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 ... 5 6 [7] 8 9 ... 78
Bugs and Issues / PDF Print DOESN'T include Diagram Name
« on: August 15, 2017, 05:03:44 pm »
If you try to print a number of diagrams to PDF, you have to manually replace the Model Name with the Diagram Name.

Why can't the Diagram name be supplied?  Since the supplied name is ALWAYS the Model name, you run the risk of overwriting a previously printed, but different, diagram.


While checking t_xref for other reasons, we discovered duplicate entries for "DefaultDiagram" rows.   That is, there was more than one entry in t_xref with the same Client and Supplier ea_guids for that Name!  The Integrity Checker [Shift+F9] has not detected them!  It should!

Since there can only be one such diagram per element, having duplicates in the DB is, by definition, an error.


Dragging a new item off toolbox will honour shapescript defsize() for rectangular form, creating one via browser (<context menu> | Add Element...) doesn’t.  The size created is related to the values in the MDG definition for the element. 
Code: [Select]
<Stereotype name="element" metatype="Element" notes="" cx="100" cy="60" When creating via browser, IF the size specified in either (or both) cx and cy is greater than the size specified in the shapescript, THEN EA will apply the shapescript size, but if the sizes are less than the shapescript size, it will apply the MDG size.   Dragging from the toolbox will always setup the correct size.


Please rectify!

Reported (I think for the second time, but at least I know WHY it happens now.)

Bugs and Issues / Connector Labels and Dotted Line styles
« on: June 21, 2017, 05:33:04 pm »
I'm sure I've reported this over the decade or so, but I can't find it in the search.

If you create a label for a connector (such as add name of Association), if the line is a Solid line, then the text of the label can be positioned so that it overrides the line and you can CLEARLY see the text.  However, if the line is one of the non-solid (setlinestyle()) lines, then the underlying line will "peek" through - making the text unreadable.  Can this be fixed, please?


General Board / Inconsistencies between Help and t_connectorTypes
« on: June 15, 2017, 03:55:15 pm »
There are inconsistencies between what the help says in .../User%20Guide/model_domains/transformingconnectors.html and the t_connectortypes table in the repository.

Connector_Type   Help Document
Abstraction      Abstraction
Aggregation      Aggregation
Assembly         Assembly
Association      Association
Collaboration      Collaboration
Connector      Connector
ControlFlow      ControlFlow
Delegate         Delegate
Dependency      Dependency
Deployment      Deployment
ERLink         ForeignKey
Generalization      Generalization
InformationFlow   InformationFlow
Instantiation      Instantiation
InterruptFlow      InterruptFlow
Manifest         Manifest
Nesting         Nesting
NoteLink         NoteLink
ObjectFlow      ObjectFlow
Package         Package
Realisation      Realization
Sequence         Sequence
Substitution      Substitution
Usage         Usage
UseCase         UseCase

Should these be identical or, if not, why not?


We use stereotyped Notelinks for various administrative metatypes in our repository.  We've observed that if we create the first link from the Toolbox, the [F3] (repeat last connector) functionality will work correctly and create a new arc of the correct metatype.  However, if we use the QuickLinker to create that first link, the [F3] function will create an unstereotyped Notelink.  Obviously, the same functionality is not being used...  Can we please ensure that the behaviour is the same (for ALL metatypes) - regardless of the first instantiation method?


Bugs and Issues / Tooltips show HTML codes
« on: June 15, 2017, 09:58:43 am »
If you enable "hovering" tooltips (don't know if this bug also occurs for the status bar tooltips) and the tooltip cutoff occurs in the middle of a HTML element, the HTML code is visible.  Since there is no formatting on a tooltip, the HTML should have been stripped off in the first place.


Increasingly, we are developing scripts that operate directly on the Selected Diagram Link (rather than the underlying Selected Connector).  Similarly with Diagram Objects.

However, there is no SelectedDiagramLink or SelectedDiagramObjects attributes for the Diagram Class.  We are forced to scan the appropriate collections to determine the diagram link from the underlying connector.

Can we have direct access to the diagram specific items?


The [ ] Use Rectangular Notation context menu item is created on the fly by EA by inspection of the shapescript.

However, for a number of element types, Activity, being the one that caught us out today, this is NOT respected.

Please rectify.  I see no reason why, if the shapescript has the requisite coding, that the context menu items shouldn't be provided.


We are finding increasingly that we can define a structured Tagged Value (typically an enum) and re-use it in a number of places.  I'm not talking about using the same named tagged value in multiple elements - that's easy enough to manage.  I'm talking about using the same set of values (with effectively the same set of semantics) multiple times for the same element.  One example is  "Degree of Fit" - where the same enum values are used for three different types of "fit" (for the same element).  We have currently created three Tagged Values defined identically, but once they've been created the fact that they have the same enum literals and semantics is lost.

I couldn't see how to use one definition in multiple named tagged values.  Is there such a mechanism?


It seems that status colours are only visible if the item is in rectangular form.  Is that correct, or have I left the secret sauce off the shapescripts I'm using?


Suggestions and Requests / Word Wrap on non spaced names
« on: May 29, 2017, 04:50:08 pm »
I'm sure there was some discussion recently on the forum about word wrapping.  But I can't find it.

Anyway, there should be the ability to word wrap names that don;t have spaces, but do have mixed case naming.  We commonly use a naming scheme we've called Complex Pascal Naming where capital letters separate words, but Acronyms and Initialisms are allowed to retain their Uppercase.  It's not that hard to write an algorithm to split the words up and determine where the word wrap should be.

Please consider.


Bugs and Issues / MDG File grows too large!
« on: May 26, 2017, 01:25:50 pm »
We recently significantly increased the size of our MDG by automatically generating a series of relationship to relationship QuickLinks (some 7000 odd in this case)

EA complained about a failure in our XML.  But it turned ut to be a size-related issue:  At line 42965 (round about column 13x - it varies depending upon the file content).  EA will fail to import the MDG.  This expansion was to be only the initial step, the file is currently 10-12 Mb in size and we expect it to grow more.  What is the current Maximum size for an MDG file?  Can this be fixed urgently?


QuickLinks "from" relationships don’t work correctly in “empty space”
  • Inter-relationship QuickLinks are listed (they shouldn’t be)
  • @Element QuickLinks don’t do the right thing (open the available set of elements from the list of available elements)

Relationship to relationship QuickLink doesn’t indicate selected target (like for elements).  This makes it difficult to ensure you're going to hit the right "target".
Please rectify.


Pages: 1 ... 5 6 [7] 8 9 ... 78