tcllib-devel

[Tcllib-devel] BWidget and Tile

[Damon C. <da...@tc...>]

    Should BWidget do what Tile used to do and just ignore the options
passed that are not Ttk options?  Tile has now removed all of these
"dummy" options, but I'm wondering if BWidget shouldn't do this to
avoid upgrade problems?

    Any thoughts?  I'm looking forward to the future of BWidgets being
object-oriented with the new Tcl OO system, but for now, I need to get
it working with Tile and playing nice.  I just need some direction
from someone other than me.  I don't like developing in a vacuum.

D

RE: [Tcllib-devel] BWidget and Tile

[Jeff H. <je...@Ac...>]

Damon Courtney wrote:
>     Should BWidget do what Tile used to do and just ignore 
> the options passed that are not Ttk options?  Tile has now 
> removed all of these "dummy" options, but I'm wondering if 
> BWidget shouldn't do this to avoid upgrade problems?
> 
>     Any thoughts?  I'm looking forward to the future of 
> BWidgets being object-oriented with the new Tcl OO system, 
> but for now, I need to get it working with Tile and playing 
> nice.  I just need some direction from someone other than me. 
>  I don't like developing in a vacuum.

IMO BWidgets won't "survive" the OO transition without a
major rewrite.  I for one am duplicating BWidgets into my
own tklib widget package, which is snidget based (and thus
closer to what the OO would be).  These have different
options, and perhaps different behaviors, and more
importantly assume you already are using all that Tk 8.4
and Ttk has to offer.

Jeff

RE: [Tcllib-devel] BWidget and Tile

[Damon C. <da...@tc...>]

> IMO BWidgets won't "survive" the OO transition without a
> major rewrite.  I for one am duplicating BWidgets into my
> own tklib widget package, which is snidget based (and thus
> closer to what the OO would be).  These have different
> options, and perhaps different behaviors, and more
> importantly assume you already are using all that Tk 8.4
> and Ttk has to offer.

    I'd be more than happy to port BWidgets to SNIT if I had some kind of
assurance that SNIT would be the megawidget package of choice for the
core.  We have too many of these damn things.  SNIT is great, but it's
a lot of work to port BWidgets over only to find out that when OO gets
in the core, a whole new megawidget syntax will be built on top of
that.

    I assume SNIT will be rewritten in the core OO once 8.5 is released. 
Can you say, with any certainty, that SNIT's syntax will be the
megawidget syntax of choice?  If so, I'll gladly put in the work to
port BWidgets to SNIT with the idea that it will be REALLY fast in the
8.5 release.

D