OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   Please keep Relax NG simple (RE: RELAX NG and type derivation)

[ Lists Home | Date Index | Thread Index ]

On Thu, 2002-07-04 at 20:18, Michael Fitzgerald wrote:
> RELAX NG can do type derivation by restriction, union, and list. In some
> ways, RELAX NG's list type derivation is more powerful that XML Schema's
> derivation.[1] All the facets that are available in XML Schema are available
> in implementations of RELAX NG. In what specific ways does RELAX NG fall
> short in type derivation?

Although I agree that, technically, you're 100% right, I find very
frightening any attempt to align, feature by feature, Relax NG on W3C
XML Schema.

I have a lot of respect for the members of the W3C XML Schema Working
Group and if WXS is so complex I don't think we can blame them but
should rather blame the wideness of the scope of the problem they've
tried to solve.

I have also a lot of respect for the Oasis Relax NG Technical committee,
but I think that the main reason why Relax NG is so much simpler is not
because they are so much more clever but because they've been wise
enough to reduce the scope of the problem they've tried to solve.

I have been very impressed in meetings I have attended to see how James
Clark splits any problem which is getting complex into smaller problems
which can be solved independently and I think that one of the main
reasons of the complexity of recent W3C specifications is that this
policy is not followed as far as it could.

If Relax NG can't make coffee that's fine with me and it's not because
specification Y or Z does it that it's a good reason to change Relax NG
to do it :-) ... The right thing to do IMO is you want to use Relax NG
to do coffee is to build a new different application (eventually using
Relax NG) to do so.

When I joined Sybase in 93, Sybase has a SQL engine which was really
different (I am not saying better but different) from other engines. It
was designed for transactional and "purely" SQL applications and only
the features needed for these applications had been implemented. As a
result, SQL Server 4.9 was much simpler to install than Oracle 7, it
required much less memory and processing power and, for these
applications, it was much faster.

The downside was that to run smoothly with SQL Server the applications
had to be written for SQL Server (there was no row level locking for
instance), that applications ported from mainframes were often a
disaster and that performances for decisional applications were poor.

When the marketing guys took control of the product and asked to the
engineering to implement the features which were "missing" guess what
happened? SQL Server got more alike Oracle, sure but their performances
too and it almost killed the company.

In other words these two languages need to be kept different and if you
ever added all the features of WXS into RNG, RNG would become WXS++ and
would not be RNG any longer :-) ...

Eric
-- 
See you in San Diego.
                               http://conferences.oreillynet.com/os2002/
------------------------------------------------------------------------
Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
(W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
------------------------------------------------------------------------





 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS