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


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] terra incognita

[ Lists Home | Date Index | Thread Index ]

----- Original Message -----
From: "Paul T" <pault12@pacbell.net>
To: "Dare Obasanjo" <kpako@yahoo.com>; "Rick Jelliffe" <ricko@allette.com.au>;
Sent: Wednesday, December 19, 2001 2:32 AM
Subject: Re: [xml-dev] terra incognita

> > a.) XML is not as efficient as a serialization technology as other methods
> > especially over the network (although if use HTTP and gzip encoding this
> can
> > be improved)
> ... 'other methods' ... such as ?

The original poster mentioned CSVs which for all intents and purposes will
always be smaller files than their XML counterparts. For complex data I assume
ASN.1 could also represent the same data as XML but use less bytes to do it (I
haven't used it just know developers who have) or even better any homegrown
data format built specificially for that purpose would be more space efficient
than XML.

> > b.) Using XML involves becoming conversant with a quarter to half a dozen
> > buzzword technologies that seem to confound the average programmer and
> even
> > leave experts unsure of their worth (from SOAP to namespaces to schemas
> and
> > more)
> So XML has buzzwords. How does it prevent me from using
> XML as a 'seralization language'? I think this is irrelevant.

In general case if one uses a minimal subset of XML for serialization then
there is little complexity and XML lives up to its reputation for being
simple. However in a few places where XML is being suggested as a
serialization format there is reliance on multiple XML technologies meaning
that developers have to become conversant with namespaces, schemas,  WSDL,
SOAP, etc.  where it may have been better to use domain specific technologies
instead of interedependent XML ones since there is less chance of information
overload on the part of the developer.

> > c.) XML was originally designed to deal with text primarily and not binary
> > data, this shows itself from time to time when using XML as a data
> > serialization and is partly the cause for the intense discussion in the
> > 1.1 thread.
> So there is some thread in XML-dev mailing list. ( BTW, my oppinion is that
> 'XML is for text, so there should be no control characters involved' )
> How does it prevent me from using XML as  a 'serialization language'.
> I think this is irrelevant.

It doesn't, my point here was that in for some people (such as the IBM
mainframe folks until XML 1.1) XML was unsuited to some of their data
serialization needs due to its original intent as a text-based technology.

> I still think that serialization into brutal XML's  subset is a nice,
> practical design.

I agree, serialization using a _subset_ of XML is typically a good idea.

I will hire an expert marksman to stand by the entrance to my fortress.
His job will be to shoot anyone who rides up to challenge me.

Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com


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

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