<?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 for Bread and Circuits</title>
	<atom:link href="http://vocamus.net/dave/?feed=comments-rss2" rel="self" type="application/rss+xml" />
	<link>http://vocamus.net/dave</link>
	<description>Philosophy for the programming set, served on home made bread</description>
	<lastBuildDate>Fri, 26 Apr 2013 17:44:39 +0000</lastBuildDate>
	
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Introducing Maker Party 2013. Join the open online course. Building Webmaker 2.0. &#124; openmatt</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154919</link>
		<dc:creator>Introducing Maker Party 2013. Join the open online course. Building Webmaker 2.0. &#124; openmatt</dc:creator>
		<pubDate>Fri, 26 Apr 2013 17:44:39 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154919</guid>
		<description>[...] Thought experiment: letting git normalize whitespace [...]</description>
		<content:encoded><![CDATA[<p>[...] Thought experiment: letting git normalize whitespace [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Lonnen</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154677</link>
		<dc:creator>Lonnen</dc:creator>
		<pubDate>Mon, 22 Apr 2013 16:30:08 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154677</guid>
		<description>Potch has been messing around with a commit hook that uses uglify to beautify the code before each commit. This way you can write whatever style you want, and it will be transformed into an AST and then into a style-compliant format.

Of course, the canonical repo may still offend your personal taste. That&#039;s where sourcemaps-aware git extensions come into play, generating mappings between repos based on .style rules...</description>
		<content:encoded><![CDATA[<p>Potch has been messing around with a commit hook that uses uglify to beautify the code before each commit. This way you can write whatever style you want, and it will be transformed into an AST and then into a style-compliant format.</p>
<p>Of course, the canonical repo may still offend your personal taste. That&#8217;s where sourcemaps-aware git extensions come into play, generating mappings between repos based on .style rules&#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by CAFxX</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154619</link>
		<dc:creator>CAFxX</dc:creator>
		<pubDate>Sun, 21 Apr 2013 14:05:22 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154619</guid>
		<description>There was a poster at the most recent LLVM dev meeting about this: CELS, Code Editing in Local Style - http://llvm.org/devmtg/2013-04/#posterabstract</description>
		<content:encoded><![CDATA[<p>There was a poster at the most recent LLVM dev meeting about this: CELS, Code Editing in Local Style &#8211; <a href="http://llvm.org/devmtg/2013-04/#posterabstract" rel="nofollow">http://llvm.org/devmtg/2013-04/#posterabstract</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Simon</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154609</link>
		<dc:creator>Simon</dc:creator>
		<pubDate>Sun, 21 Apr 2013 10:06:30 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154609</guid>
		<description>The biggest problem with your proposal seems to be that it requires the version control tool to recognise the syntax of different languages in order to determine a) how to safely remove whitespace, and b) how to put it back again.

And that&#039;s *not* a trivial task - indeed, it&#039;s a task that&#039;s orders of magnitude more complex than actually being the version control tool that it&#039;s supposed to be. Even the slightest bug can become a data corruption issue, e.g if it messes up with a whitespace-sensitive language like Python.</description>
		<content:encoded><![CDATA[<p>The biggest problem with your proposal seems to be that it requires the version control tool to recognise the syntax of different languages in order to determine a) how to safely remove whitespace, and b) how to put it back again.</p>
<p>And that&#8217;s *not* a trivial task &#8211; indeed, it&#8217;s a task that&#8217;s orders of magnitude more complex than actually being the version control tool that it&#8217;s supposed to be. Even the slightest bug can become a data corruption issue, e.g if it messes up with a whitespace-sensitive language like Python.</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Jesper Kristensen</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154608</link>
		<dc:creator>Jesper Kristensen</dc:creator>
		<pubDate>Sun, 21 Apr 2013 07:30:04 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154608</guid>
		<description>I think automatic whitespace formatting is great, but it does not belong in a version control system.

When I choose version control system for a project, I usually prefer Mercurial over Git, and the main reason is that the Windows version of Git has turned on core.autocrlf by default. It creates all sorts of problems when somebody not on Windows has created a repository, and when you check it out on Windows, &quot;git status&quot; immediately reports all files as changed, and you cannot &quot;git pull&quot; because of all the uncommitted changes, yet when you try to &quot;git commit&quot;, you cannot do that either, because there are no changes to commit. Of course I turn off core.autocrlf in my local Git installation, but dealing with people who haven&#039;t done the same is so annoying that I prefer not using Git.

The fundamental problem is that Git breaks the principle of &quot;do one thing and do it well&quot;. End of line characters is not a version control task and should not be done by a version control system. It is a text editing task, and it should be handled by your code editor/IDE. Unfortunately most code editors and IDEs do so horribly (just like they often fail to have sensible handling of character encodings and similar things). But poor code editors is no good excuse for doing a code editor&#039;s job in a VCS.

