escapeKeyEditAction="cancel" is not discarding edit values in our sample

[senordhuff]

Hi, I am attaching a sample where if you make an edit and press "Esc", the editors are hidden but the values are not discarded. From the docs, it looks like the values should be discarded when escapeKeyEditAction is "cancel": https://www.smartclient.com/smartcli...eKeyEditAction. However, I believe there is a call to ListGrid._clearEditValues() with a param of dontRefresh==true that is causing the issue. I wasn't able to recreate this in your online demos. But, I was in the attached sample. Click on an editable field like status_2, make a change and then press Esc and the editors are closed but the edited value is not discarded.

[Isomorphic]

That was quite the test case!

If possible please do try to distill down to something smaller in the future - we ultimately pared this down to the following code, to rule out any oddities within the sample code:

ListGrid.create({
    ID:"animalsGrid",
  groupByField:"status",
  width:500, height:500,
  autoSaveEdits:false,

    editEvent:"doubleClick",
    groupStartOpen:"all",
  dataSource:animals,
  autoFetchData:true,

  fields:[
            {name:"scientificName",type:"text",width:100},
            {name:"commonName",type:"text"},
            {name:"lifeSpan",width:100,canEdit:true},
            {name:"status",type:"text",width:100,canEdit:true}
    ]
})

Anyway - we've now fixed this issue. The change will be present in the next nightly build (Oct 31 and above) going forward [12.0 and 12.1 branches]

Regards
Isomorphic Software

[senordhuff]

Thank you. Looks good. Will try to make the samples smaller but that can be very time consuming and then we will often spend a lot of time laboriously adding things back to recreate the issue at hand. I understand that it is more time consuming for you to wade through a large sample though.

[Isomorphic]

With regard to simplified test cases - it makes a big difference for us to have a standalone runnable test that is as simple as possible, so we really do appreciate the effort.

In addition to being more time consuming to debug something large and complex (or pare it down to a simple test case ourselves), there's a very real concern that we'll get a large complicated test case and assign it to a developer to investigate, only to ultimately discover that we're looking at an application level coding error rather than a framework problem we can address from our end.

Anyway - in this case we're glad that we got it figured out and that the fix resolved the issue for you

Regards
Isomorphic Software