This week's giveaway is in the EJB and other Java EE Technologies forum. We're giving away four copies of EJB 3 in Action and have Debu Panda, Reza Rahman, Ryan Cuprak, and Michael Remijan on-line! See this thread for details.
Yesterday, I was reviewing domain diagram given in assignment. In domain diagram, cardinality (1,*) displayed for every association. It looks some cardinality for some associations are wrong. Can we override those cardinality? Did somebody face simillar issue?
It looks difficult to model application with given cardinality.
As per Humphrey Sheil's book Chpt 9 which is a bible for all SCEA Part2 takers you should stick to the domain diagram as closely as possible.
Sometimes it may be sub-optimal.
However it is the only basis the examiner has to evaluate the solution.
Examinees are obviously not able to be present physically to justify changes to domain diagrams.
Of course if you can make changes and provide suitable explanation, I daresay that would be fair enough.
My 2c :-)
~ Ravi Iyer, SCEA
Joined: Sep 26, 2010
Thanks, Ravi for your advice.
I think, I need to do brain storm to make myself comfortable with domain diagram.
Joined: Oct 24, 2010
Best of luck Prashant. I'm stkill waiting for my results. Not too confident since I feel I have missed quite a few things people on this forum discuss. Overall I have found the level of difficulty and detail of SCEA 1,2,3 impressive. They have done a great job designing an exam which can be prepared for mainly by experience. Books and forums like these help but experience I have found is greatest help.
I am not privy to all the opinions of the writer mentioned before; one of which, I came across perchance, I beg to differ with. In fact, Sun should declare a writer to be the official spokesperson regarding their exam rules, if they want the test takers to follow that writer, without which, it is not a fair to expect the test takers to adhere to such standards. For example, if a test taker has not read or believed my previous responded, s/he may be discriminated unjustifiably, is my concern.
Coming to the basic question, if you think that the domain model or any part of the assignment is at fault, you should first try to gather solid reasons for opinion (e.g. reference to referred books/journals, various standards, etc). Then, you should boldly make them out in your assumptions, correct the things as they should be, and proceed.
My justification in suggesting the above course of action is that otherwise, you might take a less optimized path or wait forever to find a solution to an erroneous situation.
However, you may also write back to Sun Education regarding the problems and wait for a reasonable time for their response, if you want to be absolutely sure of what you should do, before proceeding.
Pranab Das, PMP, SCEA
Joined: Oct 24, 2010
Good words :-) Das. I'm not suggesting examinees develop a solution around erroneous domain diagrams. Far from it what appears sub-optimal may have reasons which take time to understand. It may be faster to develop a better solution with a more streamlined DD but harder to explain the same to your assessor who you won't meet. The DD may be tweaked to introduce a "trick scenario" that might need accounting for in our eventual solution. Hence my suggestion that change to DD be not a first but rather last option. Again, just my opinion.
Joined: Jun 30, 2008
No dialog; just to clarify that I did not refute anything my previous respondent said; my question was about certain writer of a book that claims to be providing insider information--in the sense that it is not contained in the Sun's official communications to the test takers--which appears to be discriminatory to me, as explained before.
Moreover, many problems can be tackled in various ways; if one can fault something--with sufficient reasons--that another treats as gospel, there should be nothing wrong, IMHO.
P Das wrote:However, you may also write back to Sun Education regarding the problems and wait for a reasonable time for their response, if you want to be absolutely sure of what you should do, before proceeding.
Which will be a long wait with (probably) no response.
As the architect of this solution, you are expected to be the expert, and you are the person others are looking to for the decision.
Speaking from my own experience in Part 2, I changed the cardinality in 2 places. I wrote a detailed paragraph justifying my decision in the Design Assumptions section at the top.
Few days later I passed.
I think the only rule is that you should not neglect any BDM objects completely. You can add/modify/replace, but dont eliminate them completely.
Bottom line - Like Andrew said, you are the architect. If you feel that there is something that can be done better, then do it and justify your decision. There is no skill required in copying something 1:1. Even my manager can do that