<?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"
	>
<channel>
	<title>Comments on: Weekly Tip #2: Streamlining SpamAssassin&#8217;s Learning Process</title>
	<atom:link href="http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/</link>
	<description>Your one-stop place for Apis Networks updates</description>
	<pubDate>Fri, 16 May 2008 07:43:15 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5.1</generator>
		<item>
		<title>By: David Kozinn</title>
		<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-14</link>
		<dc:creator>David Kozinn</dc:creator>
		<pubDate>Tue, 18 Sep 2007 01:54:49 +0000</pubDate>
		<guid isPermaLink="false">http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-14</guid>
		<description>One comment and one question:

I'm using a far more manual method of moving the Spam around to be learned, but it has essentially the same effect. The suggestion that I have would be to at least learn the spam more often. It seems that these things come in waves (e.g., all those football spams), and a week later, it seems like a lot of times they've subsided. As a result, I run a nightly spam cleanup, which runs sa-learn and deletes the folder. My thought is that by doing this, I'll teach SA to ignore those new ones faster.

Now for my question: I started to experiment with "Eudora 8.0" (which is the one built using Thunderbird as a base, with the magic "Penelope" plug it to make it more Eudora-like.) I still don't have a dovecot-keywords file. Is that normally created by Thunderbird?</description>
		<content:encoded><![CDATA[<p>One comment and one question:</p>
<p>I&#8217;m using a far more manual method of moving the Spam around to be learned, but it has essentially the same effect. The suggestion that I have would be to at least learn the spam more often. It seems that these things come in waves (e.g., all those football spams), and a week later, it seems like a lot of times they&#8217;ve subsided. As a result, I run a nightly spam cleanup, which runs sa-learn and deletes the folder. My thought is that by doing this, I&#8217;ll teach SA to ignore those new ones faster.</p>
<p>Now for my question: I started to experiment with &#8220;Eudora 8.0&#8243; (which is the one built using Thunderbird as a base, with the magic &#8220;Penelope&#8221; plug it to make it more Eudora-like.) I still don&#8217;t have a dovecot-keywords file. Is that normally created by Thunderbird?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Matt</title>
		<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-13</link>
		<dc:creator>Matt</dc:creator>
		<pubDate>Mon, 17 Sep 2007 01:05:02 +0000</pubDate>
		<guid isPermaLink="false">http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-13</guid>
		<description>Testing...</description>
		<content:encoded><![CDATA[<p>Testing&#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Kozinn</title>
		<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-12</link>
		<dc:creator>David Kozinn</dc:creator>
		<pubDate>Sun, 16 Sep 2007 22:05:28 +0000</pubDate>
		<guid isPermaLink="false">http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-12</guid>
		<description>Thanks for the research Matt. Unfortunately, I found out why clicking on  doesn't work for me, and might not have worked for you. It seems the  is specific to the Mac version of Eudora, which I don't have.  (The Windows equivalent is  So, I looked for the equivalent setting for the Windows version, and apparently there isn't one. The only thing that I could find was a 2 year old comment in a Eudora forum that said that the functionality was planned for a "future version" of Eudora, which apparently never happened. There are some comments in some fairly old release notes that make it look like the Junk function should actually do something useful for an IMAP server, but I can't find anything in the documentation that explains how to get any of that functionality to actually works. Looks like I might have to look at Thunderbird/Penelope/whatever-it's called if I want to get this to work.</description>
		<content:encoded><![CDATA[<p>Thanks for the research Matt. Unfortunately, I found out why clicking on  doesn&#8217;t work for me, and might not have worked for you. It seems the  is specific to the Mac version of Eudora, which I don&#8217;t have.  (The Windows equivalent is  So, I looked for the equivalent setting for the Windows version, and apparently there isn&#8217;t one. The only thing that I could find was a 2 year old comment in a Eudora forum that said that the functionality was planned for a &#8220;future version&#8221; of Eudora, which apparently never happened. There are some comments in some fairly old release notes that make it look like the Junk function should actually do something useful for an IMAP server, but I can&#8217;t find anything in the documentation that explains how to get any of that functionality to actually works. Looks like I might have to look at Thunderbird/Penelope/whatever-it&#8217;s called if I want to get this to work.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Matt</title>
		<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-8</link>
		<dc:creator>Matt</dc:creator>
		<pubDate>Wed, 12 Sep 2007 15:17:51 +0000</pubDate>
		<guid isPermaLink="false">http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-8</guid>
		<description>After a bit of &lt;a href="http://emperor.tidbits.com/TidBITS/Talk/842/" rel="nofollow"&gt;Googling on the subject&lt;/a&gt;, it looks like label 15 will be stored server-side.  To access the label key mapping, create an empty e-mail and paste &#60;x-eudora-setting:32629&#62; into the message body.  It should be transformed into a hyperlink that you can click.   For reasons unbeknown to me I don't see that behavior.  It may be because I am running Eudora 7 in sponsored mode.  Here's the &lt;a href="http://www.eudora.com/techsupport/mac/download/X-Eudora-Settings.txt" rel="nofollow"&gt;exhaustive list of settings&lt;/a&gt; in Eudora that you can tweak.

