> From: UNICODE-MARC Discussion List
> [mailto:[log in to unmask]] On Behalf Of Smith,Gary
> Sent: 09 June, 2006 12:48
> To: [log in to unmask]
> Subject: Re: [UNICODE-M] MARC Proposal 2006-09 on lossless
> conversion of Unicode to MARC-8
>
> Stephen Toney wrote:
>
> > Gary Smith wrote,
> >> It's easier for the producing program to output
> fixed-length strings
> >> for the common cases,
>
> > ST: As the developer of MARCView(tm) and MARConvert(tm) I beg to
> differ.
> > What you say depends on the code. In our code, adding leading zeros
> > would require an extra step. I don't see how you can speak so
> generally
> > about ALL software.
>
> You're right -- I over-generalized. I don't know of a
> situation where it's significantly more difficult to produce
> the leading zeros, but I grant that they may exist.
Regardless of how easy, or not, it is to produce leading zeros,
I thought I read at one point, a W3C standard or other standard,
that it was either recommended or required that leading zeros be
eliminated for numeric or hexadecimal entities. Can anyone
produce a reference.
It would seem that
1) one should try to follow other recommendations and
2) this issue is a non-issue since it quite easy one the
receiving side to process either form; I know Gary
has an adversion to allow more than one way to do the
same thing in a standard, but...
Andy.
|