The quick question is it unadvisable to import LSS files with quite a bit of relevance/validation code from previous versions of LS (as far back as 2.06)?
A longer explanation is that I've imported LSS files created in previous versions ( 2.06, 2.6) to the latest 3.x build. The surveys have a lot of relevance equation code. The import goes through fine, but when I click into the survey and click to various questions that have the code, it displays some Qcode variables as "undefined." If I click to view the survey logic file it clears up most of the issues. For example, something simple like (currentintervention == 1) where "currentintervention" is undefined in corrects itself after the survey logic file runs. But something more complex like ((intNoofSibs.NAOK > "0") and (sibCondtions_Sib1.NAOK == "Y")) won't clear up until I view the question and resave it without make any other changes.
This seems like a bug that I will report once I get confirmation from the sages in the forums.
I'm using the latest version 3.0.5+180118, however my problem has persisted since 3.x.
jboogie21 wrote: The quick question is it unadvisable to import LSS files with quite a bit of relevance/validation code from previous versions of LS (as far back as 2.06)?
Older LSS should be importable in new versions of LimeSurvey.
The 3.X seems to be still not ready to import all kind of older LSS. LSS exports from the current LTS LimeSurvey version is important. There were already some bugreports around importing LSS. So feel free to open a bug ticket.
One of the biggest change between 2.x and 3.x is the replacement engine.
It seems that we've lost some keywords in the operation, and we're putting them back one by one.
