Re: FpML-VAL Validation Rules Syntax

[Daniel Dui <daniel.dui@xxxxxxxxx>]

Lyteck

Perhaps we should start with a bottom-up approach. Do we have a list
of rules that expose these problems?

I agree with Christian, I don't think that a validation rule should
require access to reference data. I think that the only "external"
data on which a rule may depend should be the standard calendar. A
rule should also not depend on a particular business calendar.

I don't think that if a rule is enforced in practice by the internal
logic of a producer or consumer application, it should be considered
superfluous.

For example fx-3, which states that in an FX product rate = spotRate +
forwardPoints. This rule could be enforced well before the point when
an XML message is constructed. It's nontheless useful to document it
as a rule.

The rules and the schema are there to define what a valid FpML message
looks like. An implementer may be able to produce valid messages
without implementing every single rule.

-daniel




On 4 February 2010 22:04, Lyteck Lynhiavu <LLynhiavu@xxxxxxxx> wrote:
> All:
>
>
>
> The group needs to address Harry McAllister’s concern regarding the
> expression of validation rules. We concluded at the last meeting that many
> of the existing issues and future work are dependent on finding an
> agreement.
>
>
>
> Perhaps we can start this important discussion via this mailing list. The
> goal is to gather as much feedback as possible and opinions from everyone.
> This will be helpful for when have the discussion at next meeting (TBD –
> within 1 or 2 weeks)
>
>
>
> There are a couple of reference notes that can give you some background:
>
>
>
> 1.  Excerpt of the minutes from the December 2009 SC meeting:
>
>
>
> “Harry McAllister clarified that the issue for discussion doesn’t relate so
> much on the syntax of validation rules per se but rather to the scope of
> validation rules. The validation WG has been attempting to express some
> quite complex algorithms and relationships and is running into difficulties
> because of the limitations of the syntax that has been adopted as a primary
> expression of these rules. The point is not so much about the adoption of a
> particular syntax but rather to the fact that the validation WG is in some
> cases attempting to implement rules which would already be the problems for
> any consuming system.
>
>
>
> Harry would like the group to make a decision as to whether such algorithms
> should even be in the scope of the Validation WG or whether we should
> restrict ourselves to simpler expressions.
>
>
>
> Andrew noted that we need to ensure the samples we provide are correct and
> complete.
>
>
>
> We should look into the possibility of using:
>
> - simpler expressions readily expressable in the XPath/XQuery syntax adopted
> by the group
>
> - more complex expression which may have a natural language expression but
> for which we don’t necessarily mandate a procedural expression.”
>
>
>
> 2.  Excerpt of the minutes from the last validation WG meeting (1/19):
>
>
>
> “Harry McAllister expressed some concerns (at the December 2009 Standards
> Committee meeting) regarding the technical expression of validation rules,
> in particular limitations of the chosen XPath technology.
>
> -    Background: Over the past 1½ year, the validation WG reengineered the
> way validation rules are expressed in the specification and agreed to
> provide a natural English expression and a technical (XPATH) expression for
> all validation rules. The problem is that some complex rules simply can’t be
> expressed using any particular syntax and that perhaps those rules should
> only have a natural/English expression and forgo the technical expression.
>
> -    àHarry McAllister was to send an open note to the group for discussion.
>
> -    The group discussed the issue but felt more input from Harry was needed
>
> o    Marc Addison noted XPATH was chosen as the natural implementation
> technology for XML.
>
> o    Andrew Jacobs noted some rules can’t be easily or accurately expressed;
> and implementation technology is not powerful enough and too complicated to
> implement.
>
> o    Andrew noted some open issues could be closed if the WG decided not to
> provide a technical/XPATH expression for those rules that are too complex to
> express. Irina questioned how would the group make the determination of
> whether a validation rules should have one or both representations.”
>
>
>
> What’s your take? How can we make validation rules better?
>
>
>
> Lyteck
>
>
>
>
>
> From: valwg@xxxxxxxx [mailto:valwg@xxxxxxxx] On Behalf Of Lyteck Lynhiavu
> Sent: Tuesday, January 19, 2010 5:10 PM
> To: valwg@xxxxxxxx
> Subject: FpML-VAL RE: Reboot: Meeting - Tuesday January 19th Minutes +
> **Looking for a WG Chair**
>
>
>
> Attendees:
>
> ----------
>
> - Marc Addison
>
> - Marc Gratacos, ISDA
>
> - Irina Yermakova, ISDA
>
> - Andrew Dingwall-Smith
>
> - Daniel Dui
>
> - Andrew Jacobs
>
> - Lyteck Lynhiavu, ISDA
>
> -
>
> ________________________________
> The information contained in either this email and, if applicable, the
> attachment, are confidential and are intended only for the recipient. The
> contents of either the email or the attachment may not be disclosed or used
> by anyone other than the addressee. If you are not the intended
> recipient(s), any use, disclosure, copying, or distribution is prohibited
> and may be unlawful. If you have received this communication in error,
> please notify us by e-mail at isda@xxxxxxxx <mailto:isda@xxxxxxxx> then
> delete the e-mail and all attachments and any copies thereof. This
> communication is part of an ISDA process and is not intended for
> unauthorized use or distribution.
>



-- 
____________________________________________________________
   Daniel Dui - daniel.dui@xxxxxxxxx - skype: danieldui

-------------------------------------------------------------------------------
To unsubscribe: Email majordomo@xxxxxxxx with a blank subject line
In the body include the line: unsubscribe valwg youremail@address
To view archives: http://www.fpml.org/_wgmail/_valwgmail/threads.html