Skip Logic not working after creating new version

[Omar Ahmed]

Hi,

I am using OpenXData 1.3.4 (build 149) in which skip logic keeps disappearing from the forms. I have created a new form version which has the working skip logic from the previous version but the functionality is missing when I open the properties window and Preview. I can see the logic in Xforms Source tab for the newer version. (http://pastebin.com/yJa6Gxtk). Kindly advise.

Regards,

Omar 

--

______________________________________________________________
Omar Ahmed
Engineer,
 
Interactive Research and Development (IRD)
Suite 508, Ibrahim Trade Tower
Main Shahrah-e-Faisal
T: +92(21) 34327697
W: www.irdresearch.org

[Sarah Bird]

Hi Omar,

(Brent, Omar did post this to -users list :D)

Brent is right. If you look through your xml, you will see that there after every <relevant the binding has not been updated to v2 - which must be the source of your problems.

I am totally happy to investigate if this is a consistent issue and whether or not it's been fixed on newer versions of openXdata.

Can you privately email me a copy of the original v1 - preferably the whole study.

Cheers,

Bird

On Thu, Mar 1, 2012 at 1:36 AM, Brent Atkinson <brent.a...@gmail.com> wrote:

Oh,

And this issue is more appropriately posted to the users list. Also, if this turns out to a new issue (I am not sure it is) then you should report it as an issue. If not, then it might be possible to release the fix as a 1.3 maintenance release.

Brent

On Thu, Mar 1, 2012 at 12:33 PM, Brent Atkinson <brent.a...@gmail.com> wrote:

Omar,

Check that the binding names in the contsraints are up-to-date. When you create a new version, bindings may not be updating to reflect that in the constraints. I think there was an issue reported, but I can't seem to find it in Trac at the moment. You should be able to fix the stale names and it should work AFAIK.

Brent

--
You received this message because you are subscribed to the Google Groups "openXdata Developers" group.
To post to this group, send email to openxd...@googlegroups.com.
To unsubscribe from this group, send email to openxdata-de...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/openxdata-dev?hl=en.

--
You received this message because you are subscribed to the Google Groups "openXdata Developers" group.
To post to this group, send email to openxd...@googlegroups.com.
To unsubscribe from this group, send email to openxdata-de...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/openxdata-dev?hl=en.

[Sarah Bird]

Hi all a few notes to follow-up on this.

In the end we found the problem. There was a small problem in the xml of v1, which is why v2 got screwed up and the skip logic didn't appear.

The nodeset was wrong on lines 63 & 64 - http://pastebin.com/yJa6Gxtk - this was also causing a seperate error to appear "Missing data node for :Date"

We don't know and probably can't track down how the nodeset got messed up on two questions out of 50+ it is unlikely a human error as the forms were being made by non-techies who weren't nosing around in the xml but it's going to be impossible to try and see if there's something repeatable here. so we're chalking it up to learning.

In addition, a reminder. If you do want to manually change the xml in the "xforms source" tab, to make your changes saved, you need to press the "open" button on the far left of the tool bar. This is not entirely intuitive so I have made a new ticket about it - http://trac.openxdata.org/ticket/708

Cheers,

Bird

On Thu, Mar 1, 2012 at 8:53 PM, Omar Ahmed <omar....@irdinformatics.org> wrote:

Hi Brent & Bird,

Thank you both for your prompt response. I think you are right about this. I used diff tool to find out that the "relevant" attribute of the bindings is not pointing to wrong section name which still has "v1" as opposed to "v2". The node itself points to the right section ie "v1"

relevant="/communications_evaluation_section_a_b_c_v1/C 

I am pasting the diff results so you can verify what I gathered:  http://www.diffnow.com/?report=qslho 

If this indeed is the problem, isn't it worth creating a ticket?

Regards,

Omar

(Bird: I think if correcting the version number solves it then I wont have to send copy of the study. Many thanks for your kind offer :) )