hpodder has difficulties with ISO-8859-1 encoded feeds.

To be accurate, not every feeds seems to be affected. I experienced problems with:
* http://gffstream-6.vo.llnwd.net/e1/podcast/bruno-jonas-rosstaeuscher/cast.xml
which is encoded as ISO-8859-1
* http://download.br-online.de/podcast/iq/cast.xml
which is encoded as ISO-8859-1 as well

While the following feed which contains Spanish accented letters (eg á) is NOT affected:
* http://feeds.feedburner.com/vocespodcast?format=xml
which is encoded as UTF-8

I suppose it's the non UTF-8 encoding which makes it hard for hpodder

The Australian Broadcasting Commission Radio National podcasts also use ISO-8859-1 (Latin-1) encoding, so hpodder is unable to handle their podcasts (http://www.abc.net.au/rn/podcast/feeds/)

From ~/.hpodder/feedxfer/*

<?xml version="1.0" encoding="ISO-8859-1"?>
<rss xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:media="http://search.yahoo.com/mrss/" xmlns:creativeCommons="http://backend.userland.com/creativeCommonsRssModule" version="2.0">
  <channel>
    <title><![CDATA[In Conversation]]></title>
    <description><![CDATA[In Conversation features leading science writer and broadcaster Robyn Williams talking one-on-one with science professionals and other leading figures who have an interest in the field. The program<B4>s wide-ranging interviews explore a diverse array of scientific topics that engage Robyn<B4>s local and international guests. In Conversation is published every Thursday.]]></description>
    <link>http://abc.net.au/rn/inconversation/</link>
    <copyright>Australian Broadcasting Corporation</copyright>
    <language>en</language>

Package: hpodder
Architecture: amd64
Version: 1.1.5.0+nmu2
Depends: libc6 (>= 2.4), libffi5 (>= 3.0.4), libgmp3c2, libsqlite3-0 (>= 3.6.23.1), curl, id3v2

$ hpodder update 2
1 podcast(s) to consider

Get: 2 In Conversation
0% [2 0 B/1 B 0%] 0 B/s 0shpodder: /home/rdb/.hpodder/feedxfer/ac58b865317f2be3ba150aea767682ca: hGetContents: invalid argument (Invalid or incomplete multibyte or wide character)
100% 0 B/s 0shpodder: thread blocked indefinitely in an MVar operation

So ISO-8859-1 encoded podcast XML that contains characters with the 8-th bit set will fail, because the code is attempting decode them as UTF8, rather than using the ISO-8859-1 encoding. This trips an assertion and the podcast downloads fail.

In the above report <B4> is (from man latin1)

       Oct Dec Hex Char Description
       ────────────────────────────────────────────────────────────────────
      264 180 B4 ´ ACUTE ACCENT

which is overly cute in the ABC data.

Status changed to 'Confirmed' because the bug affects multiple users.

Confirmed with Russian names as well.

Still an issue here.