>>Is it the fact that it fires for objects on the first page at initialization?
>
>2. This does not respect the documentation
>3. Read point 1.
>4. But it not fires on SCX.

I believe what the help intends, which does not excuse it being unclear, is that UIEnable does not fire for all controls on all pages at initialization. Since the first page will be activated at initialization, it makes sense that the UIEnable will fire for these controls during initialization - although, to act as documented it should fire after first page Activate() instead of during init process. Perhaps there is some internal reason it needs to work this way.
The "good" news here, though it requires some workaround, is that the UIEnable does not fire twice at initialization. i.e., it fires during Init process, but not at the first page Activate immediately following Inits. e.g.

** do form testform (scx)
TESTFORM.PAGEFRAME1.PAGE1.TEXT1.UIENABLE .T.
FORM1.PAGEFRAME1.PAGE1.TEXT1.INIT
FORM1.PAGEFRAME1.PAGE1.INIT
FORM1.PAGEFRAME1.INIT
FORM1.INIT
FORM1.PAGEFRAME1.PAGE1.ACTIVATE
FORM1.ACTIVATE

It is odd that the event sequence is not the same between VCX and SCX.

** ox=newobject("testform","testlib")
TESTFORM.PAGEFRAME1.PAGE1.TEXT1.INIT
TESTFORM.PAGEFRAME1.PAGE1.INIT
TESTFORM.PAGEFRAME1.PAGE1.TEXT1.UIENABLE .T.
TESTFORM.PAGEFRAME1.INIT
TESTFORM.INIT
** ox.Show()
TESTFORM.PAGEFRAME1.PAGE1.ACTIVATE
TESTFORM.ACTIVATE

However, with the proper workaround (perhaps Cetin will share <s>), either sequence shouldn't matter in this case.