Comments on: The Carolino Effect http://www.robweir.com/blog/2008/03/carolino-effect.html Thinking the unthinkable, pondering the imponderable, effing the ineffable and scruting the inscrutable Wed, 10 Mar 2010 01:51:22 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: Jay, writer Memberspeed.com http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1614 Jay, writer Memberspeed.com Tue, 11 Mar 2008 01:26:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1614 Well, you certainly put that into terms people would understand. Honestly, I'm not really familiar with computer codes and the like but the story has sort of created an idea inside my head. I'm still trying to grasp the concepts. Still, good introduction. Well, you certainly put that into terms people would understand. Honestly, I’m not really familiar with computer codes and the like but the story has sort of created an idea inside my head. I’m still trying to grasp the concepts. Still, good introduction.

]]> By: John Mann http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1575 John Mann Thu, 06 Mar 2008 04:37:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1575 I think the that the MS binary file documentation gives you a forward mapping .doc => description.<br/><br/>And MSOOXML gives you a forward mapping .docx => description.<br/><br/>This isn't sufficient to create a quality forward+reverse mapping from .doc => anything => .docx.<br/><br/>Similarly, if Pedro Carolino only had a Portuguese=>French phrasebook and a English=>French phrasebook (but no French=>English phrasebook) his task would have been even harder. I think the that the MS binary file documentation gives you a forward mapping .doc => description.

And MSOOXML gives you a forward mapping .docx => description.

This isn’t sufficient to create a quality forward+reverse mapping from .doc => anything => .docx.

Similarly, if Pedro Carolino only had a Portuguese=>French phrasebook and a English=>French phrasebook (but no French=>English phrasebook) his task would have been even harder.

]]> By: Anonymous http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1570 Anonymous Wed, 05 Mar 2008 21:05:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1570 Brazil shares these concerns and provide context what happened at the BRM regarding this mapping:<br/><br/>"I have never seen a person so nervous and ashamed in my life… He said that Microsoft should have this mapping and if we want, we can ask it to Microsoft but not ask it to ECMA."<br/><br/>http://homembit.com/2008/03/at-the-end-what-we-did-in-geneva.html Brazil shares these concerns and provide context what happened at the BRM regarding this mapping:

“I have never seen a person so nervous and ashamed in my life… He said that Microsoft should have this mapping and if we want, we can ask it to Microsoft but not ask it to ECMA.”

http://homembit.com/2008/03/at-the-end-what-we-did-in-geneva.html

]]> By: Anonymous http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1568 Anonymous Wed, 05 Mar 2008 18:02:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1568 Other modern examples of this effect may come from poorly translated user manuals of imported products. <br/><br/>In Canada it is frequent that the French portion of the manual is so totally mangled that French Canadians must rely on the English version to make any sense of it. In many cases some French Canadians are offended and refuse to purchase the product.<br/><br/>It is also a huge source of unintentional humor. A classic is the "Made in Turkey" label translated as "Fabriqué en dinde", confusing the country for the well known edible bird. Other modern examples of this effect may come from poorly translated user manuals of imported products.

In Canada it is frequent that the French portion of the manual is so totally mangled that French Canadians must rely on the English version to make any sense of it. In many cases some French Canadians are offended and refuse to purchase the product.

It is also a huge source of unintentional humor. A classic is the “Made in Turkey” label translated as “Fabriqué en dinde”, confusing the country for the well known edible bird.

]]> By: Matthew Holloway http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1567 Matthew Holloway Wed, 05 Mar 2008 15:42:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1567 Oh, by the way. If you're looking to visualize voting scenarios on OOXML try my http://iso-vote.com site. It lets you toggle votes on different countries to see how this might play out.<br/><br/>(I'm the same guy who talked about the NZ-0051 comment) Oh, by the way. If you’re looking to visualize voting scenarios on OOXML try my http://iso-vote.com site. It lets you toggle votes on different countries to see how this might play out.

(I’m the same guy who talked about the NZ-0051 comment)

]]> By: James http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1565 James Wed, 05 Mar 2008 13:55:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1565 Or that other classic monty sketch "your wife has lovely boobies... bouncy bouncy" perhaps?<br/><br/>Certainly the contains of that poor owner's dictionary was about as much use to him as the current DIS29500 is to any office application implementor. Or that other classic monty sketch “your wife has lovely boobies… bouncy bouncy” perhaps?

Certainly the contains of that poor owner’s dictionary was about as much use to him as the current DIS29500 is to any office application implementor.

