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 - SilverSage

Pages: [1]
I've saved linked documents to two different BPMN 2.0 elements, one an activity and the other an activity with a composite diagram. Neither of the elements display the A anywhere, let alone where it's supposed render. Is there a flag I need to set somewhere to get that A to display? I've searched the EA preference tables, the element property tabs, and the diagram property tabs. I've also searched the EA doc and this forum. Am I missing something somewhere?

Thanks in advance for any assistance.


General Board / Understanding External References
« on: January 03, 2018, 11:04:12 am »
Happy New Year Everyone!

Now that's out of the way...  ;)

Can anyone tell me where I can find EA documentation containing information about external references? I'm trying to learn whatever I can about them... what EA's definition is, potential causes, potential resolutions, relationship to repository design, relationship to dependencies and how to discover dependencies that might cause them, etc.

We're trying to copy the contents of an existing repository residing in a clusterblank of a ProgreSQL database into another instance of a clusterblank of a ProgreSQL database that was stood up and tweaked a bit better than our current instance to work with EA (13.5).

When we do the copy (We have no DBA support, so this is not a database table copy; we're performing the copy from one instance of EA connected to our source repository to another instance of EA connected to our target repository.), we get a xit-ton of external references on many of our existing BPMN 2.0 diagrams we're trying to move to this new instance. Looking at their properties, though, gives us no information. We're also losing connectors from diagrams in the target repository and some connectors themselves come across labeled as external references.

Are these external references we're seeing actually unresolved pointers back to our source database, in which case this approach will never work? Or is there a way we could make it work?

We're baffled and we're trying to get ourselves "unbaffled". If anyone can point me in the right direction, it would be a great New Year start.

Thanks for any and all assistance.


General Board / Yet Another Cloud Services Question
« on: February 12, 2016, 06:23:29 am »
I'm trying to update a shared repository located on a remote server via Sparx' Cloud Services. I work in a local repository, export my package and then import it into the remote repository. The repository is a representation of a company's enterprise business processes and value streams. As such, it contains multiple nested levels of packages, with each nesting representing a level of business process decomposition.

This is something I stumbled upon yesterday:
EA reported timeout errors each time I tried to import any package into a nested level of the shared repository - picture packages within packages within packages. I experimented with importing small maps yet the problem persisted.

Then, I tried importing a package that had repeatedly failed to import into a nested package on the server into the top level of my server-based repository. Lo and behold! The import was successful, ending months of being stymied by this problem.

Can someone provide me with any insight as to what might be happening? Nesting packages in multiple levels makes no difference when inporting a package into a local repository - or when someone local to the server imports the same package to its nested location - but makes a huge difference when importing using cloud services.

In researching this topic on this forum, I found this from Geert:

The EA client is very "chatty" with the database server. It sends up to hundreds of small sql queries to the database to get every little piece of information.
So it is not so much the bandwith as the response time that will be the bottleneck when dealing with remote clients.

The only real alternative is to use a central version control system. Then each user can have its local model, and they check-out the parts of the model they want to work on.

Checking in/out takes a while too, but as long as you keep the size of the packages small enough it is a workable solution.


I'm thinking one of the contributing factors may be network latency, although I cannot prove this. I do know the campus where I work has a 1GB fiber pipe to the Internet and supports a thousand or more people. I tested my upload speed and found it to be 2Mb/sec at the time I tested; download speed was about 12x faster. I also know people stream video to their desktop, too, which may have some effect on what I'm trying to accomplish.

By the way, I can export successfully the whole remote repository, which becomes a 100MB XML file, with no timeout issues.

I appreciate any comments / insights.


General Board / EA 12.1 Diagram Template Question(s)
« on: February 04, 2016, 08:11:29 am »
I've created a simple template diagram that contains three (3) BPMN 2.0 elements - Activity, DataStore, DataObject - and one EA problem child (BusinessProcess - the element with the chain link in the bottom-right corner of its icon) - each customized with some variation of size and/or color. I then told EA about its existence via Project->Settings->Project Template Package.

When I use the toolbox to create new activities, data stores and data objects, they render as I want and anticipate. My EA problem child, the BusinessProcess, renders with the default color, the pale whatever-it-is color. Is this an oversight that occurred during EA development or does the BusinessProcess element render using different rules and require me to do something differently to make it render the color I want when I add one to a diagram?

Also, if I drag an existing element created with the customizations from my template diagram from the project browser to another diagram, EA renders that element using that element's default properties. Why is that? If I have a template diagram defined and the element I use exists in my template diagram, it makes sense to me that I want the element properties (color, size, text, font, etc.) I defined in my template diagram to apply to each instance of that element type I place on a diagram, regardless of whether I use the toolbox or drag an existing element from my project browswer to a diagram.

Any chance this behavior will change?

Thanks for any and all responses, except for those that don't meet my expectations or disagree with my sensibilities.   ;)

I've worked with other tools, albeit using different methodologies, that will auto-populate within a decomposed process (akin to a composite activity) its child diagram with the elements connected to its parent process.  Are there any plans to implement this functionality w/in EA?  It sure would help with diagram development.


