[Bug-kawa] Re: [bugs #11577] gnu.mapping.InheritingEnvironment not seeing symbols in parent environment

[thomas kirk]

 > > ok, thanks for looking into this. I can defer moving to the newer version 
 > > for now. 
 > > I was going to spend more time understanding your current implementation, 
 > > but if 
 > > you're thinking of changing the design I'll hold off. (the security 
 > > features 
 > > currently rely on fluid bindings, so I may re-think that approach in the 
 > > meantime).
 > 
 > I checked in a fix.

thanks for the fixes...this is mostly working now, modulo the followup that 
I justed submitted to savannah.

 > 
 > You can also use gnu.mapping.ThreadLocal, which can be used as an
 > anonymous fluid-variable: Just create a unique (Symbol, property)-pair.
 > I plans to use it to implement SRFI-39 "parameter objects".
 > -- 

Good suggestion, thanks. It had occurred to me to modify my current 
implementation 
using java.lang.ThreadLocal, but I wasn't aware of your gnu.mapping.Threadlocal.
Thanks for making me aware of that. 
In fact it was worthwhile that this situation forced me to rethink my design.
The fluid binding approach had some weaknesses which have been eliminated
in the new version.