Answer thread
Sean Q. Lau
First, I will not be responding to e-mails. If you want a response, use
the newsgroup.
Second, for posts which people have said there have been no response in
the newsgroup - Professor Kontogiannis said that these questions have been
brought up and addressed during lecture. If you have any pressing
questions, I suggest you bring them to lecture.
Third, we will not comment on any specifics about the demo information,
other than that which is provided already.
Answers to previous threads below.
---
Also for "make sure manual bill generation works for a given time period",
what exactly does this mean? Does this mean we need to have something
in our UI that allows you to manually issue a bill between 2 specified
dates/times? Or does this mean you can issue a bill at any time and have
a new one start for that account after that?
> Ideally, it means exactly what is stated - we can generate a bill for
any period of time; however, given the huge variation in people's
implementation, we will settle for being able to generate a bill at the
moment it is requested.
---
For the 3 unit tests that will be selected and executed, does this mean
3 out of the 6 unit tests (3 JUnit, 3 Framework) or do you mean 3 out of
the 15 tests(if 5 per method) in JUnit or Framework?
> It means exactly what it says on the info sheet.
---
Additionally, I'd like to point out that in the thread titled "Questions
for Project Part 2", Professor Kontogiannis specifically mentioned that
only JUnit or manual tests are required for system testing. Also, in
speaking with Dr. Lian today, he further gave us the impression that only
manual tests are expected for system testing.
So, are automated tests necessary, as stated in the demo guidelines? It
seems a little unreasonable to be stating this requirement only four
days prior to the due date.
> This is not a new requirement. From the very beginning, system tests
were required to be implemented using JUnit. If this was not clear from
the assignment description, then you should have clarified it with the
instructor. The only cases which presented difficulty with JUnit in which
we were made aware of were those groups which were experiencing problems
with threads, and Professor Kontogiannis said those groups may have the
system tests run manually. If he has explicitly said otherwise in lecture,
then feel free to run the system test manually.
---
When in the spec it says "You will exercise the inferred test cases using
both a) JUnit or CPPUnit toolkits, and b) the testing framework you
implemented in the first part of the project", does this mean for EVERY
test case for ALL methods tested (excluding system tests as mentioned in
early posting) we implement the test case in both JUnit/CPPUnit
AND our framework? As opposed to doing some tests in JUnit and others in
the framework?
> The test cases which are implemented must be done in both JUnit/CPPUnit
AND your testing framework.
Jeremy Karn
> Second, for posts which people have said there have been no response in
> the newsgroup - Professor Kontogiannis said that these questions have been
> brought up and addressed during lecture. If you have any pressing
> questions, I suggest you bring them to lecture.
>
I don't think I'm the only person that has a problem with this. Is there
a particular reason that answers aren't being posted to the newsgroup?
Not only are answers not posted to the newsgroup, but it seems that
additional project requirements/clarifications are announced in class, but
not posted to
the newsgroup, not posted on the web page, and not added to the project
requirement documents. One example being the requirement to submit a hard copy of the
SRS/SDD for the first deliverable which was only ever mentioned in
lecture, and never announced anywhere else.
For a course that is one of three that has stressed documentation, why is
it that the various requirements/clarifications of this project aren't
being documented anywhere? If nothing else, it would be nice to have a
permenent reference of clarifications made so that there are no
misunderstandings about what was said, and what is expected for this
project.
Jeremy Karn
Robert Kitts
> Additionally, I'd like to point out that in the thread titled "Questions
> for Project Part 2", Professor Kontogiannis specifically mentioned that
> only JUnit or manual tests are required for system testing. Also, in
> speaking with Dr. Lian today, he further gave us the impression that only
> manual tests are expected for system testing.
>
> So, are automated tests necessary, as stated in the demo guidelines? It
> seems a little unreasonable to be stating this requirement only four
> days prior to the due date.
>
>> This is not a new requirement. From the very beginning, system tests
> were required to be implemented using JUnit. If this was not clear from
> the assignment description, then you should have clarified it with the
> instructor. The only cases which presented difficulty with JUnit in which
> we were made aware of were those groups which were experiencing problems
> with threads, and Professor Kontogiannis said those groups may have the
> system tests run manually. If he has explicitly said otherwise in lecture,
> then feel free to run the system test manually.
In case anyone is unsure, I clearly remember Prof. Kontogiannis saying in
class that the JUnit requirement for System Tests was no longer a
requirement, and all groups were free to perform them manually.
I don't think we're going to get this in writing confirmed anywhere
though. *grumble* *grumble*
Robert
Andrew L Roth
On Thu, 23 Mar 2006, Sean Q. Lau wrote:
>
> First, I will not be responding to e-mails. If you want a response, use
> the newsgroup.
