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 - fwoolz

Pages: 1 2 3 [4] 5 6 7
46
Suggestions and Requests / More XML Schema Stuff
« on: April 22, 2003, 01:29:47 pm »
A coupla things...

1. It would be nice to be able to generate just type definitions from classes rather than elements + types; this would be helpful when generating supporting schemas.
2. Some means of specifying elementless tags (e.g., <xs:import...> ) would be a nice to have; again, this would be useful in importing supporting schemas.
3. Again with the supporting schemas... the "package barrier" that seems to exist in the current implementation of the EA schema generator is not necessarily a bad feature, since it is consistent with partitioning a design into separate namespaces/schemas- but some means of importing (via <xs:import ...> is needed.

Cheers,
Fred Woolsey

PS: Great job, BTW- who needs XMLSpy!!!

47
Suggestions and Requests / Target Namespace Prefix in 3.51.608
« on: April 22, 2003, 12:56:15 pm »
Howdy,

It appears that the "targetNamespacePrefix" tag does not get carried over into the generated schema, although the "targetNamespace" information does.  In addition, if elementFormDefault is set to "qualified", the prefix does not get added to element references.  Is there a trick I am missing, or is this a glitch?

Cheers,
Fred Woolsey

48
Hi All,

Here are a few feature requests and one bug fix request:

-- Both of these requests arise from difficulties I've had in arranging use cases inside of collaboration ovals ---
1. Use case & collaboration ovals: Should have independent vertical & horizontal resizing of use cases & collaborations (allow oval to have any eccentricity/aspect ratio).
2. Use case/collaboration label: should be able to move it from the center of the oval to wherever desired.
-- These are of a more functional nature, and are based on work I've done in trying to represent network messages between nodes in a distributed system ---
3. Sequence diagram messages: Need to be able to have ability to place messages between objects at exactly the same height (i.e., the same time).
4. Sequence diagrams messages: Add support for downward sloping messages to provide visual presentation of non-instantaneous messages.  NB: This is permitted by the UML standard.
-- This would help alleviate difficulties dealing with many objects on a single diagram --
5. Sequence diagram messages: Add “off-diagram” reference to linked sequence diagrams.
-- This is a minor annoyance, but it would be nice to have it fixed! --
6. Code generation: Fix problem where initial value of an attribute isn’t reflected in code, and initial value in code doesn’t show up after synching model.

Cheers & Thanks,
Fred Woolsey

49
General Board / Where have all the members gone?
« on: July 07, 2003, 08:36:20 am »
Hi all,

The Forum members list seems to have shrunken to 17 members... whahoppen?  Looks like a possible "undocumented feature" of YaBB?

Cheers,
Fred Woolsey

50
General Board / Bookmarks in Hyperlinks
« on: May 24, 2003, 11:04:09 am »
Hello All,

I have attempted to get EA hyperlinks to recognize URL bookmarks, e.g., "file:///c:/projects/spec.htm#S1_1" without success.  EA fires up the associated file all right, but the bookmark (in this case, "#S1_1") appears to get stripped off of the URL before it is passed to the server app (browser in this case).  Am I doing something wrong, or is this just the way EA works?  If it is the latter, I would request that future releases recognize bookmarks in hyperlink addresses.

Cheers,
Fred Woolsey

51
General Board / Looking Ahead- EA and UML 2.0
« on: March 08, 2003, 04:25:47 pm »
Hi all (& Geoff in particular),

Are Sparx Systems tied into the developing standard for UML 2.0, and are there plans to migrate EA to that standard once it's issued?

Cheers,
Fred Woolsey

52
General Board / New Feature/Bug Fix Request
« on: April 06, 2003, 04:59:48 pm »
Hi All,

Here are a few feature requests and one bug fix request:

-- Both of these requests arise from difficulties I've had in arranging use cases inside of collaboration ovals ---
1. Use case & collaboration ovals: Should have independent vertical & horizontal resizing of use cases & collaborations (allow oval to have any eccentricity/aspect ratio).
2. Use case/collaboration label: should be able to move it from the center of the oval to wherever desired.
-- These are of a more functional nature, and are based on work I've done in trying to represent network messages between nodes in a distributed system ---
3. Sequence diagram messages: Need to be able to have ability to place messages between objects at exactly the same height (i.e., the same time).
4. Sequence diagrams messages: Add support for downward sloping messages to provide visual presentation of non-instantaneous messages.  NB: This is permitted by the UML standard.
-- This would help alleviate difficulties dealing with many objects on a single diagram --
5. Sequence diagram messages: Add “off-diagram” reference to linked sequence diagrams.
-- This is a minor annoyance, but it would be nice to have it fixed! --
6. Code generation: Fix problem where initial value of an attribute isn’t reflected in code, and initial value in code doesn’t show up after synching model.

Cheers & Thanks,
Fred Woolsey

53
General Board / Initial value doesn't survive round-trip
« on: March 29, 2003, 04:05:36 pm »
Hi all,

One problem that has consistently plagued EA since I began using it (and a fine product it is!) is the disappearance of attribute initial values in code generation and synchronization.  Initial values in EA don't make it to my C++ code, and initial values in C++ don't make it to EA.  Am I missing something, or is this something that needs fixing?

Cheers,
Fred Woolsey

54
General Board / Concurrent Messages in Sequence Diagram
« on: January 31, 2003, 04:07:30 pm »
I've downloaded the new release of EA (3.51)... looks great!  I am particularly glad to see that concurrent messages are now possible in sequence diagrams, since I have recently been frustrated trying to set up concurrent activities in asynchronous objects.  However, after trying to set up messages at the same height, I encountered the same problem as before- messages flip past each other, arranging themselves at different heights.  What am I missing?

