Predict output

Question

Predict output

Viewing 15 posts - 31 through 45 (of 54 total)

You must be logged in to reply to this topic. Login to reply

Richard Farris SSC Enthusiast Points: 146 More actions · Answer 1

I agree with the others here. The proffered answers were not adequate because there was too much ambiguity on what should be considered "output". If you are expecting that "output" is the result of the SELECT, then there was no correct answer. If you expected "output" to mean both messages and results, then the supposed correct answer is also not accurate.

This question was just poorly worded. Try... Catch is a concept that is very important to understand though, so I appreciate the intent.

Kindest Regards,Rick Farris

Fatal Exception Error SSChasing Mays Points: 609 More actions · Answer 2

Lynn Pettis (1/20/2010)
It isn't just the QotD that you can learn from, but also the discussion that may follow (as long as you ignore the complaints).
Does SSC vet the questions, to a point, but not nearly as well as some people would like to see. Of all the questions I have answered, I think I only really had a serious complaint about one or two. Yes, there are issues with some of the questions, but if you look closer at the concept that the question is trying to address, and less on the question and answers themself, I think you will learn something.
Take today's question. Based on the answers, you'd have expected there to be PRINT statements in the question but there wasn't. If you understand the concept of the TRY/CATCH, and that the divide by zero would be caught and handled in the CATCH part of the code, there really was only one answer that made sense, and it happened to be the only one with the message from the PRING statment in the CATCH block.

I strongly disagree. I am not trying to be a jerk but this is not the first time people have complained about ambiguous answers from arup.

What concept do new people get from incorrect answers as options? None of the answers fit at all. The fact that you assumed you knew what the author meant and selected your answer based on an assumption and not knowing proves this. What if I assumed the author missed a PRINT @i.

I learned nothing from this except:

-there can be mingling between result set and messages in the answer for a QoTD.

-It is now possible for an integer field to display a string in the answer for the QoTD.

-Ambiguity is now acceptable for listed answers to the QoTD just like irregardless is now accepted as a word by dictionaries.

Lynn Pettis SSC Guru Points: 442467 More actions · Answer 3

Fatal Exception Error (1/20/2010)
Lynn Pettis (1/20/2010)
It isn't just the QotD that you can learn from, but also the discussion that may follow (as long as you ignore the complaints).
Does SSC vet the questions, to a point, but not nearly as well as some people would like to see. Of all the questions I have answered, I think I only really had a serious complaint about one or two. Yes, there are issues with some of the questions, but if you look closer at the concept that the question is trying to address, and less on the question and answers themself, I think you will learn something.
Take today's question. Based on the answers, you'd have expected there to be PRINT statements in the question but there wasn't. If you understand the concept of the TRY/CATCH, and that the divide by zero would be caught and handled in the CATCH part of the code, there really was only one answer that made sense, and it happened to be the only one with the message from the PRING statment in the CATCH block.
I strongly disagree. I am not trying to be a jerk but this is not the first time people have complained about ambiguous answers from arup.
What concept do new people get from incorrect answers as options? None of the answers fit at all. The fact that you assumed you knew what the author meant and selected your answer based on an assumption and not knowing proves this. What if I assumed the author missed a PRINT @i.
I learned nothing from this except:
-there can be mingling between result set and messages in the answer for a QoTD.
-It is now possible for an integer field to display a string in the answer for the QoTD.
-Ambiguity is now acceptable for listed answers to the QoTD just like irregardless is now accepted as a word by dictionaries.

Okay. I'll shut up now.

KevinC. SSCommitted Points: 1570 More actions · Answer 4

Just so you know, Lynn, I appreciate your insights, especially on items such as the QotD.

I did find the answers a little ambiguous, but I learned a bit more about analysis.

Thanks.

Kevin C.

SQLRNNR SSC Guru Points: 281334 More actions · Answer 5

Lynn Pettis (1/20/2010)
mzz3lh (1/20/2010)
I have to agree here. None of the answers match what I would expect, which would be 'This is an exception' in the message pane, followed by the contents of the temp table: 1, 2, 4, 5.
However, when I ran this to try it out, I got an error:
(1 row(s) affected)
(1 row(s) affected)
this is an exception
Msg 208, Level 16, State 0, Line 10
Invalid object name '#temp'.
Which indicates that despite the try-catch block, the error caused the temp table to be dropped mid-execution.
Or that something was wrong with the code you ran. It ran fine for me, and it appears from a few other posts that it ran fine for others.
I agree, the answers may not have been displayed as many of us have seen it when we ran the code. The question is, were you able to correctly predict the output? In this case, as mentioned in another post, there really was only one option that could be correct. It may have been more difficult had there been one or two more possible answers with the exception message included.

Agreed here, there is only one answer that supplies both pieces of the answer. All other answers were invalid.

