<?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: Don&#8217;t Mess With Jay</title>
	<atom:link href="http://ma.tt/2004/06/dont-mess-with-jay/feed/" rel="self" type="application/rss+xml" />
	<link>http://ma.tt/2004/06/dont-mess-with-jay/</link>
	<description>Unlucky in Cards</description>
	<lastBuildDate>Sat, 26 May 2012 13:00:00 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.4-beta4-20941</generator>
	<item>
		<title>By: Jay Allen</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4725</link>
		<dc:creator>Jay Allen</dc:creator>
		<pubDate>Wed, 02 Jun 2004 18:50:23 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4725</guid>
		<description>&lt;em&gt;But first let&#039;s get some deployment. No point in creating a demand where there&#039;s no supply â€¦&lt;/em&gt;

Good point.</description>
		<content:encoded><![CDATA[<p><em>But first let&#8217;s get some deployment. No point in creating a demand where there&#8217;s no supply â€¦</em></p>
<p>Good point.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jacques Distler</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4718</link>
		<dc:creator>Jacques Distler</dc:creator>
		<pubDate>Wed, 02 Jun 2004 13:25:36 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4718</guid>
		<description>&quot;Does post-submissions transformation (a la Markdown, textile, etc) hinder verification?&quot;

Well, it sure prevents you from copying and pasting my &lt;em&gt;formatted&lt;/em&gt; comment and trying to verify that. (As does my use of &lt;em&gt;...&lt;/em&gt; in the previous sentence. Note, too that the 5 consecutive dashes that demarcate the signature have been turned into &quot;&#8212;&#8211;&quot;.)

Srijith&#039;s plugin gets around this by providing the &quot;raw&quot; comment in a &lt;code&gt;textarea&lt;/code&gt;, where you can copy and paste it for manual verification. With server-side verification, of course, you get around these problems because the server has access to the raw, unfiltered comment in the database.

I think server-side verification is important. Most people (including me) are too lazy to verify the comment manually, unless they are &lt;em&gt;highly&lt;/em&gt; suspicious of its provenance. But I&#039;m not averse to clicking on a link and having the server verify it for me. I won&#039;t say I do that for &lt;em&gt;every single&lt;/em&gt; PGP-signed comment, but I do it enough of the time that I&#039;m pretty likely to catch any frauds. (For comments on my own blog, I get the comments mailed to me, and GPGMail bundle takes care of verifying them.)

Even better (though this would take a bit of re-architecting) would be for the server to cache the verification status, and display an icon right there in the inline comment listings.

Originally, Srijith and I decided that verifying the comment when it is first posted would create an unacceptable delay. Fetching the commenter&#039;s public key can take some time, and posting comments in MT already requires a fair bit of page-rebuilding. Now that MT can perform such tasks in the background, perhaps that might be revisited.

WordPress, which uses dynamic pages could certainly do it that way, as there&#039;s no constraint of verifying the comment before the page is &quot;rebuilt&quot;.

&quot;Of course, the implementation won&#039;t become widespread unless users as a whole or a vendor/developer of weblogging software includes it by default or people see its usefulness. It&#039;s a chicken/egg problem.&quot;

Yeah, 6A should include Srijith&#039;s plugin with the default install (after all, it uses Ben Trott&#039;s &lt;code&gt;Crypt::OpenPGP&lt;/code&gt; module). And, whenever there&#039;s an actual &lt;em&gt;working&lt;/em&gt; version, Matt should include the OpenPGP Comment plugin with the default distribution of WordPress.

As to whether people will see the need for it, you clearly do. Matt didn&#039;t used to, but I &lt;em&gt;think&lt;/em&gt; he does now.

All it takes to convince someone of the need is to point them to your blog post.

