<?xml version="1.0" encoding="UTF-8"?>
<essay xml:lang="en" xmlns="http://docbook.org/ns/docbook" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:dcterms="http://purl.org/dc/terms/" xmlns:gal="http://norman.walsh.name/rdf/gallery#">
<info>
    
    
    
    
    
    
    
    
    
    
    
<title>A New Low</title><biblioid class="uri">http://norman.walsh.name/2004/07/01/newlow</biblioid>
<volumenum>7</volumenum>
<issuenum>110</issuenum>
<pubdate>2004-07-01T13:23:11-07:00</pubdate>
<date>$Date$</date>
<author>
      <personname>
<firstname>Norman</firstname>
	<surname>Walsh</surname>
</personname>
    </author>
<copyright>
      <year>2004</year>
      <holder>Norman Walsh</holder>
    </copyright>
<abstract>
<para>Spam. Lots of spam. Too much spam.</para>
</abstract>
<dc:coverage rdf:resource="http://norman.walsh.name/knows/where/us-ca-sanfrancisco-moscone"/>
<dc:subject rdf:resource="http://norman.walsh.name/knows/taxonomy#Email"/>
<dc:subject rdf:resource="http://norman.walsh.name/knows/taxonomy#Java"/>
<dc:subject rdf:resource="http://norman.walsh.name/knows/taxonomy#JavaOne04"/>
</info>

<para xml:id="p1">I’m making a flying visit to JavaOne (about 40 hours total time
on the ground), so I’d been offline for about 28 hours when I came
down to the convention center this morning. I’d come down early to
sync my email and practice my presentation for this afternoon.</para>

<para xml:id="p2">I settled down in a comfortable spot and fired up my laptop. The
conference provides wireless bandwidth, but electrical outlets are
pretty scarce so I was running on battery.</para>

<para xml:id="p3">I had 3,285
messages waiting. My mail syncing process uses
<command>fetchmail</command> to download mail from a couple of POP3 and
IMAP servers and runs the messages through <command>procmail</command>
to perform various sorts of spam filtering before delivering them locally.
Later on <command>gnus</command> slurps up
the local mail.</para>

<para xml:id="p4">As I watched <command>fetchmail</command> slowly grinding away,
I did a little mental arithmetic. A couple of hours of useful battery
life divided by three thousand or so messages being processed at a
rate of several tens of messages a minute. Not good. Seriously not good.</para>

<para xml:id="p5">I received so much spam in 28 hours that I wasn’t going to be able
to download all of it <emphasis>over a broadband connection</emphasis> before
my battery life was exhaused.</para>

<para xml:id="p6">Sigh.</para>

<para xml:id="p7">A few things occur to me. One is that I really have to get the
<command>procmail</command> filtering moved onto the server. For a variety
of reasons, this isn’t going to be easy.</para>

<para xml:id="p8">Another is that the throughput I’m getting with <command>fetchmail</command>
is really pretty bad.  (In fairness, this was WiFi to VPN to a proxy
to a POP server in Massachusetts and back (through the proxy, the VPN, and the WifI) to California, so I was a few levels deep. But, still!)
I used to get my mail by running a <command>expect</command>
script that <command>gzip</command> compressed my mail on the server and
downloaded it with <command>scp</command>. I want that kind of performance
back! I don’t understand why the performance of POP and IMAP sucks so badly
and why no one has done something about it.</para>

<para xml:id="p9">My last, most depressing, thought is that email is really
perilously close to collapsing under the weight of spam. Most of the
time, my box runs 24/7<footnote>
      <para xml:id="p10">In the spirit of
<citetitle>Eats, Shoots, and Leaves</citetitle>, can I just interject
here how glaringly illogical I find the advertising slogan “24/7/365”.
Folks, it’s “24/7/52” or “24/365”. Or maybe you really do mean that
your product will only run for a little more than seven
years!</para>
    </footnote> and collects mail every five minutes so I
don’t notice. Going offline for more than a few hours really drives
home just how bad it is.</para>

</essay>

