W3C home > Mailing lists > Public > www-international@w3.org > April to June 2007

[Ltru] Re: "mis" update review request (from ietf-languages)

From: CE Whitehead <cewcathar@hotmail.com>
Date: Sat, 21 Apr 2007 14:26:46 -0400
Message-ID: <BAY114-F18344C5FFF68DB8A13E859B3550@phx.gbl>
To: www-international@w3.org

Hi,  I kind of like Mark's suggestion for making the definition of zxx 
clearer; it does not hurt anything
(below are my comments also, along with some copy of the discussion at 
>Mark Davis scripsit:

>>And as a *general* statement about applying zxx to programming
>>languages, I think that unless and until we add ", or programming
>>languages" to the Description, it is unsuitable for application to
>>programming languages.

>That would only be true if there were a clear and convincing argument,
>acceptable to all, that code is written in a natural language.  Otherwise,
>it is a mere judgement about a class of borderline cases.

>>don't think we are that far apart here -- we both agree that zxx ought
>>to be the right thing. Where we disagree is that you think it already
>>has that meaning, and I can't see it at all in the standard.

>All the standard says is "Non-linguistic content".  So to assert that
>code is not "zxx", you must say that it is linguistic content, though
>you are unwilling to apply "en" (or, presumably, any other specific or
>collective language tag).  What is this linguistic content that is not
>in any language?

>You have for whatever reason decided that reading and writing code
>(presumably exclusive of comments, which are unquestionably in English
>or French or Japanese or what have you) is linguistic behavior, but why?
>Learning to program is very unlike either acquiring French or Esperanto
>as a first language, or learning them as a second language, though
>admittedly more like the latter than the former.

Sorry, I think that while loops and such have something of the logic of 
But you're right, a few keywords are in English or whatever!

>Are you also willing
>to say that mathematical formulae embedded in a math paper in English,
>French, German, Russian, or what have you are linguistic content, though
>certainly not in any of those languages?

>>  But since it is a broadening, we can make a change and then both of
>>  us would be happy (or at least happier). So what do you say we agree
>>on that, and be done with this part of the issue?

>Because it is not clear that it is a broadening, and it is very colorably
>an out-of-scope extension.  Why not mention musical notation (of
>instrumental compositions, where there are no lyrics to muddy the
>waters) as well?

>  John Cowan

John, I think Mark's clarification is a good clarification here,

though you also need to use zxx for programming code
(whether you embed it in some other declaration or not);
for if you have code with snippets of French or Chinese in a document.write 
the French or Chinese will just be the unicode representations of the 
characters anyway, & not the characters themselves

the comments in Frnech or Chinese do not display & so special characters 
would not display anyway

(French works o.k. though in comments; you just leave off the accents;
so you can have French or any Latin character set languages
in the code part of a document)

--C. E. Whitehead
I personally still think the current best way to tag snippets of code in a 
document where other languages are embedded in the code
might be a series of embeddings--
<div id="morceau_de_code" lang="fr">
Banniere Publicitaire (de Raggett):  Une adaptation francaise
<code lang="zxx">
// Dans le codage ci-dessous il s'agit d'une adaption en francais du codage 
de Dave Raggett qui se trouve
// dans le didacticiel du W3C,
// "Advanced HTML", http://www.w3.org/MarkUp/Guide/Advanced.html
if (document.images)
    imagesdesPubs = new 
    URLsdesPubs = new Array("www.csail.mit.edu",
                "www.ercim.org", "www.keio.ac.jp");
    cePub = 0;

Mortgage refinance is Hot. *Terms. Get a 5.375%* fix rate. Check savings 
Received on Saturday, 21 April 2007 18:27:18 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:04:25 UTC