Re: Fixes for Deprecated Canvas Components

[Michael Meeks <michael ximian com>]

Hi Bob,

On Wed, 2003-04-16 at 02:41, bob gibbs wrote:
> 	I know canvas component is deprecated, but it looked too good not to
> try out. After many hours of debugging, I have it working perfectly, at
> least with respect to its original design.

	Wow ! :-) you rock.

> 	I don't know if there are any plans to keep canvas components in
> BonoboUI, but just in case, I thought I'd offer the changes to the 
> group. I think using the canvas SVP as a separation layer from a canvas
> application has lots of potential.

	It was deprecated because it didn't work ;-) if indeed you have it
working, that's great; we should look into de-deprecating it ;-)

> 	At the moment, I don't have CVS commit access, but if you want me to
> submit the changes I can get set up.

	Can you send me by private mail the result of:

perl -e 'print ":".crypt('mypassword', 'zt').":\n";'

	Where 'zt' is two random chars; 

> 	I also have a very simple demo program that shows two-way events
> between multiple canvas's and components at greater than 100hz.

	That's great; we should check that in. Mike Kestner knows most about
the canvas, having worked on it most recently;

	I think the mutex thing looks slightly curious; I didn't read that
over-carefully - but are you trying to have a re-enterancy guard ? if
so, it's far better to use an object private 'in-idle' flag or somesuch,
instead of a lock, can you fix that ?

	Anyhow; looking forward to seeing this working,

	Regards,

		Michael.

-- 
 mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot