Display existing geometry when capturing new point/trace/shape

[lognaturel]

Brainstorming doc

[seadowg]

@lognaturel I'm a little confused as to where we are with this? Are we happy to move ahead with the spec detailed in your doc?

[lognaturel]

The big gap is that we don't have a XLSForm spec yet. We need to decide whether we always have form designers write raw xpath expressions there or whether we try to do some kind of pulldata-style simplification for simple filter expressions. I don't think that's likely to change the XForm spec but there's always some risk.

You and @grzesiek2010 said you're both happy with the latest version that uses an itemset, right? Would you be ready to start development right away? I can prioritize finalizing the XLSForm spec if so.

[seadowg]

You and @grzesiek2010 said you're both happy with the latest version that uses an itemset, right?

Yeah, it aligns nicely with the recent changes being made to range as well (as we intended). @grzesiek2010 you have anything you're concerned about with that approach after actually doing some implementation work with it for range?

Would you be ready to start development right away? I can prioritize finalizing the XLSForm spec if so.

Definitely not a rush as we've got other things to work on, it more looked to me like the XForm spec was ready to go, and I couldn't find any reference to an XLSForm spec. Once, that's done I can update the issue and unblock!

[grzesiek2010]

@grzesiek2010 you have anything you're concerned about with that approach after actually doing some implementation work with it for range?

No, as I mentioned in the doc, it’s actually a good thing that we’ll have a second question type with optional itemsets, as it will give us an opportunity to take a broader look at the changes we should introduce in JR.