But first let&#039;s get some deployment. No point in creating a demand where there&#039;s no supply ...</description>
		<content:encoded><![CDATA[<p>&#8220;Does post-submissions transformation (a la Markdown, textile, etc) hinder verification?&#8221;</p>
<p>Well, it sure prevents you from copying and pasting my <em>formatted</em> comment and trying to verify that. (As does my use of &lt;em&gt;&#8230;&lt;/em&gt; in the previous sentence. Note, too that the 5 consecutive dashes that demarcate the signature have been turned into &#8220;&mdash;&ndash;&#8221;.)</p>
<p>Srijith&#8217;s plugin gets around this by providing the &#8220;raw&#8221; comment in a <code>textarea</code>, where you can copy and paste it for manual verification. With server-side verification, of course, you get around these problems because the server has access to the raw, unfiltered comment in the database.</p>
<p>I think server-side verification is important. Most people (including me) are too lazy to verify the comment manually, unless they are <em>highly</em> suspicious of its provenance. But I&#8217;m not averse to clicking on a link and having the server verify it for me. I won&#8217;t say I do that for <em>every single</em> PGP-signed comment, but I do it enough of the time that I&#8217;m pretty likely to catch any frauds. (For comments on my own blog, I get the comments mailed to me, and GPGMail bundle takes care of verifying them.)</p>
<p>Even better (though this would take a bit of re-architecting) would be for the server to cache the verification status, and display an icon right there in the inline comment listings.</p>
<p>Originally, Srijith and I decided that verifying the comment when it is first posted would create an unacceptable delay. Fetching the commenter&#8217;s public key can take some time, and posting comments in MT already requires a fair bit of page-rebuilding. Now that MT can perform such tasks in the background, perhaps that might be revisited.</p>
<p>WordPress, which uses dynamic pages could certainly do it that way, as there&#8217;s no constraint of verifying the comment before the page is &#8220;rebuilt&#8221;.</p>
<p>&#8220;Of course, the implementation won&#8217;t become widespread unless users as a whole or a vendor/developer of weblogging software includes it by default or people see its usefulness. It&#8217;s a chicken/egg problem.&#8221;</p>
<p>Yeah, 6A should include Srijith&#8217;s plugin with the default install (after all, it uses Ben Trott&#8217;s <code>Crypt::OpenPGP</code> module). And, whenever there&#8217;s an actual <em>working</em> version, Matt should include the OpenPGP Comment plugin with the default distribution of WordPress.</p>
<p>As to whether people will see the need for it, you clearly do. Matt didn&#8217;t used to, but I <em>think</em> he does now.</p>
<p>All it takes to convince someone of the need is to point them to your blog post.</p>
<p>But first let&#8217;s get some deployment. No point in creating a demand where there&#8217;s no supply &#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: The Indiana Jones School of Management</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4717</link>
		<dc:creator>The Indiana Jones School of Management</dc:creator>
		<pubDate>Wed, 02 Jun 2004 13:05:47 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4717</guid>
		<description>&lt;strong&gt;I Told You So&lt;/strong&gt;
I love to say that I told you so.</description>
		<content:encoded><![CDATA[<p><strong>I Told You So</strong><br />
I love to say that I told you so.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jay Allen</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4716</link>
		<dc:creator>Jay Allen</dc:creator>
		<pubDate>Wed, 02 Jun 2004 11:52:42 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4716</guid>
		<description>And once again, I misspell your name.  My apologies.  My French originated from Cajun land, which may explain things.</description>
		<content:encoded><![CDATA[<p>And once again, I misspell your name.  My apologies.  My French originated from Cajun land, which may explain things.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jay Allen</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4715</link>
		<dc:creator>Jay Allen</dc:creator>
		<pubDate>Wed, 02 Jun 2004 11:51:44 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4715</guid>
		<description>GPGMail/GPGDropThing.    Not GPGDropMail.  See, even the names are hard, Jaques! :-)

Seriously though, as easy as signing and verifying MAY be (with the help of extras), until the implementation is widespread, it&#039;s usefulness is limited as you have demonstrated here.  Of course, the implementation won&#039;t become widespread unless users as a whole or a vendor/developer of weblogging software includes it by default or people see its usefulness.  It&#039;s a chicken/egg problem.

