OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

xliff message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [xliff] Re: XLIFF 1.1 Specification Working Draft 14 - RC 5


Yves, et al.,

--------------------------------
re: rid and id
--------------------------------

This is something we need to determine in today's conference call. The
specification is not clear enough for the it to be implemented. My comments
and recommendations follow:

CLARIFICATION

Applies to tags:

bx, ex, bpt, ept

It was unclear to me how the rid attribute should be used. There are two
possible patterns.

(1) <bx id="b1" rid="e1"/> ... <ex id="e1" rid="b1"/>
  or
(2) <bx rid="1"/> ... <ex rid="1"/>

I recommend pattern (1) because it distinguishes the role of identifier and
reference to identifier. This is similar to the ID and IDREF types in DTD,
but not strictly enforced because 'id' values do not have to be unique
within the document. This patterns requires that both the 'id' and 'rid'
attributes be required.

An example or description of how to set the 'id' and 'rid' attributes is
needed to be unambiguous.


RECOMMENDATION

Applies to tags:

bx, ex, bpt, ept

The 'rid' attribute should be required because it is needed to refer to the
matching begin/end tag.

This change is not fully compatible with XLIFF 1.0 because some documents
may not have the 'rid' attribute, but I do not know how the matching pairs
would be identified.

The 'id' attribute is currently required and should remain required so the
'rid' attribute can refer to it.



-----------------------------------
re: xml:lang
-----------------------------------

I like your wording: "the values of xml:lang are expected to be the same as
for the source-language and target-language, except in the case described".
However, I would add "If xml:lang is used" to the front as in:

"If xml:lang is used, the values of xml:lang are expected to be the same as
for the source-language and target-language, except in the case described."

It conveys the same meaning. To me, this is just an editorial matter, so
please word it as you see fit. I had originally suggested not using xml:lang
so that the XLIFF file would be smaller and to avoid redundant information.



Regards,

Doug Domeny
Software Analyst

Ektron, Inc.
+1 603 594-0249 x212
http://www.ektron.com



-----Original Message-----
From: Yves Savourel [mailto:ysavourel@translate.com]
Sent: Monday, May 05, 2003 5:08 PM
To: ddomeny@ektron.com; Tony Jewtushenko
Cc: xliff@lists.oasis-open.org
Subject: RE: [xliff] Re: XLIFF 1.1 Specification Working Draft 14 - RC 5


> 2.4 Inline Elements
> "These paired elements are related via their rid attributes."
> I'm not sure how to use the rid attribute. Is the rid attribute set
> to the value of the id attribute in the corresponding element? An
> example would be very helpful.

Actually, I'm also a little lost with that one too. I thought <bpt> and
<ept>, as well as <bx/> and <ex/> were linked with their id attributes. The
rid attribute is optional, id is mandatory. If you have more than 2 <bpt>
elements, each <ept> must have a relation back to to opening counterpart,
since id is the only mandatory attribute in <ept> it makes sense to use it
(otherwise you end up with 99% of duplication between id and rid).

I also see a problem in "Use the required id attribute to relate begin and
end <g> tags." for the use of <g>. This cannot be done as there are no
attribute in a closing tag. You always know which <g> element you are
closing. I would use <g> for place-holder, and <bx/>/<ex/> only when they
are overlapping codes.



> 2.5.2 Adding Attributes
> In the example, the <trans-unit id> attributes
> all have the value '1'. They should be different.

Oops, that one has been there for some time: my fault: a copy/paste gone
bad.



I would also add one comment on section 2 General structure.
The end of the new paragraph about the use of xml:lang reads:
"The xml:lang attribute should not be used in those elements. The exception
is that <source> and <target> elements that are children of <alt-trans> may
contain an xml:lang attribute of a different language than that of the
source-language and target-language attributes of the <file> element."

I know I'm too late to change things, but I'd like to say something for the
record:

I rather disagree with this: I think it would set a bad example and
precedent to have a guideline going against XML internationalization
practices. What we probably should say is that the values of xml:lang are
expected to be the same as for the source-language and target-language,
except in the case described. But saying xml:lang *should not* be use seems
harmful: what if a XML-enabled but not XLIFF-enabled tool needs the language
code? (like a spell-checker).

In addition, such new guideline also contradicts the wording we have already
decided (long ago) for source-language and target-language): "The source
language can be also specified by xml:lang in each <source> element. The
value of source-language and xml:lang can be different to allow having
different source languages if necessary (for example in an <alt-trans>
element)."

Just a thought.

kenavo,
-yves





[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]