Re: Comments on <draft-ietf-osids-distnames-02>

[Steve Hardcastle-Kille <[e-mail]>]

Sri,

 >From:  [e-mail]
 >To:    [e-mail]
 >Subject: Comments on <draft-ietf-osids-distnames-02>
 >Date:  Fri, 31 Jul 92 14:54 EDT

 >-----------------
 >
 >1 Why a notation is needed
 >  For example, the OSF Syntax may be more appropriate for some system
 >  oriented uses.
 >
 >     Please provide a short description of the OSF Syntax or a
 >     reference or both.

I do not have a citable document.   If someone can send me a detailed
reference, I will add it (anyone from OSF listening?).   

 >
 >2.1 Goals
 >
 >     The goals are quite fuzzy and need elaboration. What is meant by
 >     terms like, "intuitive", "fully general", "lay it out in a number
 >     of ways", and "clear representation"?

I believe that these are all important goals.   I feel that all
(except perhaps the last) are self explanatory, but I'm very hapy to
add some text if you feel that this will improve the document.

 >
 >     I think an important goal should be an unambiguous representation
 >     of a distinguished name.

Agreed.  I thought that this was there.  Will add. 

 >
 >2.3 Formal definition
 >
 >  There is an escape mechanism.
 >     Escape mechanism to what? Is it escaping from the quoting
 >     mechanism?

Yes.  Will add some text.

 >
 >  The keyword specification is optional.
 >     What keyword specification is optional? You do not give any
 >     references and do not define the terms before using them.

Will add a reference.

 >
 >Figure 1:
 >  Caption: BNF Grammar for Distinguished and Purported Name
 >     What is a "Purported Name"? This term was not used any other place
 >     in the document.

A hangover from the UFN spec (which uses the same BNF).  Will remove
this reference.

 >
 >     Figure 1 is the heart of the paper. Unfortunately, the BNF
 >     notation you use is not obvious to me. Can you point to a
 >     reference where it is defined or better yet, please give a short
 >     description so that the RFC will be self contained. For instance,
 >     what does 2*<hexchar> mean? What do parentheses signify. I am
 >     sorry it is not clear to me.

This has developed as a personal BNF style....   I should define it.
Basically, it is the RFC 822 BNF, with <> around the terminal symbols.


 >
 >3 Examples
 >
 >  OU=Dover Beach Consulting
 >      I think must read,
 >  O=Dover Beach Consulting
 >
 >  OU=University College London
 >      I think must read,
 >  O=University College London

Yes.  

 >
 >----------- My General Comments -------------
 >
 >   This RFC draft assumes extensive background knowledge. It will be
 >   helpful if you specify what is expected of your reader.  The paper
 >   specifies a notation for Distinguished Names. However a
 >   Distinguished Name itself does not appear any where in the paper.  I
 >   think that you should write out a distinguished name and point a
 >   reference to the reader to find out how it came about.

I'm not quite sure how to write out a distinguished name!  I believe
that the reference to X.500 is relevant.   This specification is
fairly meaningless unless you konw X.500

 >
 >   Does the specification meet all the goals laid out in section 2?

It meets goals 2 and 3.  Goals 1 and 4 are subjective, but I believe
that it meets them in the context of practical deployment of X.500 to
date.

 >
 >-sri (Srinivas R. Sataluri, AT&T Bell Labs., 908-949-7782, [e-mail])


thanks for these useful comments



Steve