On Monday, August 11, 2003, at 08:58 AM, Daniel Vainsencher wrote:

> Since highlighting is done by the list morph anyway, that can be fixed
> by changing PLM>>drawOn:. But the ballon help is handle by the item
> morph, so its harder to cheat there. I guess we're running into a more
> general issue - who determines the items effective width, the list or
> the item. This will be more complicated when we have general morphs in
> there. Either way, if we want the StringMorph case to stay width: 9999,
> thats trivial too.

Just to throw in my 2 cents again, I think we really should get rid of 
the width: 9999 if at all reasonably possible.  It sounds like Eddie's 
changeset maintains the highlight-the-whole-line behavior, which I 
agree would be good to keep.

For the balloon help, IMHO I think it would be reasonable if you had to 
hover over the string itself, if it proves to messy to maintain the 
current behavior of hovering anywhere on the line.  (Actually I'm 
having trouble finding a place in the image where balloon help is used 
on PLM strings...)

- Doug


> Daniel
>
> Bob Arning <arning at charm.net> wrote:
>> On Mon, 11 Aug 2003 14:51:17 +0300 Daniel Vainsencher 
>> <danielv at netvision.net.il> wrote:
>>> [how wide the StringMorphs should be made]
>>> This part of the thread is moot. As Eddie said, removing the
>>> #width: call completely is the correct solution. It is no
>>> longer needed, and selection just works.
>>
>> It may be a matter of personal taste, but limiting the highlighting 
>> width to that of the characters in the string doesn't seem like an 
>> improvement. The same shortcoming applies to balloon help on list 
>> items.
>>
>> Cheers,
>> Bob