By John Gruber
Sonar is a new Mac app for GitHub and GitLab issues.
There are a couple of intriguing aspects to Apple’s recent foray into podcasting with iTunes 4.9. First and foremost is the degree to which they’re promoting it. Podcasting is the new feature in iTunes 4.9 — it’s pretty much the entire point of the release. The Podcasts icon scored the second-to-top position in the iTunes source list, and it’s getting a huge amount of play on Apple.com and a top spot in the iTunes Music Store.
[Quick primer, which will seem quaint to many of you: Podcasting is the use of RSS for the aggregation of audio content, generally intended for synching with iPods and other portable music players. The RSS feed for a particular podcast contains a list of (usually) recent episodes, with the associated audio file as an “enclosure”. Unlike email enclosures, however, the enclosures in an RSS feed are not embedded in the message itself; rather, they refer to the enclosures by URL for subsequent downloading. This means you can publish a podcast RSS feed listing 10 recent episodes, each of which may be several megabytes, but the RSS feed itself is just a few kilobytes of XML-formatted text. It is up to your podcasting software to determine when and why it will download the enclosures.]
Podcasting is still very, very new. As in like still dripping-wet new. Compare and contrast with the use of RSS for aggregating regular (written) web content — RSS had been out for a few years and was relatively mature by the time Apple added support for it to Safari. Whereas podcasting really only hit the nerdosphere about a year ago, and yet here is iTunes 4.9, already taking it to the mainstream.
On the whole, I think it’s a sign that Apple can still be rather nimble — this strikes me as about as quickly as something new can be adopted and embraced by a company of Apple’s size.
There are several signs, in fact, that Apple’s podcasting features were rushed to market:
During the first week, there were rampant problems with feeds hosted by Apple at the ITMS. E.g. the RSS feed for The Al Franken Show, which was (and remains) one of the three most-popular subscriptions amongst all iTunes users, was utterly broken. I tried subscribing and re-subscribing repeatedly, each time only to have iTunes complain that the feed contained no enclosed items. No wonder: I looked at the feed’s RSS source and each of the enclosure items looked like this:
<enclosure url="" length="" type=""/>
(The value of the
url attribute is supposed to be, you know, a
URL.) The workaround was simply to subscribe to Franken’s feed
directly from his show’s web site, bypassing the ITMS
entirely. I.e. the original feed, hosted at sundancechannel.com,
worked fine; but the cached feed hosted by Apple at — and I swear
I’m not making this domain name up — ax.phobos.apple.com.edgesuite.net was hosed.
This has since been fixed, but the problem persisted for days. And remember, Franken’s is currently the second-most popular feed listed at the ITMS (first is Apple’s New Music Tuesdays podcast).
You can’t turn off the Podcast item in the iTunes source list. This strikes me as simply an oversight, since you can turn off all the other optional features in the source list, such as Party Shuffle, Radio, and even the Music Store. In the same way the Store panel in iTunes’s preferences window has a “Show iTunes Music Store” checkbox, there ought to be a “Show Podcasts” checkbox in the Podcasts panel.
Even the iPod firmware updates seem rushed. The updates add a new top-level Podcasts menu (which, unlike in iTunes, is optional) and allow iPods to treat podcasts much like they do audiobooks, in that when you resume playing a podcast, it goes back to where you were, rather than starting over at the beginning of the track. But, Apple has confirmed that the firmware updates for clickwheel models breaks an existing feature: the ability for Smart Playlists to update dynamically. E.g. if you have a playlist for all “five-star” songs, it used to update dynamically when you changed the star ratings for songs on your iPod; after installing iPod Updater 2005-06-26, however, Smart Playlists only update when you sync with iTunes. Oops.
The Knowledge Base article concludes:
This document will be updated as more information becomes available.
Which was translated into plain English in MDJ 2005.07.09 as “current Cupertino-speak for ‘We know it’s a bug and it’s not fixed yet.’”