I do wonder one thing though: Does post-submissions transformation (a la Markdown, textile, etc) hinder verification?  I can&#039;t see how it wouldn&#039;t.  There&#039;s another problem to solve.</description>
		<content:encoded><![CDATA[<p>GPGMail/GPGDropThing.    Not GPGDropMail.  See, even the names are hard, Jaques! <img src='http://s.ma.tt/blog/wp-includes/images/smilies/icon_smile.gif' alt=':-)' class='wp-smiley' /> </p>
<p>Seriously though, as easy as signing and verifying MAY be (with the help of extras), until the implementation is widespread, it&#8217;s usefulness is limited as you have demonstrated here.  Of course, the implementation won&#8217;t become widespread unless users as a whole or a vendor/developer of weblogging software includes it by default or people see its usefulness.  It&#8217;s a chicken/egg problem.</p>
<p>I do wonder one thing though: Does post-submissions transformation (a la Markdown, textile, etc) hinder verification?  I can&#8217;t see how it wouldn&#8217;t.  There&#8217;s another problem to solve.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jacques Distler</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4714</link>
		<dc:creator>Jacques Distler</dc:creator>
		<pubDate>Wed, 02 Jun 2004 06:25:03 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4714</guid>
		<description>-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jay,

I, too, use the &lt;a href=&quot;http://www.sente.ch/software/GPGMail/English.lproj/GPGMail.html&quot;&gt;GPGMail bundle&lt;/a&gt; for Mail.app too. I sign all my outgoing email (and have been doing so for almost a decade).

For signing other stuff (like blog comments), try &lt;a href=&quot;http://prdownloads.sourceforge.net/macgpg/GPGDropThing-0.4.3.dmg.gz?download&quot;&gt;GPGDropThing&lt;/a&gt; from the &lt;a href=&quot;http://macgpg.sourceforge.net/&quot;&gt;MacGPG&lt;/a&gt; project.

The real issue is not whether signing comments is easy (it is). It&#039;s whether verifying the signatures on comments is easy. That&#039;s what Srijith&#039;s &lt;a href=&quot;http://www.srijith.net/codes/openpgpcomment/&quot;&gt;MovableType plugin&lt;/a&gt; achieves.

I&#039;ll sign this comment just for giggles. You won&#039;t be able to verify its authenticity until the under-development &lt;a href=&quot;http://bronski.net/data/openpgpcomment/&quot;&gt;WordPress plugin&lt;/a&gt; actually works (and is installed on this blog).

In the meantime, you&#039;ll have to just &lt;em&gt;assume&lt;/em&gt; this was &lt;strong&gt;me&lt;/strong&gt;.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (Darwin)

