<?xml version="1.0"?>
<rss version="2.0">
  <channel>
    <title>Advogato blog for hovinen</title>
    <link>http://www.advogato.org/person/hovinen/</link>
    <description>Advogato blog for hovinen</description>
    <language>en-us</language>
    <generator>mod_virgule</generator>
    <pubDate>Fri, 5 Sep 2008 17:23:31 GMT</pubDate>
    <item>
      <pubDate>Sun, 30 Jan 2000 02:49:32 GMT</pubDate>
      <title>30 Jan 2000</title>
      <link>http://www.advogato.org/person/hovinen/diary.html?start=0</link>
      <guid>http://www.advogato.org/person/hovinen/diary.html?start=0</guid>
      <description>I just did some work on GDict today, mainly improving my
reimplementation of the protocol implementation (since MIT's
dictionary server appears to be gone, I threw out the old
protocol implementation and rewrote it to use the one given
at www.dict.org). My new implementation has some pretty nice
features: non-blocking operation, automatic creation and
destruction of parallel connections, and so forth. Getting
the UI to the point where it really takes advantage of this
protocol is an entirely different story. I am a tad
concerned that, while MIT's protocol supported cross
references (which basically make the dictionary act doubly
as a thesaurus), RFC 2229 does not. It would not be
difficult for a server to implement that functionality, by
adding an additional strategy for the MATCH command, but
that support would be nonstandard at best. Spell checking
also presents a problem in that regard. Hopefully I'll get
around to fixing up the UI code some time soon.

&lt;p&gt; It's looking pretty good, though. </description>
    </item>
  </channel>
</rss>
