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.


Messages - Paolo F Cantoni

Pages: 1 [2] 3 4 ... 544
16
Since I haven't been using v16 (for reasons previously noted), I hadn't noticed this.  However, it seems to be a significant backward step (as you imply)!  Asking for the rationale has been pointless so far.

Looks like I'll be moving to v16 with b1605, so I have new irritations to increase my blood pressure.

Paolo

17
General Board / Re: Shapescript with a compartment for the namespace
« on: June 15, 2022, 09:01:55 am »
Thanks Thomas and Eve, I'll give a try.

Eve, I suspect that packagepath may not do what I am after, please see this thread for context https://sparxsystems.com/forums/smf/index.php/topic,46931.0.html.
Call me a literalist, but I would have thought PackagePath should yield the path to the package (as described in the browser.)  Whereas, PackageNamespace might return the Namespace created by traversing the package path from the namespace root (which could be anywhere on that path) and the package - taking into account any suppressed packages (that shouldn't contribute to the namespace).  The results could be quite different for the two properties (or identical in many cases).

Perhaps that's what Modesto means?

Just a thought.
Paolo

18
While we, like some other organisations, haven't been able to upgrade to v16 due to an inheritance bug when defining models via the metadata[1] - we have been persevering with v15.2.  We're noticing that when we use our model-generated QuickLinker definitions between certain stereotypes, some of the options go "walkabout" - that is, they disappear!  Try to link to another stereotype (which should produce the same result and they are present!).
When this happens, our research occasionally finds there are (legitimate) behind-the-scenes reasons why EA decides not to provide the option.  However, often, there seems to be NO "rhyme or reason" for this (that we can see).

So after spending a wasted two hours trying to establish any logical reason why the option was missing, we tried the SAME MDG specification on the SAME repository under v16 to discover that under V16, the options are "all present and accounted for"!

Can anyone, but preferably a Sparxian who can speak knowledgeably about the subject, indicate what inheritance improvements that affect the QuickLinker behaviour were made in the v16 release?  Those of us who have been here a long time are aware that these explanations are often obfuscated under arcane descriptions of version changes.  We'd like to understand clearly what has been improved so we can quickly decide that a new QuiuckLinker failure under v15 has been fixed in v16 rather than have to waste a lot of energy starting up v16, cross-checking and then determining that the issue still occurs under v16.

In addition, are there any more changes to inheritance under MDGs for the next release of EA?

Paolo

[1]We have since been told that the defect will be fixed in the next release.

19
Quote
Yes, it makes sense that the user would want at least some types from the new technology. In an already tight restriction, it makes sense to require explicit selection of the newly allowed types.

Sparx has always had a problem creating a useful user Selection experience. See my many posts in the past.  They appear to be living in the 1980s when User Experience was not considered.

As Geert said "Pff"!

Paolo

20
When adding a new technology to a strict model based perspective set, nothing in that new technology is selected by default.
You have to manually select all the elements, diagrams etc... before they can be used.

The default should be that everything is selected, so you can uncheck what you don't need.

Reported

Geert
Wot 'e sed!
Paolo

21
Thanks for you answers
Now after studiously ignoring SQL for 40 years (I'm an Embedded Sweng) I'm going to get my hands "dirty"  :o
[SNIP]
(my emphasis)
Look on the bright side, Yonatan!  You can now add "Experience with a Functional Language" to your CV ;D

Happy Friday, everybody!   ;)
Paolo

22
No, it's some of the MDG settings he is after. I'm a bit out of synch at the moment. But IIRC Paolo had some (manual) black magic to set these properties. Can't recall the details and don't have time at the moment to check it by myself.

q.
We still make this part of the MDG by hand, we take an existing diagram specification and manually adjust it from the values in a live diagram.

Sorry, I can't help further.  We'd also be interested in how to create these directly via the model.

Paolo

24
I received the following official response from Sparx:

"This issue has been fixed, and the change has been merged for the next release. Unfortunately, I cannot provide a release date at this stage."

Let's hope it is released soon!