All of which is not to say I’m complaining. Or at least I’m not saying Apple shouldn’t have released its podcasting 1-2-3 punch when it did. And let’s be clear: anyone who deems unacceptable any of the aforementioned problems (or other problems) is implicitly stating that Apple shouldn’t have yet released its podcasting support.
You can’t say that Apple should have released it when they did, but without the bugs and shortcomings, unless you’re of the (mistaken) opinion that Apple’s engineers were just sloppy, lazy, and/or incompetent. Real artists ship, and one of the main tricks to shipping software is knowing when to say, “This is good enough.” Recognizing what qualifies as “good enough” is art, not science. (Although a good QA process with lots of regression testing certainly helps.)
The process of releasing a single major software update is hard enough; but Apple’s podcasting support required simultaneous, coordinated updates to the whole troika: iPod, iTunes, and ITMS.
On the whole, shortcomings notwithstanding, I think Apple’s podcasting support was clearly good enough to ship. Shipping sooner than later, even with more bugs, has instantly made Apple the undisputed leader of podcasting. The whole problem that most other desktop podcasting apps attempt to solve is getting audio files from RSS feeds into iTunes, for synching with iPods. All of a sudden, iTunes 4.9 has turned an entire class of software into an unnecessary middleman. (Podcasting support in dedicated RSS aggregators like NetNewsWire still seems like an important feature to me — many RSS feeds only contain occasional podcast enclosures. iTunes is really only intended for use with RSS feeds in which every item represents a podcast episode.)
And consider poor Odeo, a startup (co-founded by Blogger co-founder Evan Williams) based solely on podcasting. It’s downright amazing that Odeo — the first serious podcasting startup — was beaten to market by Apple. It’s not because Odeo moved slowly; it’s because Apple moved fast.
Apple’s podcasting support is effectively a loss-leader: iTunes is free, and podcast downloads from the ITMS are free. So how does Apple expect to recoup their engineering and hosting costs? Where they’ve been making money hand-over-fist all along: iPod sales. I seriously doubt very many people will be buying iPods just because of podcasting, but every little reason to jump on the iPod bandwagon can help. The trend line on the iPod sales graph is still sloping up, steeply, and the main point to take away from Apple’s serious embrace of podcasting is that they are not about to get cocky with regard to features.
A large part of Apple’s interest in podcasting is that they’ve never shied away from promoting the use of iPods for the consumption of content that you don’t have to pay for. Or in the case of ripping music from your CDs, consuming music that you’ve already paid for. Despite all the hubbub about the imminent 500 millionth download from the ITMS, a big part of the appeal of Apple’s music platform is that they in no way shove the ITMS down your throat. (E.g. as stated previously, you can turn off the Music Store source list item in iTunes.)
Apple has flipped the old Gillette maxim — they’re making money selling the razors (iPods), not the blades. There’s definitely a huge potential upside — big, big bucks — if the ITMS continues growing at the current rate for a few more years. And it’s hard to imagine that anything even remotely resembling any of the current iPods will still be a high-profit-margin product 10 years from now. But at the moment, Apple’s music revenue and profits are coming from multi-hundred-dollar iPods, not 99-cent songs.
The other bit of good fortune is the name: podcasting. Good fortune for Apple, at least. Clearly the “pod” in “podcasting” is about the iPod. Apple couldn’t have come up with a better name for this phenomenon if they’d gotten to choose it themselves. If the whole “audio enclosures via RSS” scene were still known as “audioblogging”, as it was when Maciej Ceglowski recorded his seminal “Audioblogging Manifesto”,1 I seriously wonder whether Apple would have done this now.
If you’re an engineer, you might be tempted to argue that RSS-with-enclosures by any other name is still just RSS-with-enclosures, and that it makes no technical difference whether you call it “podcasting” or “audioblogging” or “noodlepants”.