Some code editors and IDEs actually do a pretty good job of whitespace formatting. For example in Eclipse, Ctrl+Shift+F works very well. But it is not perfect. It does not support using different formatting rules for saving versus displaying, which is what you want here, but I personally don&#039;t really care what whitespace formatting rule I use, as long as it is consistent within a project. The Eclipse formatter also does not support saving the rules in a standardized IDE independent format. But the fix should be to improve the IDE functionality, not to add IDE functionality to the VCS.</description>
		<content:encoded><![CDATA[<p>I think automatic whitespace formatting is great, but it does not belong in a version control system.</p>
<p>When I choose version control system for a project, I usually prefer Mercurial over Git, and the main reason is that the Windows version of Git has turned on core.autocrlf by default. It creates all sorts of problems when somebody not on Windows has created a repository, and when you check it out on Windows, &#8220;git status&#8221; immediately reports all files as changed, and you cannot &#8220;git pull&#8221; because of all the uncommitted changes, yet when you try to &#8220;git commit&#8221;, you cannot do that either, because there are no changes to commit. Of course I turn off core.autocrlf in my local Git installation, but dealing with people who haven&#8217;t done the same is so annoying that I prefer not using Git.</p>
<p>The fundamental problem is that Git breaks the principle of &#8220;do one thing and do it well&#8221;. End of line characters is not a version control task and should not be done by a version control system. It is a text editing task, and it should be handled by your code editor/IDE. Unfortunately most code editors and IDEs do so horribly (just like they often fail to have sensible handling of character encodings and similar things). But poor code editors is no good excuse for doing a code editor&#8217;s job in a VCS.</p>
<p>Some code editors and IDEs actually do a pretty good job of whitespace formatting. For example in Eclipse, Ctrl+Shift+F works very well. But it is not perfect. It does not support using different formatting rules for saving versus displaying, which is what you want here, but I personally don&#8217;t really care what whitespace formatting rule I use, as long as it is consistent within a project. The Eclipse formatter also does not support saving the rules in a standardized IDE independent format. But the fix should be to improve the IDE functionality, not to add IDE functionality to the VCS.</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by mzu</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154606</link>
		<dc:creator>mzu</dc:creator>
		<pubDate>Sun, 21 Apr 2013 05:46:23 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154606</guid>
		<description>Dave^H^H^H^H Mozilla you should use tabs.
And this ;-)

https://www.kernel.org/doc/Documentation/CodingStyle</description>
		<content:encoded><![CDATA[<p>Dave^H^H^H^H Mozilla you should use tabs.<br />
And this <img src='http://vocamus.net/dave/wp-includes/images/smilies/icon_wink.gif' alt=';-)' class='wp-smiley' /> </p>
<p><a href="https://www.kernel.org/doc/Documentation/CodingStyle" rel="nofollow">https://www.kernel.org/doc/Documentation/CodingStyle</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Havvy</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154604</link>
		<dc:creator>Havvy</dc:creator>
		<pubDate>Sun, 21 Apr 2013 05:04:35 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154604</guid>
		<description>Aww, the whitespace in my code blocks got ignored....</description>
		<content:encoded><![CDATA[<p>Aww, the whitespace in my code blocks got ignored&#8230;.</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Havvy</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154603</link>
		<dc:creator>Havvy</dc:creator>
		<pubDate>Sun, 21 Apr 2013 05:04:06 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154603</guid>
		<description>There would still be a style guide. For example, the following function could be written in two ways:

&lt;code&gt;function repeatTenTimes (fn) {
    for (var ix = 0; ix &lt; 10; ix++) {
        fn();
    }
}

function repeatTenTimes (fn) {
    var ix;
    for (ix = 0; ix &lt; 10; ix++) {
        fn();
    }
}&lt;/code&gt;

Not to mention variable names, which loop type to use, ect.

Whitespace is only a small portion of the information in a file that can vary without actually affecting the code that gets written.

Also, how does git know whether whitespace is significant or not?</description>
		<content:encoded><![CDATA[<p>There would still be a style guide. For example, the following function could be written in two ways:</p>
<p><code>function repeatTenTimes (fn) {<br />
    for (var ix = 0; ix &lt; 10; ix++) {<br />
        fn();<br />
    }<br />
}</p>
<p>function repeatTenTimes (fn) {<br />
    var ix;<br />
    for (ix = 0; ix &lt; 10; ix++) {<br />
        fn();<br />
    }<br />
}</code></p>
<p>Not to mention variable names, which loop type to use, ect.</p>
<p>Whitespace is only a small portion of the information in a file that can vary without actually affecting the code that gets written.</p>
<p>Also, how does git know whether whitespace is significant or not?</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Thought experiment: letting git normalize whitespace by Tomas Carnecky</title>
		<link>http://vocamus.net/dave/?p=1602&#038;cpage=1#comment-154594</link>
		<dc:creator>Tomas Carnecky</dc:creator>
		<pubDate>Sat, 20 Apr 2013 21:35:51 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1602#comment-154594</guid>
		<description>One argument against storing text compressed (whitespace removed) is that it makes diffs pretty much useless.

Take a look at clean/smudge filters (https://www.kernel.org/pub/software/scm/git/docs/gitattributes.html), they can be used to automatically clean up files when checking them in/out.</description>
		<content:encoded><![CDATA[<p>One argument against storing text compressed (whitespace removed) is that it makes diffs pretty much useless.</p>
<p>Take a look at clean/smudge filters (<a href="https://www.kernel.org/pub/software/scm/git/docs/gitattributes.html" rel="nofollow">https://www.kernel.org/pub/software/scm/git/docs/gitattributes.html</a>), they can be used to automatically clean up files when checking them in/out.</p>
]]></content:encoded>
	</item>
	<item>
		<title>Comment on Toward Webmaker Custom Elements, Web Components by New online course for mentors. Popcorn recipe videos. Webmaking in space. &#124; openmatt</title>
		<link>http://vocamus.net/dave/?p=1598&#038;cpage=1#comment-154402</link>
		<dc:creator>New online course for mentors. Popcorn recipe videos. Webmaking in space. &#124; openmatt</dc:creator>
		<pubDate>Fri, 12 Apr 2013 17:53:01 +0000</pubDate>
		<guid isPermaLink="false">http://vocamus.net/dave/?p=1598#comment-154402</guid>
		<description>[...] Toward Webmaker Custom Elements, Web Components [...]</description>
		<content:encoded><![CDATA[<p>[...] Toward Webmaker Custom Elements, Web Components [...]</p>
]]></content:encoded>
	</item>
</channel>
</rss>
