rule based engine?
[1/7] from: maarten:koopmans:surfnet:nl at: 24-Dec-2002 12:57
Hi,
I remember discussing this on the list before with Robbert, but has
anyone ever looked into a rule-based engine for Rebol?
--Maarten
[2/7] from: robert:muench:robertmuench at: 24-Dec-2002 14:17
> -----Original Message-----
> From: [rebol-bounce--rebol--com] [mailto:[rebol-bounce--rebol--com]]
<<quoted lines omitted: 4>>
> I remember discussing this on the list before with Robbert, but has
> anyone ever looked into a rule-based engine for Rebol?
Hi, yep it's still on my agenda and the topic is moving upwards the
priority list. I want to add such an engine to my RFM. I remember
someone (Gregg?) having done some state-machine stuff.
What I want to start with is the Rete algorithm. It uses some graph
based datastructure. The graph stuff is already done for RFM.
What kind of rules are you thinking about? Robert
[3/7] from: greggirwin:mindspring at: 24-Dec-2002 10:44
Hi Maarten,
MK> I remember discussing this on the list before with Robbert, but has
MK> anyone ever looked into a rule-based engine for Rebol?
I did a very simple MAKE dialect, which isn't much, then I started
tinkering around with some more advanced things and decided I really
needed to understand the fundamentals better, since it's not an area
I've done much in. To that end, I started porting some stuff out of
the Winston/Horn LISP book (basically writing a Prolog interpreter).
I got it mostly done before getting distracted, but it's not 100%. I can
send over what I have if you think it would help but it was really
more a learning excercise than anything; with the goal of building a
clean slate REBOLish rule system eventually.
-- Gregg
[4/7] from: maarten:koopmans:surfnet:nl at: 24-Dec-2002 22:48
>>I remember discussing this on the list before with Robbert, but has
>>anyone ever looked into a rule-based engine for Rebol?
<<quoted lines omitted: 6>>
>based datastructure. The graph stuff is already done for RFM.
>What kind of rules are you thinking about? Robert
I want to look into a policy engine (authorization) which is
goal-directed, hence backtracking. Looks like we're going differnt ways :(
The hard thing seems to do it the Rebol way, utilizing all dynamic code
features with a minimal evaluator.
--Maarten
[5/7] from: robert:muench:robertmuench at: 25-Dec-2002 8:33
> -----Original Message-----
> From: [rebol-bounce--rebol--com] [mailto:[rebol-bounce--rebol--com]]
<<quoted lines omitted: 5>>
> goal-directed, hence backtracking. Looks like we're going
> differnt ways :(
Hi, backtracking is nice but a bit harder to do ;-) that's why I first
want to start with forward chaining. But I'm interested in the other
part as well.
But are you sure that you really need backtracking? For forward chaining
it might even be possible to use PARSE.
You might have a look at 11. at
http://www.dcs.kcl.ac.uk/staff/pjm/publications/
> The hard thing seems to do it the Rebol way,
> utilizing all dynamic code
> features with a minimal evaluator.
Did you already started with something? Robert
[6/7] from: robert:muench:robertmuench at: 25-Dec-2002 8:33
> -----Original Message-----
> From: [rebol-bounce--rebol--com] [mailto:[rebol-bounce--rebol--com]]
<<quoted lines omitted: 6>>
> was really more a learning excercise than anything; with the
> goal of building a clean slate REBOLish rule system eventually.
Hi Gregg, I would be interested in having a look at what you have done
:-)). So, if you don't mind please send me a copy. Thanks. Robert
[7/7] from: greggirwin::mindspring::com at: 26-Dec-2002 10:56
Hi Robert,
RMM> Hi Gregg, I would be interested in having a look at what you have done
RMM> :-)). So, if you don't mind please send me a copy. Thanks. Robert
Will do.
-- Gregg
Notes
- Quoted lines have been omitted from some messages.
View the message alone to see the lines that have been omitted