RE: Bug in IOTP's DTD

["Smith, Chris " <CHRIS.SMITH@royalbank.com>]

Title: RE: Bug in IOTP's DTD

Interesting. The following is a rather subtle point, easily missed
when viewing the DTD (and NOT actually an error), but...

IOTP PackagedContent has

 <!ELEMENT PackagedContent (#PCDATA) >
 <!ATTLIST PackagedContent
  Name        CDATA            #IMPLIED
  Content     NMTOKEN          "PCDATA"
  Transform   (NONE|BASE64)    "NONE" >

while IOTP,DSIG has (corrected with removing the #IMPLIED)

 <!ELEMENT Value (#PCDATA) >
 <!ATTLIST Value
  ID          ID               #IMPLIED
  encoding    (base64|none)    'base64' >

Is there a specific reason that the attributes
"encoding" and "Transform" don't agree on parameter
content and the default value of the parameter,
even though they are almost exactly the same
thing? Also, I can live with 'encoding' but this
is risking confusion with 'encoding' as it refers
to character sets like ISO-8859-1 (and as used on
the <?xml ...> XML declaration).

I realize that IOTP and DSIG are being worked on
independently, but could we have them be a little
less independent?

------------------------------------------------------------
 Chris Smith                                +1.416.348.6090
 Royal Bank                       chris.smith@royalbank.com

> -----Original Message-----
> From: David Burdett [mailto:david.burdett@commerceone.com]
> Sent: October 14, 1999 08:55 PM
> To: 'hiroya@sdl.hitachi.co.jp'
> Cc: IETF Trade (E-mail); 'kawatura@bisd.hitachi.co.jp';
> 'kent@differential.com'
> Subject: RE: Bug in IOTP's DTD
>
>
> Masaaki
>
> I accept that there is a mistake in the IOTP DTD, however
> this part of the IOTP DTD was copied **directly** from the
> the IOTP dsig spec at
>
>      
> http://www.ietf.org/internet-drafts/draft-ietf-trade-iotp-v1.0
> -dsig-03.txt
>
> which contains the same error. The only changes made when
> copying were the removal of some blank lines and page headers
> and footers.
>
> The reason for copying in the IOTP dsig DTD is to avoid
> having to use namespaces in IOTP.
>
> So really what I think we should do is:
> 1. Fix the DTD in the IOTP dsig spec (Yoshiaki ?) and then
> 2. Fix the DTD in the IOTP spec.
>
> Regards
>
> David
>
> -----Original Message-----
> From: Masaaki Hiroya [mailto:hiroya@sdl.hitachi.co.jp]
> Sent: Thursday, October 14, 1999 3:33 AM
> To: David Burdett
> Cc: hiroya@sdl.hitachi.co.jp
> Subject: Bug in IOTP's DTD
>
>
>
> David
>
> There is a bug in ietf.org/draft-ietf-trade-iotp-v1.0-protocol-06.
>
> >  <!ELEMENT Value ( #PCDATA ) >
> >  <!ATTLIST Value
> >          ID               ID           #IMPLIED
> >          encoding    (base64|none)    #IMPLIED      'base64'
>
>
> I think #IMPLIED in the "encoding" attribute should be removed.
>
> Draft-ietf-trade-iotp-v1.0-dsig-03.txt defines the Value
> element as follows:
>
> >    <!ELEMENT Value ( #PCDATA ) >
> >    <!ATTLIST Value
> >            ID                 ID            #IMPLIED
> >            encoding      (base64|none)     'base64'
>
> Please correct the DTD in iotp-v1.0-protocol-06.
>
> Masaaki
>
> -----
> Masaaki Hiroya
> Systems Development Laboratory
> Hitachi, Ltd.
> tel: +81-44-549-1531
> fax: +81-44-549-1640
> email: hiroya@sdl.hitachi.co.jp
>