But names do matter. And what makes this so delicious for Apple is that the more popular “podcasting” becomes as the name for publishing audio via RSS, the less likely it will be that a new name will ever take hold. Which leaves Apple’s competitors — including Microsoft, Sony, and the various other gadget-makers producing Windows Media-based players — in the extremely uncomfortable position of choosing from the following courses of action:
Embracing the word “podcasting”, even though it contains the name of the competitor they’re chasing, and which name subtly implies that podcasting is meant for use with iPods, which implication sort of further implies that every other digital music player is just an iPod knock-off. I mean, can you imagine Apple using a term like “walkmancasting”, “dellcasting”, or “wincasting”? It’s embarrassing.
Devising and using a new term for “podcasting” that doesn’t use “pod”. Good luck with that, considering that everyone — everyone — who is publishing podcasts is already calling them “podcasts”. [Update: According to this story in the Seattle Post-Intelligencer, Microsoft employees are pushing “blogcasting” as a “pod”-free alternative. Good luck with that.]
Ignoring the whole podcasting phenomenon.
There are no other options. The best-case scenario for Apple’s competitors is for this whole podcasting thing to turn out to be nothing more than a fad. That makes #3 a reasonable course of action. But if it isn’t a fad, they’ve got to choose between #1 and #2, both of which are marketing nightmares. And these guys are all already in a deep hole, marketing-wise, versus Apple and iPod.
Hence Apple’s impatience to get their podcasting support out and in use. If podcasting continues to grow and turns into something big, the simple fact that the name includes “pod” is a significant and permanent advantage in Apple’s favor.
One other complaint about Apple’s foray into podcasting is that their entire effort is focused on podcast consumption — finding, subscribing, and listening to podcast episodes — but they’re offering only meager tools to help with podcast production.
They do have a tutorial for creating podcasts with QuickTime Pro, but that pretty much boils down to “click the record button and do your podcast in one continuous take, then save it”. Even if you’re only vaguely serious about production quality, you’re going to need some sort of editing tool.
They also have a tutorial for using GarageBand for podcast recording. This sounds better than using QuickTime Player, since GarageBand allows you to layer separate tracks, splice, and edit. But GarageBand is in many ways overkill, or at least overcomplicated, for recording a podcast. It’s meant for creating your own music, not for recording spoken-word content. (It doesn’t even work with the iSight microphone.)
What’s needed, I think — and I suspect this thought has occurred to various indie Mac developers — is something along the lines of Macromedia’s old SoundEdit / SoundEdit 16 (which was discontinued a few years ago). A sound editor with waveform-based editing, good recording controls, and output options geared toward minimizing file size. The coup de grâce would be a nice human interface for adding chapters and artwork. Apple’s only tool for this now is the command-line Chapter Tool — better than nothing, but a far cry from the sort of creative software tools that Apple and the Mac are known for.
[Update: Several readers recommended Amadeus II, a $30 app from HairerSoft, which seems to provide many of the above features. A few others mentioned Felt Tip Sound Studio, which costs $40. Neither offers anything related to chapter tools, but they do provide waveform-based splicing and fading.]
The question facing indie developers considering making such a tool, of course, is whether Apple itself will be adding podcast recording and editing features to iTunes itself. (I think if Apple does offer a recording tool, it will be part of iTunes, not a separate app — bloat be damned — because they’d want it to work for Windows users, too, because Apple might be tempted to steer podcasters toward publishing AAC files instead of MP3s, because AAC won’t work on Windows Media-based players.)
On the whole, however, I don’t think the current sparsity of podcasting production software is much of a big deal. The vast majority of people interested in podcasting only want to listen to them, not create them. (And, unsurprisingly, most people are mostly interested in podcasts from name-brand professional broadcasters.)
The lowest-level, nerdiest issues that have surfaced regarding Apple’s foray into podcasting have pertained to RSS. To wit:
Various RSS experts have complaints about the spec for Apple’s RSS extensions, e.g. Edd Dumbill. Others, such as Dave Winer, chided Apple for not having asked the RSS community for feedback before releasing the software.
