Re: [MiNT] AES editable field extension proposal

[Arnaud BERCEGEAY <arnaud.bercegeay@free.fr>]

Hello,

It seems that i wasn't clear.

IMO, the password stuff is something AES related, and the AES should 
provide a way to type password in editable field. It's the AES aim to 
provide such feature.

About windom: yes this feature will be included in windom. If the 
application (compiled against windom) has the password feature, then 
windom will let the AES manage this. In other case (old AES), the editable 
field will be replaced by a userdef etc... by windom.

Another point: from the application side, it's not "just" a userdef... 
it's much more ! First, the application has to replace the FTEXT object by 
a userdef object, and this userdef object shall be an editable object, but 
the AES is not able to manage editable userdefined object, so the 
application has to have its own objc_edit() function too... and as 
consequence, the form_do() function from the AES could not be used: the 
application has to manage by hand the whole formular. It's far from "just 
a userdef".

Implementation of this feature in the AES is the good way because (IMO) 
modern AES should has natively that kind of feature, and furthermore, i 
don't think it's so hard to implement this is the draw subroutine of FTEXT 
objects.

I think FTEXT drawing function doesn't use pe_tvalid (it should only be 
usefull for validating characters in objc_edit), so maybe a flag or an 
extended ob_type is a better design ? To simplify, we can only check the 
1st character of pe_tvalid (other characters are not relevant if the 1st 
character is "*").

My aim is to find a "not so bad" design for this feature. Then windom will 
adopt this design, and i hope some AES will do too... and because it's not 
a windom-only feature, i'd like to hear opinions from AES authors.


Best regards,
Arnaud.