From: Jeff H. <jeffh@ActiveState.com> - 2005-09-28 23:53:00
|
> > New patch 1302146 posted to SourceForge: > > This changes how the [ttk::entry] widget uses the 'invalid' state bit > > as follows: [...] > > Please try it out and let me know. > > OK, so no feedback yet. I'm going to go ahead and commit the > patch; it might not be right, but it's at least closer. The patch changes appeared to satisfy one of the key problems, but I didn't get a chance to apply the patch and test. I will certainly test after you apply. > On a related note: what should [$e validate] return? Whatever the result of the -validatecommand was, 1 if none? > The core manpage entry(n) just says: > > | pathName validate > | This command is used to force an evaluation of the validateCommand > | independent of the conditions specified by the validate > | option. This is done by temporarily setting the validate option > | to all. It returns 0 or 1. > | ^^^^^^^^^^^^^^^^^ > > Based on the manpage, the [ttk::entry] validate method is > compatible -- it does, in fact, always return 0 or 1 -- but > the conditions under which it does so might be different -- > see #1236979. Hmmm, the extra cases for returning 0 seem odd, but otherwise it looks OK. > (Which brings up another minor incompatibility: -- The core > entry widget performs validation "by temporarily setting the > -validate option to 'all'". The ttk::entry widget performs > validation by performing validation. I don't think this is > too serious a problem ....) I think that is not an issue. Thanks, Jeff |