[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: version numbers and infosets

• To: <xml-dev@l...>, <xml-editor@w...>
• Subject: Re: version numbers and infosets
• From: "Rick Jelliffe" <ricko@a...>
• Date: Wed, 24 Jul 2002 14:47:37 +1000
• References: <200207231859.OAA21280@m...>

From: "John Cowan" <jcowan@r...>

> People use mainframes, and XML too.  If XML 1.0 had insisted
> that only CR-LF and LF were acceptable line terminators, don't you
> think an argument based on justice for Mac users would have been
> appropriate?

No, supporting Mac-style line terminators does not need to grandized into
an issue of justice (by which John means egalitiarianism, I think*) but

 * pragmatism: the number of people typing documents in editors
  on Macs is very large

 * operational considerations: a programmer generally has control over the
  line terminators they use, and mainframes are typically programmed

 * XML 1.0 was starting fresh, and therefore only had SGML and RFC
  compatability to cope with: XML 1.n must be a trade-off between the
  particular benefits and the costs in distruption. 

 * CR and LF were in play any way, so a decision on how to support CR
  alone is necessary even as a matter of error handling

 * CR is part of ASCII, and so it is low-hanging fruit to adopt. 

 * XML had the 80/20 rule, and I am not sure that NEL would
  have made it even if it into XML 1.0 in any case.

NEL on the other hand is rare, specialist, non-ASCII, and not low-hanging fruit.

> "It is intolerable to have no better reason for a legal rule than
> that *thus* it was laid down in the time of King Henry the Second."

For some reasons, see above. 

Cheers
Rick Jelliffe

* IMHO It is not injust if W3C were to decide on only ASCII characters in names,
for example.  But would be anti-egalitarian, and obviously bad for the market.

• References:
  • Re: version numbers and infosets
    • From: John Cowan <jcowan@r...>

• Prev by Date: Re: version numbers and infosets
• Next by Date: Re: URIs harmful (was RE: Article: Keeping pa cewith James Clark)
• Previous by thread: Re: version numbers and infosets
• Next by thread: Re: version numbers and infosets
• Index(es):
  • Date
  • Thread