This is reasonable, because everyone can benefit from the answer.
However, I will be asking questions specific to our implementation.
> Second, for posts which people have said there have been no response in
> the newsgroup - Professor Kontogiannis said that these questions have been
> brought up and addressed during lecture. If you have any pressing
> questions, I suggest you bring them to lecture.
This is unreasonable.
It is reasonable that things said in class be recorded in writing.
If I did something on an assignment, and said "but you said in class we
didn't have to", would you believe me? Maybe. If I said, "on the
newsgroup post subject ... on date ..., it said we didn't have to", then
you definitely would. The latter situation is better.
It is reasonable to expect clarifications to be recorded in writing on the
newsgroup or web page.
Thank you for posting some answers here.
> Third, we will not comment on any specifics about the demo information,
> other than that which is provided already.
This is also unreasonable.
If the demo information is ambiguous, why shouldn't they be clarified? We
are expected to be prepared for the demo; we can't do that if we don't get
clarifications on ambiguous parts. It's also unreasonable to say it
doesn't need clarifications; past assignments certainly have, why not this
one?
Please see specific questions below.
>> Ideally, it means exactly what is stated - we can generate a bill for
> any period of time; however, given the huge variation in people's
> implementation, we will settle for being able to generate a bill at the
> moment it is requested.
Our system allows us to make a bill for a specific month and year. Is
that adquate?
> For the 3 unit tests that will be selected and executed, does this mean
> 3 out of the 6 unit tests (3 JUnit, 3 Framework) or do you mean 3 out of
> the 15 tests(if 5 per method) in JUnit or Framework?
>
>> It means exactly what it says on the info sheet.
Let's put it another way. Are our framework tests considered unit test?
> So, are automated tests necessary, as stated in the demo guidelines? It
> seems a little unreasonable to be stating this requirement only four
> days prior to the due date.
>
>> This is not a new requirement. From the very beginning, system tests
> were required to be implemented using JUnit. If this was not clear from
> the assignment description, then you should have clarified it with the
> instructor.
The first time the demo was introduced in class, Professor Kontogiannis
specifically said we didn't have to automate the tests. This is not a
new clarification / point of confusion.
As mentioned on March 10, in "Re: Questions for Project Part 2" by Prof.
Kontogiannis, in response to "For system tests, what will we be using
to perform/automate the test? Our own framework? JUnit? Just some
extra code? (Or perhaps even manual test?)", said "Only JUnit or manual
test as required, but not your framework."
In http://www.student.cs.uwaterloo.ca/~sqlau/Demo.pdf, under "Demo Flow",
"Part II - Automated Tetsing" has "3 system tests will be executed". This
contradicts what Prof. Kontogiannis said and posted!
-Andrew Roth
Daniel Urquhart
for System Tests would not be required. That said, can I prove it? No,
since:
>Second, for posts which people have said there have been no response in
>the newsgroup - Professor Kontogiannis said that these questions have been
>brought up and addressed during lecture. If you have any pressing...
"Robert Kitts" <rki...@student.cs.uwaterloo.ca> wrote in message
news:Pine.GSO.4.64.06...@cpu18.student.cs.uwaterloo.ca...
Philip Mitchell
WANT to put stuff in writing? So that a) you know what you've told us,
b) we will stop asking you questions over and over again, and c) we can
put together a project that has a hope of meeting your criteria (making
our marks better and making it overall far easier for you to mark). Sure,
putting stuff in writing causes you to have to honour your statements, but
I don't really think that's a bad or difficult thing, nor am I accusing
anyone of trying to deceive us. Occasionally something could be mentioned
to a single group and everyone just assumes everyone was informed.
Sometimes people hear things incorrectly or forget mundane details.
Mistakes happen. Putting things in writing is the only way to avoid such
mistakes.
Perhaps this is actually part of the project. One thing engineers have to
fight with constantly is employers who don't want to put things in writing
so that when something goes wrong, the employer can't be blamed. Perhaps
this is meant to give us experience with such things (I'm not actually an
engineer, though...also the analogy is poor, because we're the ones who
are paying...if I see this sort of thing from someone I'm paying, I'm not
inclined to deal with them again).
Forgive my cynicism, but after 3 terms of this sillyness, I'm somewhat
tired of poorly specified project requirements. The questions that have
been asked are honest and good requests for clarification where there is
vagueness or conflicting information in the specs. Giving us clear
responses in writing (that we may reference them later) is, it would seem
to me, the least you can do.
Hopefully this isn't taken as childish whining or flaming, as the tone I'm
going for is constructive criticism of, what many of us feel are poor
decisions. This is a genuine concern that we are trying to address.
Thanks,
Phil