TMSWhite wrote: Yep, I made another error too ... look for more

flocs, how would you use a "this" selector for tailoring?

For use within templates, the easiest solution would be to create new substitution variables like QUESTION_XXXX, where XXXX are each of the new EM-related attributes.

Or, is the thought that you would have a lot of similar questions and using "this" would avoid the need to modify the question text?

TMSWhite wrote:
I think it is useful especially when using javascript workaround to extend capability of questions. I know that EM will make a lot of javascript workaround useless, but in some case a kind of this selector would be convenient. I'm thinking about some questions I've created in a survey (e.g. Track respondents answers across multiple surveys or an image selection question) that could be reused without any modifications if this exists

flocs-

Having a "this" variable for simple questions would probably be do-able. However, how would you want to refer to (e.g. create names for) questions that have sub-questions? Or, would you want "this" to refer to each sub-question as it is processed?

If you can provide some sample code that would use "this", I can think through this in more detail.

/Tom

Hello,

I tried to test the EM and condition to relevance conversion and it seems to work fine with our survey (which seems to be slow in 1.91+ because of its 11 conditions in one matrix question). The relevance now looks likeand should be the same as the conditions we set in 1.91+.
Though it would be interessting to see if it's really much faster. But I can't test that at the moment.

Hello,

I tested a survey using "&newtest=Y". I stoped at the second page, then started the survey. I then used the link without "&newtest=Y" but was sent to the second page and got the message:
"Please use the LimeSurvey navigation buttons or index. It appears you attempted to use the browser back button to re-submit a page."
After a click on okay I answerd the question and clicked on the "next" button. I then came out the welcome page of my survey.
It seems there is a problem to differentiate between normal survey participation and test mode.

[quote="vierundzwanzig" post=71668
It seems there is a problem to differentiate between normal survey participation and test mode.
[/quote]
This sounds to be worth a bug ticket

Mazi wrote: This sounds to be worth a bug ticket [/quote]

reported:
bugs.limesurvey.org/view.php?id=5656

Hello,

With 1.91+, we have the opportunity to use focusfist function, but now this is in LEMsetTabIndexes function, we have $(this).focus(); .

I think it's not a good idea to have an automatic focusfirst function, can we remove it ?

Denis

PS: there are a problem for ranking question and input Index, but in 1.91+ too.

vierundzwanzig and Mazi-

The "newtest" issue is an artifact of how I'm handling sessions to keep a person from submitting the same page twice. I create a random number and send it in the HTML, plus store it in the $_SESSION. If you submit a page that doesn't have the same number, then you might have had two windows open, or clicked the browser back button, or other reasons.

This is not the most elegant solution. Yii will have better session handling which will obviate the problem.

Sounds like I just need to detect when "newtest" is passed on the command line and have the functions that look for the random number ignore it.

/Tom