<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: The L20n Format Shootout</title>
	<atom:link href="http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/</link>
	<description>localization and community at mozilla</description>
	<lastBuildDate>Tue, 07 Jun 2011 10:48:15 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.2.1</generator>
	<item>
		<title>By: Robert Kaiser</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143255</link>
		<dc:creator>Robert Kaiser</dc:creator>
		<pubDate>Thu, 27 Aug 2009 18:38:58 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143255</guid>
		<description>YAML also seems to run into problems with strings that should start or, what we have more frequently in our code, end with a space. Also, multi-line strings are hard to represent from what I see, but again, this is something we have quite frequently in our code.
That said, I&#039;m generally not a fan of indentation that is relevant to evaluating the code, but that might just be me. I think it&#039;d be confusing enough for a number of localizers, though.

I&#039;m not necessarily a fan of all details in terms of punctuation choices in the proposed LOL format, but I think it&#039;s the most simple to understand for the person editing such a file, i.e. the localizer (though I&#039;m sure with time there will be tools to abstract it away from the non-technical localizers, there are always more technical ones who at least want to look at the files, maybe directly edit them).</description>
		<content:encoded><![CDATA[<p>YAML also seems to run into problems with strings that should start or, what we have more frequently in our code, end with a space. Also, multi-line strings are hard to represent from what I see, but again, this is something we have quite frequently in our code.<br />
That said, I&#8217;m generally not a fan of indentation that is relevant to evaluating the code, but that might just be me. I think it&#8217;d be confusing enough for a number of localizers, though.</p>
<p>I&#8217;m not necessarily a fan of all details in terms of punctuation choices in the proposed LOL format, but I think it&#8217;s the most simple to understand for the person editing such a file, i.e. the localizer (though I&#8217;m sure with time there will be tools to abstract it away from the non-technical localizers, there are always more technical ones who at least want to look at the files, maybe directly edit them).</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jeremy Hiatt</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143217</link>
		<dc:creator>Jeremy Hiatt</dc:creator>
		<pubDate>Wed, 26 Aug 2009 19:07:09 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143217</guid>
		<description>Thanks everyone for your comments.

@Dave: You&#039;re right that LOL is the most effective format for expressing these objects. The issue is that using a proprietary format makes it that much harder to port l20n to other platforms, so we wanted to gather some feedback on that. Still, I&#039;d have to agree that JSON and YAML really don&#039;t work particularly well and that LOL looks like the best choice.

@johnjbarton: I&#039;d like to hear more about why you don&#039;t like DTD. Is it the angle bracket  syntax, or something else? As far as your properties suggestion, I do like a few things about that format, especially that it doesn&#039;t use a lot of punctuation. However, that also means that it&#039;s clunky to express an object with a lot of properties, requiring a lot of repetition of the prefix. One issue with the example you proposed is that it can&#039;t be parsed unambiguously, since the intended index could be appName or appName.gender. That&#039;s not to say that there aren&#039;t other ways to make this work in a properties-like format, but ultimately it seems like it would end up being quite confusing. I&#039;d be interested to see you prove me wrong, though!

@Sebastian, Archaeopteryx: I&#039;m going to pass the blame on to Axel Hecht here, who wrote the example. I don&#039;t speak German, so I&#039;m not really qualified to weigh in on this one. He&#039;d also be the one to ask about the choice of beer as the example. Thanks for bringing this up!</description>
		<content:encoded><![CDATA[<p>Thanks everyone for your comments.</p>
<p>@Dave: You&#8217;re right that LOL is the most effective format for expressing these objects. The issue is that using a proprietary format makes it that much harder to port l20n to other platforms, so we wanted to gather some feedback on that. Still, I&#8217;d have to agree that JSON and YAML really don&#8217;t work particularly well and that LOL looks like the best choice.</p>
<p>@johnjbarton: I&#8217;d like to hear more about why you don&#8217;t like DTD. Is it the angle bracket  syntax, or something else? As far as your properties suggestion, I do like a few things about that format, especially that it doesn&#8217;t use a lot of punctuation. However, that also means that it&#8217;s clunky to express an object with a lot of properties, requiring a lot of repetition of the prefix. One issue with the example you proposed is that it can&#8217;t be parsed unambiguously, since the intended index could be appName or appName.gender. That&#8217;s not to say that there aren&#8217;t other ways to make this work in a properties-like format, but ultimately it seems like it would end up being quite confusing. I&#8217;d be interested to see you prove me wrong, though!</p>
<p>@Sebastian, Archaeopteryx: I&#8217;m going to pass the blame on to Axel Hecht here, who wrote the example. I don&#8217;t speak German, so I&#8217;m not really qualified to weigh in on this one. He&#8217;d also be the one to ask about the choice of beer as the example. Thanks for bringing this up!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Archaeopteryx</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143179</link>
		<dc:creator>Archaeopteryx</dc:creator>
		<pubDate>Tue, 25 Aug 2009 18:17:20 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143179</guid>
		<description>As Sebastian already indicated: &quot;Ein hübsches&quot; is neutral, &quot;Eine hübsche&quot; female.
Should we be concerned why you use alcoholic products to demonstrate the user of gender?</description>
		<content:encoded><![CDATA[<p>As Sebastian already indicated: &#8220;Ein hübsches&#8221; is neutral, &#8220;Eine hübsche&#8221; female.<br />
Should we be concerned why you use alcoholic products to demonstrate the user of gender?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Sebastian Redl</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143178</link>
		<dc:creator>Sebastian Redl</dc:creator>
		<pubDate>Tue, 25 Aug 2009 13:58:18 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143178</guid>
		<description>&quot;Ein hübscher ${appName}.&quot;
&quot;Eine hübsche ${appName}.&quot;</description>
		<content:encoded><![CDATA[<p>&#8220;Ein hübscher ${appName}.&#8221;<br />
&#8220;Eine hübsche ${appName}.&#8221;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: johnjbarton</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143176</link>
		<dc:creator>johnjbarton</dc:creator>
		<pubDate>Tue, 25 Aug 2009 05:30:20 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143176</guid>
		<description>Nothing similar to DTD please. I did not see anything about just using property files, eg
complex.indices.appName.gender.male = Ein hübscher ${appName}s.
complex.indices.appName.gender.female = Ein hübsches ${appName}s.

appName.gender.male = Jägermeister

jjb</description>
		<content:encoded><![CDATA[<p>Nothing similar to DTD please. I did not see anything about just using property files, eg<br />
complex.indices.appName.gender.male = Ein hübscher ${appName}s.<br />
complex.indices.appName.gender.female = Ein hübsches ${appName}s.</p>
<p>appName.gender.male = Jägermeister</p>
<p>jjb</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Dave</title>
		<link>http://blog.mozilla.com/seth/2009/08/24/the-l20n-format-shootout/comment-page-1/#comment-143169</link>
		<dc:creator>Dave</dc:creator>
		<pubDate>Tue, 25 Aug 2009 01:20:34 +0000</pubDate>
		<guid isPermaLink="false">http://blog.mozilla.com/seth/?p=547#comment-143169</guid>
		<description>I see no reason to not just pick the simplest format that&#039;s written for the job. Why bother trying to hack something else into doing this? JSON would be a mess here and YAML doesn&#039;t look much better. LOL seems the obvious choice to me.</description>
		<content:encoded><![CDATA[<p>I see no reason to not just pick the simplest format that&#8217;s written for the job. Why bother trying to hack something else into doing this? JSON would be a mess here and YAML doesn&#8217;t look much better. LOL seems the obvious choice to me.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

