RE: [RAF] small comment on discriminator "declarative vs. procedural"

Good point.
- rules can be defined and executed declaratively OR procedurally
- rule conditions and actions can be order dependent or independent (cf declarative vs procedural)

Although procedurally defined / executed rules are not very interesting to the academic community (I suspect), they are very common in areas such as BPM and workflow. However, as I did not note any use cases for (interchange of rules for) workflow it could be that procedural rules are not relevant to RIF.

Order dependency for rule conditions is probably not much of an issue other than to rule engine writers.

Order dependency for rule actions / consequents will probably depend if the rule is "logical rule" or an "executable process rule" - in the latter case it could be important (eg my actions could be order-dependent: I set the customer account to invalid and THEN send the email to the customer).

Paul Vincent
Fair Isaac Blaze Advisor --- Business Rule Management
OMG Standards for Business Rules, PRR & BPMI
mobile: +44 (0)781 493 7229 ... office: +44 (0)20 7871 7229 

-----Original Message-----
From: public-rif-wg-request@w3.org [mailto:public-rif-wg-request@w3.org] On Behalf Of Axel Polleres
Sent: Friday, February 10, 2006 2:11 PM
To: Christopher Welty
Cc: public-rif-wg@w3.org
Subject: [RAF] small comment on discriminator "declarative vs. procedural"


Dear Chris,

I had some discussion with Harold about the "declarative vs. procedural" 
discriminator mentionedin the RAF wiki page (see mails below).

Basically, I think that the discriminator "declarative vs. procedural"
does not properly reflect the text:

" what was meant here is the availability (or not) in a rule of
   procedures (expressed in some programming language) that perform
   computations (such as doing arithmetic) that are difficult or
  impossible to express in logic"
at http://www.w3.org/2005/rules/wg/wiki/Rulesystem_Arrangement_Framework

Is IMHO misleading: In LP usually one understands the different between 
procedural and  declarative in the sense that the order of rules does 
not play a role for the semantics, i.e. that a set of rules is indeed 
handles as a set with a declartive semantics. This is at least the main 
issue which gives PROLOG its procedural style vs declarative logic 
programming. It is not necessary about arithmetic or procedural 
attachments alone.

So, I'd suggest to change thw current heading for this discriminator
to something like "builtins vs. no builtins" and redefine the 
discriminator "declarative vs. procedural" as a new one.

best,
axel

-------- Original Message --------
Subject: RE: [RAF] small comment
Date: Fri, 10 Feb 2006 08:52:20 -0500
From: Boley, Harold <Harold.Boley@nrc-cnrc.gc.ca>
To: <axel@polleres.net>

Hi Axel,

> This is mostly what I meant. I will look at this in detail asap.

I now found the time to see that you commented on the RAF part recently
changed by ChrisWelty, going back to F2F1 breakout session notes by
PascalHitzler. I did not add explanations to their parts.

However, I alluded to a distinction in:
"For example, since the [WWW] Phase 1 rule semantics will be essentially
Horn Logic, the "procedural vs. declarative" Discriminator (along with
many "procedural" Subdiscriminators) of the "Suggested discriminators
..." could be added for Phase 2."

I agree with you that this "procedural vs. declarative" distinction
is not quite the same as in LP. The problem may have to do with the
distinction of "procedural attachments vs. no procedural attachments"
(which I would therefore rename into "builtins vs. no builtins"):
This latter one is the one, I guess, ChrisWelty's comment refers to.

After your comment, my suggestion would now be to clearly separate
these distinctions, at least as subdistinctions.

You can modify/forward this email to ChrisWelty.

Best,
Harold


-----Original Message-----
From: Axel Polleres [mailto:axel.polleres@uibk.ac.at]
Sent: February 9, 2006 10:14 AM
To: Boley, Harold
Subject: [RAF] small comment

Boley, Harold wrote:
>> [NEW] ACTION: Harold will explain what Lloyd Topor extensions etc
>> mean
> 
> [recorded in http://www.w3.org/2006/01/31-rif-minutes.html#action18]
> 
> I updated the 
> http://www.w3.org/2005/rules/wg/wiki/Rulesystem_Arrangement_Framework
>  section "Phase 1 Expressive Discriminators", part "Syn: Syntactic 
> Discriminators", adding brief explanation sub-bullets plus links
> (except for the self-explaining item 1.).

Hi Harold!

I have a small comment  on the RAF page:

> procedural vs. declarative: Rules are generally declarative, I assume
> what was meant here is the availability (or not) in a rule of
> procedures (expressed in some programming language) that perform
> computations (such as doing arithmetic) that are difficult or
> impossible to express in logic

In LP usually one understands the different between procedural and
declarative in ther sense that the order of rules does not play a role
for the semantics, i.e. that a set of rules is indded handles as a set
with a declartive semantics. This is at least the main issue which gives

PROLOG its procedural style vs declarative logic programming paradigms,
right? It is not necessary about arithmetic or procedural attachments
alone.

best,
axel


-- 
Dr. Axel Polleres
email: axel@polleres.net  url: http://www.polleres.net/




-- 
Dr. Axel Polleres
email: axel@polleres.net  url: http://www.polleres.net/

Received on Friday, 10 February 2006 14:26:25 UTC