Thanks,
Fred Woolsey

55
General Board / Feature Request: Tree View for All Relationships
« on: January 20, 2003, 05:47:23 am »
Hi all,

While it is possible to arrange relationships in a tree view automatically for generalizations, this feature does not seem to be supported for other relationship types.  It would be nice if all relationships could have an automatic tree view option as well... otherwise, you have to set up the relationship ends so that they overlap exactly.

Cheers,
Fred Woolsey

PS: It is possible to start out with generalizations in tree view, then change each relationship to a different type; this produces the desired result.  However, any changes to the diagram that shift relationship anchor points make it necessary to align the ends manually.

56
General Board / Grabbing Time Constraints from EA
« on: January 07, 2003, 06:40:10 am »
Hi All,

Is there a (relatively) easy way to get time constraints out of an EA model (assuming you put them in there in the first place!).  I am thinking here of some automation method that will grab certain constraint values or tag values (e.g., "Time = XX ms") from operations in a state or activity diagram.  The goal would be to use the results to do deadline sanity checking, schedulability analysis, etc.

Cheers and Best Regards,
Fred Woolsey

BTW: Happy New Year to All!

57
General Board / Suggestions for Additional Diagram Elements
« on: December 20, 2002, 02:26:09 pm »
Hi all,

While working with EA (a great tool, by the way), I've noted some missing diagram elements that would make life a little easier if they were included in a future revision to EA, to wit:

1. Junction pseudo-states, represented by a small filled-in circle

The UML 1.4 standard says:

3.81.2 Notation
Two or more transitions emanating from different non-concurrent states or pseudostates can terminate on a common junction point. This allows their respective
compound transitions to share the path that emanates from that junction point. A junction point is represented by a small black circle.


Diamonds w/o guards on outgoing transitions could be used per the standard [Alternatively, it may be represented by a diamond shape (see Section 3.87, “Decisions,” on page 3-160)], but these could be confused with decision points.

I currently produce these by setting the border width of a synch state to a value that fills the circle in completely.

2. Synch states with finite capacity

Per the standard:

3.83.2 Notation
A synch state is shown as a small circle with the upper bound inside it. The bound is either a positive integer or an asterisk ('*') for unlimited. Synch states are drawn on the boundary between two regions when possible.


Currently, synch states have only infinite capacity '*'.  Implementing the ability to set the capacity to a finite integer value provides the ability to define the equivalent of Petri net "places" with a finite token capacity.

3. Visible labels on fork and join bars ( V & H synch) to allow stub states to be shown conveniently

Per the standard:

3.82.2 Notation
The submachine state is depicted as a normal state with the appropriate “include” declaration within its internal transitions compartment (see Section 3.75, “State,” on page 3-138  ). The expression following the include reserved word is the name of the invoked submachine. Optionally, the submachine state may contain one or more entry stub states and one or more exit stub states. The notation for these is similar to that used for stub ends of stubbed transitions, except that the ends are labeled. The labels represent the names of the corresponding substates within the invoked submachine.


4. Substate, sub-activity icon

(Section 3.76.2)
In some cases, it is convenient to hide the decomposition of a composite state. For example, the state machine inside a composite state may be very large and may simply not fit in the graphical space available for the diagram. In that case, the composite state may be represented by a simple state graphic with a special “composite” icon, usually in the lower right-hand corner. This icon, consisting of two horizontally placed and connected states, is an optional visual cue that the state has a decomposition that is not shown in this particular statechart diagram (Figure 3-74 on page 3-142)

Not a big deal, but it is a useful visual cue.

5. Dynamic choice points

(Section 3.81.2)
Two or more guarded transitions emanating from a common dynamic choice point are used to model dynamic choices. In this case, the guards of the outgoing transitions are evaluated at the time the choice point has been reached. The value of these guards may be a function of some calculations performed in the actions of the incoming transition (s). A dynamic choice point is represented by a small white circle (reminiscent of a small state icon).

6. Custom Concurrent Substate dividers

Similar to custom routing for connectors, it would be helpful to be able to resize concurrent substate dividers independently of other substates; use a combination of vertical and horizontal dividers, etc.

Enuf for now...

Cheers and Best Regards,
Fred Woolsey

58
General Board / C++ Code Generation Bugs
« on: December 12, 2002, 05:28:12 pm »
Hi all,

Just a note concerning a few glitches I've encountered generating C++ code & round tripping it (in EA 3.50.585):

- Class notes (comments) sometimes get duplicated several times in the generated code.
- After generating code, then renaming a class, code generation creates constructors/destructors with both the new and old class names.  This happens even after deleting the earlier source files.
- Similarly, after generating code, then renaming methods, method prototypes are generated for both the old & new names.
- Importing a declaration such as Class* m_Class[10] from a source file seems to work OK, but subsequent code generation produces Class* [10] m_Class.

Thanks...
Fred Woolsey

59
General Board / Accessor/Mutator methods not generated?
« on: December 11, 2002, 03:29:05 pm »
Hey y'all,

Using EA 3.50.585 & C++, I can't seem to get the code generator to produce the gets & sets for private attributes automatically.  Any advice?

Thanks,
Fred Woolsey

60
General Board / Comments <<echo, echo, echo...>>>
« on: December 11, 2002, 03:22:01 pm »
Hey all,

I am running EA 3.50 build 585 (Enterprise Ed.) and have noticed that notes for a class get duplicated each time the code for that class is generated (generation options: C++, Javadoc comments).  After several runs, you might wind up with 5 copies of the same comments.  Not a major problem, but a trifle annoying.

Cheers,
FCW

BTW: EA is a GREAT value... who needs Rhapsody in C++ when you can have peace and harmony in EA?

Pages: 1 2 3 [4] 5 6 7