iD8DBQFAvXKCnyqPIXpYcjcRAlm8AKDA3EjbZZ7B6RArviE7pMcxSiQoWACg+omK
NDyeMrAc5IICFuSyiyE717A=
=m4ds
-----END PGP SIGNATURE-----</description>
		<content:encoded><![CDATA[<p>&#8212;&#8211;BEGIN PGP SIGNED MESSAGE&#8212;&#8211;<br />
Hash: SHA1</p>
<p>Jay,</p>
<p>I, too, use the <a href="http://www.sente.ch/software/GPGMail/English.lproj/GPGMail.html">GPGMail bundle</a> for Mail.app too. I sign all my outgoing email (and have been doing so for almost a decade).</p>
<p>For signing other stuff (like blog comments), try <a href="http://prdownloads.sourceforge.net/macgpg/GPGDropThing-0.4.3.dmg.gz?download">GPGDropThing</a> from the <a href="http://macgpg.sourceforge.net/">MacGPG</a> project.</p>
<p>The real issue is not whether signing comments is easy (it is). It&#8217;s whether verifying the signatures on comments is easy. That&#8217;s what Srijith&#8217;s <a href="http://www.srijith.net/codes/openpgpcomment/">MovableType plugin</a> achieves.</p>
<p>I&#8217;ll sign this comment just for giggles. You won&#8217;t be able to verify its authenticity until the under-development <a href="http://bronski.net/data/openpgpcomment/">WordPress plugin</a> actually works (and is installed on this blog).</p>
<p>In the meantime, you&#8217;ll have to just <em>assume</em> this was <strong>me</strong>.<br />
&#8212;&#8211;BEGIN PGP SIGNATURE&#8212;&#8211;<br />
Version: GnuPG v1.2.4 (Darwin)</p>
<p>iD8DBQFAvXKCnyqPIXpYcjcRAlm8AKDA3EjbZZ7B6RArviE7pMcxSiQoWACg+omK<br />
NDyeMrAc5IICFuSyiyE717A=<br />
=m4ds<br />
&#8212;&#8211;END PGP SIGNATURE&#8212;&#8211;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jay Allen</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4713</link>
		<dc:creator>Jay Allen</dc:creator>
		<pubDate>Wed, 02 Jun 2004 05:32:55 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4713</guid>
		<description>Thanks Matt.  It&#039;s all over now.  The last visit was around May 19th.  I have nothing more to say about that guy.

Anyway, Jacques, PGP-signed comments suffer from the same problem that PGP-signed emails suffer from: poor, little or no implementation. and ease of use.  Sure, the super-techies might be able to sign their own comments, but who can figure out what to do with that and what about the non-techies or even techies like me who could never figure it out?  Thanks to GPGDropMail, and Mail.app, I&#039;m now signing all of my emails, but I&#039;ve been wanting an easy and reliable way to do that for five years and only now have I found it.

Make a dead simple integrated solution and people will use it.  Until then, it&#039;s just too difficult.</description>
		<content:encoded><![CDATA[<p>Thanks Matt.  It&#8217;s all over now.  The last visit was around May 19th.  I have nothing more to say about that guy.</p>
<p>Anyway, Jacques, PGP-signed comments suffer from the same problem that PGP-signed emails suffer from: poor, little or no implementation. and ease of use.  Sure, the super-techies might be able to sign their own comments, but who can figure out what to do with that and what about the non-techies or even techies like me who could never figure it out?  Thanks to GPGDropMail, and Mail.app, I&#8217;m now signing all of my emails, but I&#8217;ve been wanting an easy and reliable way to do that for five years and only now have I found it.</p>
<p>Make a dead simple integrated solution and people will use it.  Until then, it&#8217;s just too difficult.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jacques Distler</title>
		<link>http://ma.tt/2004/06/dont-mess-with-jay/#comment-4712</link>
		<dc:creator>Jacques Distler</dc:creator>
		<pubDate>Wed, 02 Jun 2004 03:12:02 +0000</pubDate>
		<guid isPermaLink="false">http://photomatt.net/2004/06/01/dont-mess-with-jay/#comment-4712</guid>
		<description>&lt;a href=&quot;http://zeldman.com/daily/0504h.shtml&quot;&gt;Zeldman, too&lt;/a&gt; has recently had someone going around leaving blog comments in his name.

Do you still think &lt;a href=&quot;http://golem.ph.utexas.edu/~distler/blog/archives/000321.html#c000719&quot;&gt;PGP-signed comments are overkill&lt;/a&gt;?</description>
		<content:encoded><![CDATA[<p><a href="http://zeldman.com/daily/0504h.shtml">Zeldman, too</a> has recently had someone going around leaving blog comments in his name.</p>
<p>Do you still think <a href="http://golem.ph.utexas.edu/~distler/blog/archives/000321.html#c000719">PGP-signed comments are overkill</a>?</p>
]]></content:encoded>
	</item>
</channel>
</rss>

