Another LC9.6.9rc2 regression? This time with Datagrids

[Paul Dupuis]

Hi Mike,

I suspect some change was made in the Datagrid scripts, such as a move 
of some code from a stack to a script-only stack or some engine change 
in how comparisons are done, variables interpreted, etc. that 
inadvertently impacted the Datagrid code, but while we use a lot of 
Datagrids, I have never really tried to follow or understand the 
internal scripts and I'd can't invest the time just now (too much else 
taking priority). I have I think, clearly demonstrated this is a 
regression - code works under LC 9.6.8 and fails under LC 9.6.9rc2, so 
it's now up to Livecode, Ltd. (or yourself IF you want to dive right in 
;-) to solve the bug.

Our choice boils down to [A] Build under 9.6.8 and not have a fix for 
the Non-Gregorian calendar crash under macOS which was fixed in 969rc2 
OR [B] wait until LC staff or someone fixes this and we find no other 
regression in 9.6.9. I think we've now found 3 regressions in 3 weeks, 
which is not a good sign for 9.6.9.

BTW, thank you again for your help with the Ask and Answer regressions!

-- Paul


On 12/13/2022 4:31 PM, Mike Kerner via use-livecode wrote:
> is this in a built app or interpreted?
> the "silent fail" has been a feature of LC apps on mobile for a decade.
> i wonder if these issues are both caused by some stack moving from binary
> to SOS. the ask dialog certainly was migrated to an SOS, recently.
> i see you included a recipe stack in the bug report. just like with the ask
> dialog, the temptation to dive in might be too great for me...
>
> On Tue, Dec 13, 2022 at 4:20 PM Paul Dupuis via use-livecode <
> use-livecode at lists.runrev.com> wrote:
>
>> Anyone else having problems with:
>>
>> set the dgProp["someProperty"] of grp "someDatagrid"
>>
>> under LC 9.6.9rc2 vs LC 9.6.8?
>>
>> In bug https://quality.livecode.com/show_bug.cgi?id=24046 were seeing
>> lines of script where we set the dgProp of a datagrid just stop (like it
>> hit and 'exit to top'). The same code works without stopping in LC 9.6.8
>> and consistently fails in 9.6.9rc2.
>>
>> Just wondering if anyone else has run into this. I did not see a
>> specific bug reported against 9.6.9rc2 that might have been this issue
>> before reporting it.
>>
>>
>> _______________________________________________
>> use-livecode mailing list
>> use-livecode at lists.runrev.com
>> Please visit this url to subscribe, unsubscribe and manage your
>> subscription preferences:
>> http://lists.runrev.com/mailman/listinfo/use-livecode
>>
>