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

RE: [Ltru] Re: For review: Tagging text with no language

From: CE Whitehead <cewcathar@hotmail.com>
Date: Sat, 14 Apr 2007 13:12:06 -0400
Message-ID: <BAY114-F204E71C0AAA51A96796238B35C0@phx.gbl>
To: petercon@microsoft.com, www-international@w3.org, ltru@lists.ietf.org

>The meaning of zxx must be interpreted in terms of the coding standard of 
>which it is a part. ISO 639 is explicitly about coding human languages. "No 
>linguistic content" in the case of zxx means 'no content in any human 
>language'. If a language tag must be applied to something like "ifdef 
>DEBUG", then the appropriate language subtag would be zxx. ISO 639 cares 
>not if an IETF tag zxx-Latn or zxx-progrmng is constructed; but consistency 
>with ISO 639 would imply that the language subtag zxx be used.

Hi, Peter, I would only really disagree with what you say if there were 
programming languages where the words used to control the code were in other 
languages in use,
so that one language say made use of "while"
and another  of "lorsque"
but otherwise,

zxx I gather keeps the spelling checker from checking the language of the 
but as authors embed alerts, which are in one language, in code where the 
keywords or whatever are in quite another language,
one might want a way to turn off the zxx code or to include a second code 
too and there is none that I know of.

(Example: you might have the programming language in English and the alert 
in German.
I guess it would be tough to pick a language identifier in that case.
I do not know what the spelling checkers do with mis or mul either
. . . though I know that zxx is the only subtag recommended so what the 
spelling checkers do is irrelevant if everyone uses zxx

You have to check the content language of alerts separately I guess)

--C. E. Whitehead

>From: Karen_Broome@spe.sony.com
>To: Jukka K. Korpela

>I agree with you. I would also say that indicating that computer languages 
>have "no linguistic content" is not quite accurate for the same reasons you 
>cite below and disagrees with the origin of the zxx tag, which was to 
>indicate a total absence of language, not content which is not in a natural 
>Karen Broome
>www-international-request@w3.org wrote on 04/13/2007 01:08:33 PM:
> >
> > On Fri, 13 Apr 2007, Karen_Broome@spe.sony.com wrote:
> >
> > > With respect to computer language snippets, isn't that what the <code> 
> > > is for -- at least in XHTML?
> >
> > <code> has been in HTML since the beginning, and it indeed indicates
> > computer code, or "computer language".
> >
> > But to make things more difficult, computer code - such as program code,
> > operating system commands, or markup - can be regarded as being in some
> > human language(s) in the sense that the identifiers may have been formed
> > from words in human languages and the comments are written in some human
> > languages. Knowing the language of computer code can be relevant at 
> > for two purposes:
> > a) the understandability of the code to human readers depends on whether
> >     they understand the human language(s) used
> > b) speech synthesis would benefit from the use of language information.
> > Moreover, a checking program could e.g. check that the code follows the
> > syntactic rules of the programming, command, or markup "language" _and_
> > check the comments using a suitable human language spelling checker.
> >
> > Thus, I would say that the human language some text and the role of that
> > text as computer code are in a sense orthogonal.
> >
> > --
> > Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/
> >
> >
> >

Mortgage rates near historic lows. Refinance $200,000 loan for as low as 
Received on Saturday, 14 April 2007 17:12:21 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 21 September 2016 22:37:28 UTC