Settings in your SSMS would cause slightly different output. One scenario includes set nocount, and whether or not it is on or off.

set nocount on gives me the following in the message pane:

this is an exception

set nocount off gives the following

(1 row(s) affected)

this is an exception

(1 row(s) affected)

(4 row(s) affected)

I think being overly picky about the presentation of the available answer in this scenario is going beyond what the question was trying to derive.

We have been shown that both parts of the correct answer are only available in one of the answers.

Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events

Richard Farris SSC Enthusiast Points: 146 More actions · Answer 6

Agreed, Jason. The intent of the question was good and purposeful, no doubt.

Funny that most of us, as technologists, are so insistent on a perfectly valid scenario. 🙂 The point wasn't lost on me, but I am adamant about well thought-out surveys. I remain left to wonder, for example, how much thought went into this question and for what reasons. This is and will be my main reason for engaging in this discussion, as has been my practice in prior discussions.

Kindest Regards,Rick Farris

prahaladd Grasshopper Points: 20 More actions · Answer 7

Nice QoTD :-).

An important point to remember is that a TRY..CATCH can be used to catch all database errors having the error number >10 that do not terminate a database connection.

MSDN has more information on this.

http://msdn.microsoft.com/en-us/library/ms175976.aspx

pg53 Ten Centuries Points: 1303 More actions · Answer 8

Apart from that, what's going on with the transaction? Nobody's mentioned that, but isn't that kind of weird? (Don't shoot me if it's obvious, I'm primarily a .Net guy rather than a DBA, so Try . . .Catch was familiar ground; but that rollback in the middle - can you do that?)

nomlad SSC Enthusiast Points: 152 More actions · Answer 9

that's because the 'correct' answer wasn't correct.

vk-kirov SSCertifiable Points: 7686 More actions · Answer 10

pg53 (1/21/2010)
Apart from that, what's going on with the transaction? ... that rollback in the middle - can you do that?

You can roll back a transaction in the middle, in the end, or even in the beginning of a batch – it's no problem with that 🙂

Saurabh Dwivedy SSCommitted Points: 1887 More actions · Answer 11

I too, like most others, initially felt inclined to believe none of the answers was PERFECTLY CORRECT because the text - in the middle of the answer set - was a bit misleading - mainly because everyone expected the OUTPUT to come from the select * from #temp, which could not "emit" the output displayed in the answer choices.

However, I chose the right answer because there didn't seem to be any other answer coming close. The "right answer" had all the ingredients though not exactly in the 'proper way'.

I see both sides of the argument - one asserting that "There is no correct answer" and the other inviting us to focus on the "larger picture".

My only submission is for us to be a little more considerate for the poster. Someone has spent some time and effort creating the question. Creating is not an easy business and I quite liked the frankness displayed by another poster on the forum who openly accepted he wasn't quite wired for doing such a thing. Let's be a little considerate, folks. Let's also keep in mind that even those posting the questions may not be experts at creating questions. I don't think we have folks dedicated here to proof-read questions thoroughly before the questions are printed. If we are so irked by "inappropriate/incorrect" question why don't we form a volunteer group (on a revolving basis maybe) who would do the needful.

Somebody brought GRE type questions into the picture which was again very surprising. This is a forum for learning and sharing. Let's give kudos and help make this forum a great learning experience for everyone.

If anything - pure criticism without being properly backed by a plan of action to fix the source of criticism is a sheer waste of time IMHO.

Regards

Saurabh

Saurabh Dwivedy
___________________________________________________________

My Blog: http://tinyurl.com/dwivedys

For better, quicker answers, click on the following...
http://www.sqlservercentral.com/articles/Best+Practices/61537

Be Happy!

prahaladd Grasshopper Points: 20 More actions · Answer 12

Hi,

The CATCH clause in SQL server does exactly what a catch(Exception ex) in C# does. Also the ROLLBACK is perfectly valid; imagine what would happen if there was no TRY.. CATCH within the given sample code. In that case after opening a transaction, there would be a divide by zero error and this is analogous to an error in the transaction which would eventually result in it's roll back.

Kelsey Thornton SSCrazy Points: 2157 More actions · Answer 13

Saurabh Dwivedy (1/21/2010)
<snip>
My only submission is for us to be a little more considerate for the poster. Someone has spent some time and effort creating the question. Creating is not an easy business and I quite liked the frankness displayed by another poster on the forum who openly accepted he wasn't quite wired for doing such a thing. Let's be a little considerate, folks. Let's also keep in mind that even those posting the questions may not be experts at creating questions.
<snip>
Regards
Saurabh

Hear hear!

Kelsey Thornton
MBCS CITP

ChiragNS One Orange Chip Points: 26137 More actions · Answer 14

Learnt some thing . But does'nt Set ansi_warning have an impact.

"Keep Trying"

HardCoder Ten Centuries Points: 1045 More actions · Answer 15

Output answer should be

1

2

4

5

why this option was not there??

Mohammad Irfan
http://matespoint.blogspot.com
http://www.irfit.com