General Board / Why does EA complain re: location of data objects?
« on: June 30, 2015, 04:26:11 am »
I have three data objects whose movement I'm trying to show through a BPMN 2.0 composite activity diagram.  I'm decomposing higher-level activities into lower-level activities and am creating links from those data objects in parent activities to show where those objects are required in the lower-level activities.  Each data object to which I create a link appears only one time in each diagram and only one time in the project browser. IOWTW, each of the linked data objects in the lower-level activities are the same objects that appear in their parent activities; they are not instances.

When I run the validator in my lower-level diagrams I get warnings that my data objects don't appear in the correct location in my project browser. I understand the warning messages I get, but do these warnings matter?  If so, why? If not, why does EA flag them?

Thanks for your help.


General Board / Insert Other Element... (EA Context Menu)
« on: June 26, 2015, 04:04:24 am »
What is the difference between inserting "Link" and inserting Instance (Artifact)?  Why does EA allow me to insert only one Link to an object per diagram?

Thanks for your help.

Uml Process / UML Class - Notes Box
« on: November 11, 2015, 08:55:35 am »
Is there a maximum size for this field? We're having an issue with text and hyperlinks disappearing from a notes box in one of our classes.  There are a ton of hyperlinks created in this area - we had to jerry-rig a method to create a Word-based report that EA doesn't provide and can't be created through EA scripting, so a VBA for Word program was written to create what we needed and necessitated a class be created to hold hyperlinks that the VBA program reads - and I'm thinking the reason for the random link disappearances is that we're bumping up against the maximum size of this field.


Thanks in advance for your help.

Uml Process / Questions re: BPMN 2.0 Diagram Interpretation
« on: June 26, 2015, 01:55:23 am »
I don't see a forum topic area devoted to BPMN specifically, so I am posting my BPMN questions here.  Please inform me if there's a more appropriate thread / sub-forum for me to post questions about BPMN.  BTW, I know this is an EA forum, and I do have specific questions pertaining to EA that I've posted in the proper threads.

Now for my question:  If a map exists with N number of DataObjects with the same name - e.g. Bob - with each one connected to a different activity via a data association (the intent being to show how one or more physical documents move through a business process), is the correct way to interpret that map that each DataObject named Bob is a separate, unrelated entity and that there are N number of separate "Bobs", each with data associated only to the activity to which an association is indicated?

My thinking is that the "correct" way to illustrate movement of an object/data through a process diagrammed using BPMN 2.0 would be to place the object on a diagram one time and then create the appropriate data input and data output associations between that one DataObject and each activity that requires it.

I know BPMN is not intended to document how DataObjects and information move through a business process, but that's what management wants to see and I'm trying to determine the "correct" or "best" way to illustrate that in my maps.

Thanks for any assistance.

Uml Process / Illustrate in BPMN Object Class and members
« on: June 13, 2015, 04:48:23 am »
The company for which I'm creating BPMN diagrams is just getting started with BPMN; my group is doing nothing with UML.  In fact, our diagrams may never see the light of day as far as development goes.  BUT... My managements wants one diagram that will be suitable for presentation to IT execs and business executives, yet also will be useful for developers and business process improvement folks.  They want to be able to decompose a process into deeper and deeper layers to see more detail.  That being said, I know BPMN is not a decomposition methodology, but perhaps there are techniques I can use to communicate what I need.

Is there a way I can visually illustrate in BPMN, without making executives' eyes cross, that a collection of shipping documents (electronic and/or physical) breaks down into a manifest, an air waybill, certificate of origin, etc. and show how some of this collection will go to one activity and some of it to a differnt activity?  Would I need to create subclasses?  If so, how would I illustrate visually within a BPMN diagram that SubClass A1 and SubClass A2 are children of SubClass A?

I might be on the opposite side of the world from many of you. As I write this, it's 1:37 PM Central Daylight Time on Friday, 12 June 2015, which is probably several hours behind you.  Be that as it may, I won't be able to read or respond to any messages until 15 June 2015 in my location.

Thanks in advance for your suggestions.

Michael Preiss

Uml Process / BPMN 2.0 Biz Process vs. Composite Activity
« on: June 12, 2015, 06:20:07 am »
I'm trying to learn the meaning and correct usage of a BPMN 2.0 business process element (eyeglasses in lower right corner) and the difference between that and a composite activity.  I haven't found anything really meaty in the EA help file and we're having quite the argument spirited discussion about whether and where to use a business process element vs. a composite activity element.

I've noticed a few things about the business process element:
  • You can automatically create a child diagram upon creation of the business process.
  • You cannot draw a sequence connector between an activity and a business process.
  • If you have a business process nested within another business process and run a simulation against the parent process, the simulation will not run against the child process.
Now for the composite activity:
  • Will not automatically create a child diagram upon creation of the composite diagram.
  • If you have a composite activity nested within a business process or a composite activity nested inside another composite activity, a simulation will run against all elements.
Thanks in advance for your help.

Pages: [1]