2012 SD Exam Discussion

[billyjackson768]

If you are sure of that I think now we may both lose a mark. You're 3001 in that case should have been a 4000. Then my 3500, which I am sure for 1 that I wrote down should have been that 2999 that I'm not entirely sure I wrote down. Well, unless the assessors are very generous with accepting answers that still have a sensible justification.

Well everything you wrote is right, it's just that they asked for 4 things not 5.... 4000 and 2999 have the same purpose so you are right it's just I don't see how you wrote 5 down...

I mustn't have. I defiantly didn't add anything extra. I so maybe I didn't include the 2999. My brain is on holiday now though so you'll have to wait till it's back before I can be sure. 

[Uberjew]

I was pretty pleased with the exam. The programming focus was a bit better than previous years too.

I'm also pretty happy with the last two questions of section B - actually having to write out some pseudocode.

Yeah that annoyed me. I'm not sure I did the best job for that. I think I just pretty much made temporary vales to equal length and width then made each of them take on the temporary value of the other so that they swap. I didn't really include much in the way of control structure. Just had the above described within and Begin and End statement.

Begin
A <- 1
B <- 2
Repeat
A <- A + A
B <- B + 1
Until A >= B
End

Did that with same indentation as the pseudo code question later in section C.

[billyjackson768]

I was pretty pleased with the exam. The programming focus was a bit better than previous years too.

I'm also pretty happy with the last two questions of section B - actually having to write out some pseudocode.

Yeah that annoyed me. I'm not sure I did the best job for that. I think I just pretty much made temporary vales to equal length and width then made each of them take on the temporary value of the other so that they swap. I didn't really include much in the way of control structure. Just had the above described within and Begin and End statement.

Begin
A <- 1
B <- 2
Repeat
A <- A + A
B <- B + 1
Until A >= B
End

Did that with same indentation as the pseudo code question later in section C.

I was talking about the other question. But was that one greater than or equal to? -_-

[Uberjew]

I was talking about the other question. But was that one greater than or equal to? -_-

Yes it was.

[billyjackson768]

I was talking about the other question. But was that one greater than or equal to? -_-

Yes it was.

I should probably take your word for that but... I want to at least be able to go to bed tonight dreaming of a 50, so I better not dig to much into everything which I may have already. 

[Yendall]

The case study section asked you to fill in a UCD (first time ever) and a DFD (first in a long time). What was your answer to the DFD and UCD? Also, the test table asked for values to test the requirements form 3000 to 3999, what values did you provide? Does that imply inclusive, exclusive or one of each?

I did 2999, 3000, 3001 and 3999.  That tests all boundaries, expected data and invalid data.

The conditions were <3000 and >3999 so my test data was:

• 2999 (<3000)
• 3000 (=3000 AND <3999) [Lower Bound]
• 3999 (=3900 AND > 3000) [Upper Bound]
• 4000 (>3999)

Your test data doesn't account for >3999

[Uberjew]

The case study section asked you to fill in a UCD (first time ever) and a DFD (first in a long time). What was your answer to the DFD and UCD? Also, the test table asked for values to test the requirements form 3000 to 3999, what values did you provide? Does that imply inclusive, exclusive or one of each?

I did 2999, 3000, 3001 and 3999.  That tests all boundaries, expected data and invalid data.

The conditions were <3000 and >3999 so my test data was:

• 2999 (<3000)
• 3000 (=3000 AND <3999) [Lower Bound]
• 3999 (=3900 AND > 3000) [Upper Bound]
• 4000 (>3999)

Your test data doesn't account for >3999

Okay well 2999 and 4000 serve the same purpose in my reasoning because they're both invalid data.  3000 and 3999 are borderline, there needs to be something in between.  This is were i think reasoning comes in because generally you want before lower border, lower borer, inbetween borders, upper border and above upper border.

Technically 3000 and 3999 are both in the allowed values so they would count as just one.

[Yendall]

3000 and 3999 are between?

[Uberjew]

3000 and 3999 are between?

Well technically anything below 3000 and above 3999 isn't in the allowed values so as those 2 are, they technically would serve the same purpose.  The border is really not existent as it's not an equal to sign.

My reasoning for just using 2999 and not 4000 was that it was enough to test invalid data.

[Yendall]

It's not simply to test "invalid data" as a value. You have to test all conditions. You need to make sure that anything under 3000 and over 3999 are invalid. If you don't test over 3999, how do you know that if someone inputs a number over 3999, like 4000, that it's going to return the correct value? There could be a logic error or something that you will never know because you didn't test it.
The boundaries are between: 2999 and 4000 at a minimum.
You need to test at least one value above and below to test both conditions, and at least two between.

[Uberjew]

It's not simply to test "invalid data" as a value. You have to test all conditions. You need to make sure that anything under 3000 and over 3999 are invalid. If you don't test over 3999, how do you know that if someone inputs a number over 3999, like 4000, that it's going to return the correct value? There could be a logic error or something that you will never know because you didn't test it.
The boundaries are between: 2999 and 4000 at a minimum.
You need to test at least one value above and below to test both conditions, and at least two between.

I used 2999 as my test and invalid data as my reasoning.  I should be fine imo

[Yendall]

It's not simply to test "invalid data" as a value. You have to test all conditions. You need to make sure that anything under 3000 and over 3999 are invalid. If you don't test over 3999, how do you know that if someone inputs a number over 3999, like 4000, that it's going to return the correct value? There could be a logic error or something that you will never know because you didn't test it.
The boundaries are between: 2999 and 4000 at a minimum.
You need to test at least one value above and below to test both conditions, and at least two between.

I used 2999 as my test and invalid data as my reasoning.  I should be fine imo

We'll see. In my eyes you didn't test all conditions.

[Uberjew]

It's not simply to test "invalid data" as a value. You have to test all conditions. You need to make sure that anything under 3000 and over 3999 are invalid. If you don't test over 3999, how do you know that if someone inputs a number over 3999, like 4000, that it's going to return the correct value? There could be a logic error or something that you will never know because you didn't test it.
The boundaries are between: 2999 and 4000 at a minimum.
You need to test at least one value above and below to test both conditions, and at least two between.

I used 2999 as my test and invalid data as my reasoning.  I should be fine imo

We'll see. In my eyes you didn't test all conditions.

Well i think that you need something between 3000 and 3999 which is expected, usual, non-boundary values.

[Yendall]

It's not simply to test "invalid data" as a value. You have to test all conditions. You need to make sure that anything under 3000 and over 3999 are invalid. If you don't test over 3999, how do you know that if someone inputs a number over 3999, like 4000, that it's going to return the correct value? There could be a logic error or something that you will never know because you didn't test it.
The boundaries are between: 2999 and 4000 at a minimum.
You need to test at least one value above and below to test both conditions, and at least two between.

I used 2999 as my test and invalid data as my reasoning.  I should be fine imo

We'll see. In my eyes you didn't test all conditions.

Well i think that you need something between 3000 and 3999 which is expected, usual, non-boundary values.

I would agree with you if the boundaries were : <= 3000 and >= 3999.