Jan 2005

RSS 1.1

I had no idea that people were working on RSS 1.1!

A quick look suggests it’s a cleaned-up version of 1.0. The text of the draft says as much, “RSS 1.1 is hence to be considered a bugfix and streamline release of RSS 1.0 for users of RSS 1.0 who do not want to migrate to Atom.”

It looks good. I like it. My favorite change is removing that items sequence thing that was in RSS 1.0.

But there is a little bad news in here too, unfortunately. I’ll explain:

In RSS 1.0, every <item> had to have an rdf:about attribute that functioned as a unique identifier. NetNewsWire used this as the equivalent of an RSS guid or an Atom id.

Unique identifiers are hugely important to aggregators. Huge. I can’t stress this often enough or strongly enough.

In RSS 1.0 these unique identifiers were mandatory. In RSS 1.1, they are not mandatory. This is a step backwards.

Is there any way that this could be changed?

Update 5:00 p.m.: I heard from Sean Palmer—and it’s not too late to change it. In the next draft of the RSS 1.1 specification the rdf:about attribute will be required. I’m quite pleased.

Feed: protocol

In Really Simple Subscription, FeedDemon developer Nick Bradbury suggests using the feed: protocol to make subscribing to RSS and Atom feeds easier.

I agree.

Nick also says that the solution is not perfect, and I agree with that too. But it does work and it has broad support in existing aggregators.

It’s just not used on the web that much yet. But, if it were used, we’d have really simple subscription today, since it’s already supported in so many aggregators.

(Someone will ask, “Why not use MIME types?” Answer: because aggregators need the URL of a feed, not the content, in order to subscribe.)

Sometimes I think the whole issue is just that we haven’t agreed on a graphic. Nick posted a “Feed” graphic on his post. Last year, Bryan Bell made a few: perhaps one of these would be good?

Blue feed

Purple feed

Orange feed

I think that orange should not be used, to avoid confusion with the orange XML graphic.

Macworld keynote stuff

My thoughts about Macworld keynote stuff...

I’m at home working, not at the show, and I haven’t seen the keynote yet, just read the news. But I have a few thoughts so far:

1. The headless Mac is cool. Period. I want five. I won’t even turn them on, I just want them.

It’s crazy to me how Apple hardware can provoke such desire, but it does.

2. I don’t know if the iPod shuffle is cool. It’s not for me—but it doesn’t need me to buy one in order to sell a few million of them.

Which proves that not every Apple product provokes desire in the heart of every Mac geek. (By the way, I do have an iPod already, if that explains it.)

3. iWork is an odd duck.

I so wanted a great—or even good—word processor and spreadsheet to replace AppleWorks.

I don’t use Microsoft Office, and the open-source office products are, so far at least, still using non-native UI stuff. (They look and feel annoyingly weird. To me. But this will change. Eventually. I hope.)

Dori Smith says that Pages is not a word processor, it’s entry-level page layout software. From what I’ve seen so far I agree. Think PageMaker, not Word.

So: no word processor, no spreadsheet.

But—you know what?—I bet somebody out there makes good word processors and spreadsheets for OS X. It’s just a matter of finding ones I like. (Any recommendations?)

4. Tiger, iLife, etc.—yes, all cool, I’m a happy Mac user.

isbn.nu feeds, or how to find good prices on books, also: I’m mad as a March TiVo

Glenn Fleishman told me about something clever—at his site isbn.nu, you can get feeds for individual books. The feeds show prices, so you can see how much a book is at Amazon, how much at Powell’s, and so on.

Say you want to buy Tom Negrino’s latest book about managing your personal finances with Quicken. Since you’re smart about money (or at least hope to be), you’re looking for a good deal.

Here’s what you could do:

1. Get the ISBN number of Tom’s book. (You could search for it at isbn.nu, look it up on Amazon, etc.)

2. Add a new feed with a URL that looks like this: http://isbn.nu/0321293657.xml

(The number part is the ISBN, which is different for each book.)

When a price changes, the feed is updated.