When fetching feeds, iTunes has no support for bandwidth-saving HTTP features like ETags, Last-Modified headers, or gzip compression. Every other major RSS aggregator supports all of these features; iTunes supports none. The result is that every time iTunes checks a feed for updates, it downloads the entire RSS feed, uncompressed, regardless if anything has changed.
When clients support ETags and/or Last-Modified headers, they can essentially ask the server hosting the feed, “Hey, I’d like to download this feed, but only if it has changed since the last time I downloaded it.” It’s a tremendous bandwidth saver, and not a hard feature to implement, programming-wise.
iTunes’s RSS parser is weirdly inconsistent with regard to case sensitivity (it’s case-insensitive in many places, but XML is a case-sensitive technology), date formats, and a few other areas. Sam Ruby and Mark Pilgrim (posting in the comments on Ruby’s site) have documented these issues wonderfully, including specific test-cases.
The idea that Apple should have sought community approval for their RSS extensions is a non-starter. To seek approval implies that if issues were found, or consensus could not be reached quickly (and it seems it wouldn’t have), that Apple would have delayed the release of their podcasting support until those issues were resolved. That wasn’t going to happen; Apple’s podcasting support was going to be released when Apple deemed it ready, and no later. Seeking community approval beforehand but then ignoring it would have been worse than not having sought it in the first place.
The RSS-parsing and standards-adherence issues raised by Ruby and Pilgrim may seem like a bunch of niggles to the untrained eye. Who cares if their XML parser treats “Podcast” the same as “PodCast”? But the accuracy and strictness of Apple’s iTunes RSS parser matters, simply because it is so popular. As Pilgrim wrote:
Apple is an 800-lb. gorilla in this space (at least until Microsoft releases an RSS-enabled IE in Longhorn). iTunes is to podcasting as Internet Explorer is to HTML. RSS interoperability, at least as far as podcasting goes, now means “works with iTunes.” Thousands of people and companies will begin making podcasts that “work with iTunes,” but unintentionally rely on iTunes quirks (e.g. Disney’s incorrect namespace). This in turn will affect every developer who wants to consume RSS feeds, and who will be required to emulate all the quirks of iTunes to remain competitive.
The question shouldn’t be “Why weren’t these issues resolved from the start?”, but instead “Will Apple care enough to resolve them going forward?”
Judging by this report from Tantek Çelik, who met with some friends of his on Apple’s iTunes team, the answer appears to be “yes”:
At that point we started discussing the iTunes podcast extensions (PDF link), and the reactions from the blog community. Blog comments ranged from constructive to whiny, from polite to rude, from objective to snarky. But regardless of the tone that bloggers took, it was clear to Kevin and me that the iTunes team was very much open to feedback and constructive criticism, and subsequent fixes in the podcast extensions specification, and iTunes itself. [...]
The most thorough comments came from Sam Ruby and Mark Pilgrim, e.g. in Sam’s post titled Insensitive iTunes. We went over Sam and Mark’s specific comments and came up with several things that could be fixed in the the spec and/or implementation. Mark’s test cases were particularly helpful.
Kevin and I volunteered to help out with iterations on the spec. Kevin knows a thing or two about RSS and podcasting, and I know a thing or two about spec-writing. We’re going to be providing feedback within the next day or so. [...]
It’s easier to ask forgiveness than to seek permission — and, let’s face it, Apple was not going to seek permission from anyone on something like this. Assuming Çelik is correct about the iTunes team’s willingness to clarify the spec and fix the iTunes RSS parser’s failures, forgiveness ought to come easily. Apple is in a position to do whatever they want regarding podcasting, but for the most part, it appears that what they want to do is the right thing.
Which you should listen to before asking me why I’m not podcasting. ↩︎