Paolo

25
General Board / Re: Control flows and object flows
« on: May 09, 2022, 09:45:06 pm »
What selection criterium did you use? You'll have to make sure your object flows are included in your selection.
I have tried with Filter set to Connector - None, Connector - Type, or Connector - Stereotype, and Applies to set to ControlFlow or ObjectFlow depending on the type. The end result is always the same, no change.
Hey Modesto,
Are you able to get the Legends working with ANY connector?  I have a sneaking recollection that they don't actually work with connectors[1].  Perhaps others can correct me.  I just checked our common legends and NONE apply to connectors, only shapes.

HTH,
Paolo

[1] Perhaps, another case of connectors being second class citizens?
No, it works for connectors as well. (at least in v15.2.1559)
We use that to indicate the lifecycle on our architecture diagrams for both elements as connectors.

Geert
MyBad, then.  Sorry!

However, I would still suggest testing the legend on other connector types and on other machines.  One of our machines behaves anomalously and so we first confirm any anomalies we find on other machines before reporting a defect.


Paolo

Paolo

26
General Board / Re: Control flows and object flows
« on: May 09, 2022, 09:17:00 pm »
What selection criterium did you use? You'll have to make sure your object flows are included in your selection.
I have tried with Filter set to Connector - None, Connector - Type, or Connector - Stereotype, and Applies to set to ControlFlow or ObjectFlow depending on the type. The end result is always the same, no change.
Hey Modesto,
Are you able to get the Legends working with ANY connector?  I have a sneaking recollection that they don't actually work with connectors[1].  Perhaps others can correct me.  I just checked our common legends and NONE apply to connectors, only shapes.

HTH,
Paolo

[1] Perhaps, another case of connectors being second class citizens?

27
General Board / Re: ArchiMate instance of class
« on: May 09, 2022, 09:10:52 pm »
Hi Jim,
I would recommend re-reading (and attempting to absorb) my post #9. in the original thread, you referenced.  I believe it contains the solution to the problem. Note especially my final comment about standards...

We decided to create a modelling environment that works rather than one that adheres to standards that don't actually work.  However, it DOES require that you actually think about what you are trying to do rather than blindly follow "authority".  Yes, if you look at many of my posts, I DO hold contrarian views.

HTH,
Paolo

28
General Board / Re: Archimate instance of class
« on: May 09, 2022, 02:38:26 pm »
I don't think I've looked at it at all since that discussion.

From memory, ArchiMate is a bit weird. I believe it actually recommends modelling instances as a specialization of the type. So if you have an instance of 'Data Owner' that you want to model, you create a new Business Role with a specialization to Data Owner.
That's, effectively, how we went.  Our placeholder items are generalizations of specific items.  However, we have TWO kinds of specializations:  (the more common) inheritance (IS A TYPE OF) and restriction (IS WHERE). The specific items are restrictions on the placeholders.  You can consider the placeholders the equivalent of a query "A Data Owner is an X where".  If the specific item meets the query requirements, then it is a restriction on the placeholder.

Works REALLY well for us!

HTH,
Paolo

29
General Board / Re: Stereotype Properties _AttPri, _AttPro, etc
« on: May 08, 2022, 06:40:45 pm »
AH, its amazing what a good rant will do for you.
About 10 minutes after I finished the last post I realised what I had done wrong.
The _AttPkg, etc attributes are inverse from the others (i.e. 1 = hide).

All working nicely, now onto the next bugs :-)

Thanks guys :-)
Concistency, konsistency, consistensy! TMUffe - after Paolo
and...
Manage Complexity,
   .   .   Reduce Ambiguity,
   .   .   .   .   Eliminate Inconsistency!
TM


Paolo

30
General Board / Re: Stereotype Properties _AttPri, _AttPro, etc
« on: May 07, 2022, 09:52:46 am »
Hi Jays,
Is this the exported XML from the MDG generation or a handcrafted one?
If the latter, try generating from an MDG model to see what EA creates. What you've done looks OK, but occasionally, as we know, EA has surprises "up its sleeve".

Paolo

Pages: 1 [2] 3 4 ... 544