I hope that helps you and your Model T-equivalent e-mail client out ;)</description>
		<content:encoded><![CDATA[<p>After a bit of <a href="http://emperor.tidbits.com/TidBITS/Talk/842/" rel="nofollow">Googling on the subject</a>, it looks like label 15 will be stored server-side.  To access the label key mapping, create an empty e-mail and paste &lt;x-eudora-setting:32629&gt; into the message body.  It should be transformed into a hyperlink that you can click.   For reasons unbeknown to me I don&#8217;t see that behavior.  It may be because I am running Eudora 7 in sponsored mode.  Here&#8217;s the <a href="http://www.eudora.com/techsupport/mac/download/X-Eudora-Settings.txt" rel="nofollow">exhaustive list of settings</a> in Eudora that you can tweak.</p>
<p>I hope that helps you and your Model T-equivalent e-mail client out <img src='http://blog.apisnetworks.com/wp-includes/images/smilies/icon_wink.gif' alt=';)' class='wp-smiley' /></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Kozinn</title>
		<link>http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-7</link>
		<dc:creator>David Kozinn</dc:creator>
		<pubDate>Tue, 11 Sep 2007 15:56:54 +0000</pubDate>
		<guid isPermaLink="false">http://blog.apisnetworks.com/2007/09/10/weekly-tip-2-streamlining-spamassassins-learning-process/#comment-7</guid>
		<description>Matt -- These tips are great. The version of Eudora that I'm using (only the "8th life" version as of yet) apparently doesn't pass the tags back to the server, so this method doesn't currently work. What I've been doing is to just move the spams (manually, in the client) to a Spam folder and then run a nightly job to run sa-learn on the spam folder and delete it when done. Since I'm doing the sorting pretty much completely manually, it's pretty much zero risk of learning a false positive, but on the other hand, I do have to do it all manually.

The other difference is that in my case, I don't leave email on the server after I'm done reading, at least not permanently. I often read at the office, do a little processing, then collect things at home, where I keep permanent archives of certain emails. I wind up using both POP and IMAP to manage the emails, mostly because I can't quite do what I want completely with either.  Hopefully Eudora/Penelope will handle things better.</description>
		<content:encoded><![CDATA[<p>Matt &#8212; These tips are great. The version of Eudora that I&#8217;m using (only the &#8220;8th life&#8221; version as of yet) apparently doesn&#8217;t pass the tags back to the server, so this method doesn&#8217;t currently work. What I&#8217;ve been doing is to just move the spams (manually, in the client) to a Spam folder and then run a nightly job to run sa-learn on the spam folder and delete it when done. Since I&#8217;m doing the sorting pretty much completely manually, it&#8217;s pretty much zero risk of learning a false positive, but on the other hand, I do have to do it all manually.</p>
<p>The other difference is that in my case, I don&#8217;t leave email on the server after I&#8217;m done reading, at least not permanently. I often read at the office, do a little processing, then collect things at home, where I keep permanent archives of certain emails. I wind up using both POP and IMAP to manage the emails, mostly because I can&#8217;t quite do what I want completely with either.  Hopefully Eudora/Penelope will handle things better.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
