I was checking the LSS format and i noticed something possibly buggy in the SGQA identifier format: the Array(text) and Array(number) SGQA identifier.

Let's suppose we have ...
... a survey id: 95463
... a group id: 1
... a question id: 1 being Array (Text) type

... a h-subquestion id: S_1_S with description "H1"
... a h-subquestion id: S_1 with description "H2"
... a v-subquestion id: S_2 with description "V1"
... a v-subquestion id: 2 with description "V2"

yes! i now it's a bit freaky to think about it BUT try making another question (say ... question id 2) and use the first question cells as conditions. you will see a funny thing about testing/comparing the value in cells [H2][V1] and [H1][V2]: they will have the same field name: 95463X1X1S_1_S_2

in fact, it's bc the underscore is not just a separator: is also a valid code character.

my 1st suggestion: restrict user from having "_" in the question/answer/whatever code.

another funny bug would occur if an answer code or subquestion code is FOO in question 21, and an answer code or subquestion code is 1FOO in question 2, both being, say, multiple choice or radio list.

my 2nd suggestion: another underscore between question id and subquestion id or answer id: 95463X1X21_FOO and 95463X1X2_1FOO.



well ... i don't know ... tell me what do you think.

^^.

Yes, theoretically this is possible but it is checked on survey activation and semi-automatically fixed if necessary.

yes, but condition placement is still buggy about that, bc even when fields are fixed later, they are not fixed when the condition is created, and so, if you put a condition on S_1_S:2 you will see both texts, and if you put a condition on S_1:S_2 you will see both texts, and then both conditions will refer to the same item (IMHO the check/fix should occur when question structure is set/modified).

If you can create such a case which leads to problems and you can reproduce the problem, please open a ticket at our bugtracker (please attacha sample survey and step by step instructions) and we'll see what we can do.