Time for me to boycott LUGRadio

LUGRadio is now into its fourth series. A Linux-based podcast with a massive listenership and following, it’s grown from a chat between four members of Wolverhampton LUG to probably the...

LUGRadio is now into its fourth series. A Linux-based podcast with a massive listenership and following, it’s grown from a chat between four members of Wolverhampton LUG to probably the most listened-to Linux-based podcast in the world. I didn’t listen to much of the first series because I found the constant swearing, shouting and laddishness hard to stomach. Don’t get me wrong, I’m no prude, I swear far too much and have quite a bit in geek-common with the four guys who make LUGRadio, and I like the guys. I just found the first series hard to listen to. Since then they’ve improved the audio quality, had more discussion and tempered some of the language, soI have subscribed. I also chat in the #lugradio channel on Freenode regularly and have met all four of the guys in person. So this blog post isn’t from some random bloke who has downloaded one episode and then fired off a rant.

I am subscribed to the podcast using the podcatcher of my choice – iPodder on Ubuntu Linux. iPodder is written in Python, is open source, but has been replaced by Juice Receiver. Unfortunately there is no version of Juice for Linux available right now.

LUGRadio isn’t the only podcast I download and listen to. I currently have around 60 in my list of which LUGRadio is one I make sure I listen to every episode of.

So it’s with some sadness and immense frustration that I feel that I need to boycott LUGRadio, no longer downloading or listening to the show until they hopefully address my concerns. My beef is two-fold, one a cause of the other.

• The admins don’t listen to reasoned argument.
• The feed is broken.

For some time it’s been known that the URL used to deliver the MP3/OGG files to podcasting clients (podcatchers) causes problems for some of their listeners. The problem is that in the XML feed the web hosts use the following format for their media URLs:-

http://audio.lugradio.org/foo/lugradio-bar.ogg?podcast

The problem being the “?podcast” at the end. The result of this is that many podcasting clients will either barf completely when they see that URL or will download ok resulting in a file called:-

lugradio-foo.ogg?podcast

A simple test with the “wget” tool can show this.

alan@bishop:~$ wget http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.mp3?podcast
--10:07:07--  http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.mp3?podcast
           => `lugradio-s04e02-250906-high.mp3?podcast'
Resolving audio.lugradio.org... 80.68.94.16
Connecting to audio.lugradio.org[80.68.94.16]:80... connected.
HTTP request sent, awaiting response... 200 Ok
Length: 39,518,336 [audio/mpeg]

100%[==============================================================>] 39,518,336     8.02M/s    ETA 00:00

10:07:14 (7.67 MB/s) - `lugradio-s04e02-250906-high.mp3?podcast' saved [39518336/39518336]
alan@bishop:~$ ls -l lugradio*
-rw-r--r--  1 alan alan 39518336 2006-09-21 14:26 lugradio-s04e02-250906-high.mp3?podcast

This means that your average media player will look at the file and not know what to do with it because most software uses the file extension to figure out what kind of file it is. In my case when I attempt to copy the files to my MP3 player it refuses to copy that file. I have to fix the file after it downloads. I do this every time, every damn time.

This is the _only_ podcast (of 60 odd I subscribe to) that I download that does this. This is also not a new issue. For a long while now this has been the situation. I and others raised this in the past and have had our arguments for change shot down by the developers and web hosts who look after the LUGRadio feed and hosting.

As I understand it their argument is based around the following points:-

1. My podcast client of choice (iPodder) is broken and should be fixed. It should not download the file from the URL above without stripping off the “?podcast” text from the end.

2. The “?podcast” parameter is used by the admins for some reason (possibly statistical analysis) and it would be difficult / annoying / painful for them to remove it now.

3. LUGRadio are just adhering to the standards, and all the podcast clients are not implementing the standards properly.

In answer to those points I’d say:-

1. If iPodder, and I am led to believe, many other clients are broken when downloading LUGRadio, but work fine for every other podcast then surely LUGRadio should bear at least some of the responsibility to do something about this. I know they have called for patches to podcast clients in the past and have sucessfully patched bashpodder. I find it incredibly arrogant however to point the finger at every other application except their own when this issue could be fixed in a trice by the admins. Indeed the LUGRadio guys have themselves bitched about projects such as MythTV who sport an arrogant development team who are unwilling to help their userbase. Pot? Kettle?

2. Are LUGRadio seriously suggesting this is the only way in which stats can be gathered about which method is being used to access the files being downloaded? I put it to LUGRadio that there are numerous ways to identify where people come from to get the files being served, indeed some (such as using vhosts) have been suggested and not implemented. It also appears to have reached stage where the admins are being deliberately stubborn and obtuse to the LUGRadio listenership. Join the #lugradio IRC channel or the web forums and ask why this happens and you’ll immediately get told your podcasting client is broken, fix it. Is this the image LUGRadio really want to convey? I think not.

3. Get real. Standards are great and I can see a well-meaning element behind this point. But it’s being laboured with again an arrogant tone. There are plenty of applications out there that don’t correctly adhere to whatever standards there are. Don’t let LUGRadio be a martyr to the cause of standardisation. Whilst I appreciate that on Linux we are lucky to have access to the source code for many podcasting clients (the one I use included), that isn’t the case for everyone. I have a podcatcher on my Palm PDA which I don’t have the source to. There are mobile phones with podcatchers available – also closed source – for which there is no immediate fix. If it was iTunes that was ‘broken’ and that meant thousands of iPod users couldn’t easily download and listen would the answer still be “your client is broken, fuck off”. I somehow doubt it. I suspect those people would just unsubscribe.

Worth noting that it’s incredibly easy to screw up the LUGRadio stats. The following links go directly to one of the audio files and specifies the ?podcast parameter. Clicking on these, or a spider doing so will cause “GET”s of the files with the parameter even though the RSS feed was not used.

http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.mp3?podcast
http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.ogg?podcast

Also note the stats can be further screwed:-

http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.mp3?FixYourFeed
http://audio.lugradio.org/season4/ep2/lugradio-s04e02-250906-high.ogg?FixYourFeed

Using virtual hosts “audio.lugradio.org” and “podcast.lugradio.org” they could get the stats they require about where people come from who download the show. If they set the content-type and content-name when they deliver the content this may also resolve the problem. Both of which are trivial compared with “fixing” every “broken” podcasting client out there.

I’ve left the IRC channel, and removed the feed from my podcast client until the LUGRadio admin guys see sense and fix their feed.