>I have another proposal to add to the list. I have tried to cover it off in another wish #1094 but it has garnered 3 votes with the average of 1.67. I think there's one or two people out there who feel obligated to rate all of my wishes at 1 < s >.

Some months ago I too have complained about this. It was my impression that someone is screwing up the ratings, and there's no way to see who it is, eventually.

>A good remedy for the potential Mdot problem is to never have a table selected, so if we could assure that state throughout our code then it wouldn't be the problem it is today.

Your idea insprired me to do a test. I wondered what the speed penalty is for having an object that on init saves the workarea and selects an empty area, and that restores the saved area when the object is destroyed. My test reveals that there is a speed penalty indeed, but that it is only worth mentioning for routines that are occasionally called many many times, for example in a sql-select. Another consequence is that its subroutines too will have an empty area. Whether this is an advanatage or disadvantage depends on the situation, I guess.

use <some table>

t = seconds()
for n = 1 to 1000
	do doit
next
? seconds() - t
return

function doit
	local loSelect0
*	?? ">" + alias() + "<"
	loSelect0 = createobject( 'select0' )
*	?? ">" + alias() + "< "
	return


define class select0 as custom
	oldSele = 0
	function init
		this.oldSele = select()
		select 0
	endfunc
	function destroy
		select ( this.oldSele )
	endfunc
enddefine