Re: [xliff] RE: [xliff-comment] its namespace URL in xliff 2.1

[David Filip <david.filip@adaptcentre.ie>]

Thanks, Yves, Marc, all,

Since this needs fixed in XLIFF 2.1, we will accept this as a late cos01 public review comment.

We already agreed with OASIS Admin to withdraw our currently open cos01 progression ballot.

Unfortunately, I am on vacation until 3 Jan. But I will log this as a new cos01 public review issue the first thing in the morning on 3rd January and prepare a cos02 draft for the TC meeting on 16th January..

Cheers and thanks for your attention

dF

On Dec 26, 2017 7:55 PM, "Yves Savourel" <ysavourel@enlaso.com> wrote:

Hi all,

 

Yes, it is clear that the ITS namespace identified is http://www.w3.org/2005/11/its (http not https)

See https://www.w3.org/TR/its20/#notation-terminology

 

And it seems the XLIFF 2.1 current candidate document gives the wrong URI:

http://docs.oasis-open.org/xliff/xliff-core/v2.1/cos01/xliff-core-v2.1-cos01.html#d0e17864

There is a mix of http and https throughout the document.

 

So the candidate document must be fixed if possible before it moves further.

 

-ys

 

 

From: David Filip [mailto:david.filip@adaptcentre.ie]
Sent: Thursday, December 21, 2017 11:25 AM
To: Marc Mittag (MittagQI) <marc@mittagqi.com>
Cc: xliff-comment@lists.oasis-open.org
Subject: Re: [xliff-comment] its namespace URL in xliff 2.1

 

Hi Mark,

 

[technically you should have logged this against OKAPI Lynx, but any of the XLIFF 2 validators including the XLIFF 2.1 schemas will reject this..]

the validator is correct, the namespace identifier is "http://www.w3.org/2005/11/its" and is NOT "https://www.w3.org/2005/11/its"

 

It doesn't matter that the URL when used in browser does resolve into the https based URL.

I always wonder why using URLs (as opposed to URNs) became so wildly popular to identify namespaces, as it keeps confusing people. Namespace identifiers aren't supposed to resolve.. They often do but that's immaterial for validation..

 

Also notice that both resolve into a folder name "https://www.w3.org/2005/11/its/", still it would be wrong to use the trailing slash in the namespace identifier. The namespace identifier might resolve or not; also it might start resolving into something completely different, still the namespace identifier will always be just this exact string [within the quotes]: "https://www.w3.org/2005/11/its"..

 

I hope this helps

dF    

 

Dr. David Filip

===========

OASIS XLIFF OMOS TC Chair

OASIS XLIFF TC Secretary, Editor, Liaison Officer

Spokes Research Fellow

ADAPT Centre

KDEG, Trinity College Dublin

Mobile: +420-777-218-122

 

 

On Thu, Dec 21, 2017 at 2:37 PM, Marc Mittag (MittagQI) <marc@mittagqi.com> wrote:

Hello together,

Chase pointed me to commenting here on this:

I tried to validate 2 different xliff 2.1 files at

http://okapi-lynx.appspot.com/validation

One does validate, the other does not (one had errors in the its
module). The only difference between the 2 files was, that the one that
validates uses the namespace

xmlns:its="https://www.w3.org/2005/11/its"

instead of

xmlns:its="http://www.w3.org/2005/11/its"

So the only difference is the protocol that is used.

The validator did not check for correct its module syntax in my xliff,
when I used the https protocol in my namespace.

Yet in the xliff 2.1 spec it suggests to use the https protocol and this
is the one used at w3c at the current stage as well
(http://www.w3.org/2005/11/its/ gets forwarded to
https://www.w3.org/2005/11/its/)


So from my point of view the validator needs to be corrected here.

best

Marc
--
Marc Mittag
MittagQI - Quality Informatics

Konrad-Lorenz-Weg 10
D-72116 Mössingen
Germany
Tel.:   ++49 (0)7473/220202
Fax:    ++49 (0)7473/220211
mailto: Marc@MittagQI.com
Web:    www.MittagQI.com

Info zum Anhang anbei:
Standardmäßig liegt jeder E-Mail bei MittagQI die
PGP-Signatur unseres jeweiligen Mitarbeiters bei,
um Ihnen die Möglichkeit zu geben,
E-Mails an uns zu verschlüsseln.