In completely other news...

daveXtreme says: “Here’s how I’d describe podcasting: it’s like TiVo for radio.”

See? Can’t get away from TiVo comparisons.

My love is like a red, red TiVo. Shall I compare thee to a TiVo’s day? Whether ’tis nobler in the mind to suffer the slings and arrows of outrageous TiVos, or to take arms against a sea of programming, and by opposing TiVo them?

I got TiVo on a cloudy day—when it’s cold outside, I got the month of TiVo.

The value of podcasting

When the buzz about podcasting first started, I wondered what the value of it was. And, even now, some people are still asking that question. I believe I have an answer.

There are obvious ways to criticize podcasting. It’s hard to quote a portion of a podcast. You can’t link to a certain spot in an audio file. Reading text is much faster than listening to it.

It’s kind of like the difference between email and a fax—a fax is a picture of text, while email is text. A podcast is the sound of words.

But all of that is beside the point—the value of podcasting is that people enjoy it. They like the spoken sound of words.

And that’s it. It’s not more mysterious than the appeal of radio. The added bonus with podcasting is that you’re listening to people who don’t have the restrictions of radio. They might even talk about things that interest you.

It’s not a replacement for traditional weblog content. There will always be plenty to read. No worries.

Interview with me

Jeff Harrell interviewed me for his weblog, The Shape of Days.

Email

Email is like TiVo for chat.

I’m not sure why I felt like writing that sentence, except that I’m a big fan of the “blank is like TiVo for blank” comparisons. At some point, everything is either like TiVo or like a thing being TiVo’d.

I sometimes enjoy making up absurd comparisons. Clouds are like TiVo for rain. Soup is like TiVo for vegetables. That kind of thing.

NetNewsWire + MarsEdit lines of code

A little while ago folks were posting the number of lines of code of their apps as generated by SLOCcount. I’m just now getting around to it.

But first—here’s Fraser Speirs, Gus Mueller, and Michael McCracken.

I can’t tell you NetNewsWire and MarsEdit stats separately, since both apps share a number of frameworks that we’ve written. Here’s the stats on both apps together, including the frameworks they share:

Total Physical Source Lines of Code (SLOC) = 86,433
Development Effort Estimate, Person-Years (Person-Months) = 21.60 (259.25)
Schedule Estimate, Years (Months) = 1.72 (20.66)
Estimated Average Number of Developers (Effort/Schedule) = 12.55
Total Estimated Cost to Develop = $ 2,918,470

I think the main virtue of SLOCcount is that it’s an ego gratification tool for developers. “I did the work of twelve point five five people!” Etc.

ATPM reviews MarsEdit

Wes Meltzer reviewed MarsEdit 1.0 for About This Particular Macintosh. It got a “Very Nice” rating—which is nothing to sneeze at for a 1.0. And there’s some great feedback in there we can use for future development.

(Oh, heck, of course we wanted an “Excellent” rating—but, hey, even Delicious Library got a Very Nice rating. Good company to be in.)

One of the things I love about the review is that it talks about Bryan Bell’s work:

In fact, this is one of the application’s greatest strengths. I’m no expert on iconography, though I play one on TV, but all the icons seem immediately recognizable for what they are. That’s a formidable challenge with a weblog editor, since it doesn’t have any real-life analogues, and piggy-backing off of Mail could be confusing. I am especially fond of the way the “… Weblog” command icons look: an action symbol superimposed on a Safari-esque window. And the “… Post” command icons use a sheet of paper with action symbols. It’s a very clean way of saying a lot with a little, and they make an abstract concept much more approachable.

I like this because it recognizes Bryan’s excellent work and also because it’s a mini-tutorial on developing toolbar icons. It’s not enough to make them attractive and clean—the use of metaphors that make sense and are consistent is a huge part of the icon designer’s art.

Though not an icon designer myself, I know just enough to have some small idea how wickedly difficult it can be.

And so not only Bryan but all the great OS X icon designers—Jon Hicks, definitely, and others whose names I don’t know—deserve our fondest thanks for helping make OS X apps usable and fun.

Archive