of elements in a batch.


Automation Interface, Add-Ins and Tools / Alternate image in MDG
« on: April 08, 2010, 08:29:19 pm »
I tried to create an alternate image for my MDG but I can't get it to work...  I'm obviously missing the "special sauce".

Can any kind soul show me how it's done?  There were no effective references in a search.


The help doesn't say so but experiments I've done seems to suggest that the Collection GetByName method is case insensitive.

Can anyone (preferably Sparx) please confirm this?

If it is case insensitive, what happens if more than one object has the same name?

How are case sensitive languages such as C# handled?  CHours and CHOURS are differnt classes, but GetByName (for class "chours") will only find one - the small amount of experimenting I've done suggests the choice by EA is arbitrary...


It's been a long day and I can't find any method for detecting if a diagram that I want to get from the repository (say via GetDiagramByGuid()) is already open in the Workspace.

Can someone tell me if it's possible and if so how?


It's pretty obvious the setting is unique for each repository, but where is it stored?


I can create Composite Elements by the methods indicated in other posts.  This post is in regard to the "chain link" symbol that EA provides to indicate the attached diagram and allow you to activate it by means of a double click.

I've spent a fruitless day trying to get this to work in a consistent manner (actually to work at all...).

As I said above, I can get the item in the diagram to indicate that it is "Composite" on the context menu for the item - after my processing.  Indeed, on any currently open diagram, the item shows as composite in the context menu.

If I open a (previously unopened but preexisting) diagram that contains the composite item, it will correctly display with the "Chain Link".  However, any previously open diagrams won't display the "Chain Link".

To get the previously open diagrams to display the "Chain Link" I need to manually reload each diagram (through the UI).

I've tried all manner of combinations of the various available methods - all to no avail.  Part of the debugging seemed to indicate some sort of race condition - as when I had a breakpoint set in the code, I could more consistently (but not fully consistently) get the "Chain Link" to appear unaided.  However, since I couldn't get the code to pass the "Alice in Wonderland (Or was it through the Looking Glass) Test":  Consistently produce the same output 3 times in a row through 3 separate invocations of the program - it's difficult to be definitive.

Does anybody know how to get it to work?  Or is it just a bug?


Pages: 1 ... 74 75 [76] 77 78 ... 80