Re: [TCLCORE] (No subject)

[Christian W. <Chr...@t-...>]

On 08/09/2025 06:58 PM, Brian O’Hagan via Tcl-Core wrote:

Brian,

> I’ll fix it in my fork of TkTable.

Great. If you are at it, maybe my observations help for further review.
The problem became slowly apparent using tksqlite on some complex query,
which gives one result column and possible zero rows. It took me quite
a time to come to the conclusion, that the visual representation in this
case was misleading. The first culprit was tksqlite, but the real problem
turned out to be tktable in this case. Now the bad point is, that the
many combinations of number of header rows and/or columns plus the number
of modes on how to deal with the selection might yield, that my naive
patch isn't even a tenth of the real solution. It only fixed my issue
in the aforementioned tksqlite case.

Hope that helps for improving tktable,
Christian