]]> By: Anonymous http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1563 Anonymous Wed, 05 Mar 2008 13:06:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1563 In addition the feature mapping there has been no information provided on which features support which versions.<br/><br/>For example, it would be impossible to round-trip a file with DrawingML to an Office 2000 binary as there is no equivalent.<br/><br/>NZ-0051 said that "The Specification contains the commands for every single feature that any Office version ever had, but does not tell the user which version."<br/><br/>In response Ecma said "it is not possible to create a simple one-to-one mapping between specific Office Open XML functionalities and specific versions of Microsoft Office applications because the features have evolved over time with each version of each application."<br/><br/>Now this statement of Ecmas' is just factually incorrect.<br/><br/>Here's the simple logic,<br/><br/>1. All features appeared at a particular time in Microsoft Office files. These can be documented, categorised.<br/><br/>2. If a file-format feature's usage changed (using the same semantics) then this can be documented too. Eg, in Office '97 font weight "bold" meant an OpenType weight of 300, but in Office 2003 it had an OpenType weight of 400 (this is a hypothetical example).<br/><br/>Without such a mapping the stated goals of OOXML appear more as advertising fluff than engineering criteria. In addition the feature mapping there has been no information provided on which features support which versions.

For example, it would be impossible to round-trip a file with DrawingML to an Office 2000 binary as there is no equivalent.

NZ-0051 said that “The Specification contains the commands for every single feature that any Office version ever had, but does not tell the user which version.”

In response Ecma said “it is not possible to create a simple one-to-one mapping between specific Office Open XML functionalities and specific versions of Microsoft Office applications because the features have evolved over time with each version of each application.”

Now this statement of Ecmas’ is just factually incorrect.

Here’s the simple logic,

1. All features appeared at a particular time in Microsoft Office files. These can be documented, categorised.

2. If a file-format feature’s usage changed (using the same semantics) then this can be documented too. Eg, in Office ‘97 font weight “bold” meant an OpenType weight of 300, but in Office 2003 it had an OpenType weight of 400 (this is a hypothetical example).

Without such a mapping the stated goals of OOXML appear more as advertising fluff than engineering criteria.

]]> By: Rob Brown http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1562 Rob Brown Tue, 04 Mar 2008 22:55:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1562 Oh, crud... that last comment came out all wrong. It should have ended with something like "as long as <b>any</b> complete and unambiguous mapping is available, then any other should be easy to create. And by not providing any usable mapping, OOXML prevents <i>any</i> faithful representation.<br/><br/>/RobBrown crawls back under his rock Oh, crud… that last comment came out all wrong. It should have ended with something like “as long as any complete and unambiguous mapping is available, then any other should be easy to create. And by not providing any usable mapping, OOXML prevents any faithful representation.

/RobBrown crawls back under his rock

]]> By: Rob http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1560 Rob Tue, 04 Mar 2008 22:39:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1560 Hmmm... I wonder. Manuel, on Faulty Towers, talks as if he could have been reading from this phrase book. So maybe there is a Python connection. Hmmm… I wonder. Manuel, on Faulty Towers, talks as if he could have been reading from this phrase book. So maybe there is a Python connection.

]]> By: Rob Brown http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1561 Rob Brown Tue, 04 Mar 2008 22:39:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1561 I'd go further, and say that with one endpoint (the binary file) and the mapping, the other endpoint becomes arbitrary.<br/><br/>I fail to understand how the "existing corpus" of documents could <i>not</i> be represented 100% faithfully in Braille, Morse code, cave drawings, or even (shock) ISO26300 ODF, as long as the mappings are available, complete, and unambiguous. I’d go further, and say that with one endpoint (the binary file) and the mapping, the other endpoint becomes arbitrary.

I fail to understand how the “existing corpus” of documents could not be represented 100% faithfully in Braille, Morse code, cave drawings, or even (shock) ISO26300 ODF, as long as the mappings are available, complete, and unambiguous.

]]> By: Anonymous http://www.robweir.com/blog/2008/03/carolino-effect.html#comment-1559 Anonymous Tue, 04 Mar 2008 22:34:00 +0000 http://www.robweir.com/blog/2008/03/the-carolino-effect.html#comment-1559 <a HREF="http://slashdot.org/firehose.pl?op=view&id=554916" REL="nofollow" rel="nofollow">Submitted it to Slashdot</a>. I wonder if this has any relation to that 'My hovercraft is full of eels' sketch? Submitted it to Slashdot. I wonder if this has any relation to that ‘My hovercraft is full of eels’ sketch?

]]>