this is sippey.typepad.com

make a beat, eat the beat

How great would it be to play with the Skittle-based beat sequencer with the kids? How better to teach them about the near infinite glories of what you can do with sixteenth notes and 4/4 time than with a bunch of colorful candies. Make a beat, eat the beat. Yum. (Via Waxy, of course.)

(But you know, come to think of it, a simple sequencer like this would be a great iPhone app, even if you can't eat the Skittles after making your beat. Instead, you could share them online. iPhone as drum machine. See, I knew I had to have an iPhone 2.0 reference today somewhere...)

my head just exploded a little

Check out the video of processing.js running inside a TiddlyWiki.[1] (Learn more and get your own.)

[1] I've always hated the term "TiddlyWiki" because it's so twee it hurts. No, I don't have a better alternative, and this post really isn't about the tiddly-ness, it's more about the processing-ness.

i miss my keyboard shortcuts

There's very little motivation for Microsoft to do this, but oh how much would I love to have all the keyboard shortcuts for menu and dialog navigation be available in Excel 2008 for the Mac that were available in Office for Windows? Being forced to mouse around in Excel to do simple things like insert a row is absolutely infuriating, when in pre-switch times it would be a split second alt-I-R away. Not having these available is turning my trust swiss army knife of business computing into a dull, oversized butter knife.

(Oh, and have I told you lately that when I start a band it will be called Alabaster Whine? Wanna join? I'm looking for a bass player.)

a quick note to the kindle team

Dear Kindle team,

I just finished a week of travel with a new Kindle, and I'm well into reading my third book on it. So far so good -- the screen is very readable, even for long stretches of time; I don't have a problem with the next page / prev page button placement; the little e-ink flash when you turn pages isn't that annoying; and the battery life is great.

The UI for finding and buying books is pretty good, and it will be interesting to watch how you use the limited screen real estate and UI controls as a merchandising advantage as more and more titles come online. I can't imagine using the device for browsing the web or reading email -- the screen refresh is fast enough for reading books, but not for skimming web content.  And while I wish the keyboard were virtual (I only use it when searching the store), I get why it's there and appreciate the tradeoff decision re. price & touchscreen.

A key feature of the device is a simple and elegant way to highlight and clip sections of text -- and once you get the hang of it you find yourself doing it all the time. When reading real books I'm always marking pages or making notes in the margin; the Kindle makes this a quick click and scroll action. But I was surprised to learn that these highlights stay on the device, and don't sync up to your Amazon account. (You can download them with the included USB cable, but who wants to do that?)

So here's the feature request that must be on your roadmap: send my clippings back up into the cloud, where I can copy and paste them for future use. Bonus points for giving me the opportunity to connect my Kindle account to my blog, and have the service automatically post new clips via Atom or MetaWeblog. Extra bonus points for illustrating those with a cover thumbnail, and embedding my Associates code in the URL back to the store.

Best,
Michael

thoughts on the social graph api

Congrats to Brad Fitzpatrick on the unveiling of Google's Social Graph API.  I love that the work he started at Six Apart with OpenID and the open social graph continues on at GOOG.  Brad's video is a great introduction to the API; he is a master of the whiteboard (I've seen it in action many many times), and I love the fact that the first whiteboard scene looks like it's been shot in front of a whiteboard that's eight stories tall.  (Oh, and the line where he says "once we throw away all the links that aren't marked up like this" is perfect.)

Clearly from Google's perspective it's obvious and elegant to have the Googlebot reporting on the graph; if it's public data it's being indexed by Google already, and it encourages the use of simple markup like XFN and FOAF to declare social relationships. But I wouldn't be surprised to see some sort of privacy blowback on this when the API starts to get baked into consumer services: "Wait -- Google knows who my friends are, too? How do they know this?" It's going to be incumbent upon social services that are (a) enabling individuals to connect with one another and (b) expressing that connection in an indexable way to make it clear that those relationships are public (if they are) and demonstrate the benefits of those relationships being public (and the risks of exposing relationships you don't want public).  I'm not talking about language in a privacy policy, I'm talking about features that create a better user experience for the data being available.

A short comparative story.  About a year ago there was a pretty significant shift in how normal everyday people thought about blogs.  (I'm not talking about what people in the echo chamber thought about blogs -- I'm talking about people who could care less about the people and service you know and love.)  Perception shifted from "why would I want to put that much information about myself out there" to "I need to do this to own my results page on Google." This wasn't about blogging for SEO, but rather about having control over what information is available about you online. Blogs are an easy way for people to do that, and it's driving the next big wave of adoption.

Here's the point:  connections being discoverable (like they are today -- check out these results for sippey.com) will drive the next wave of adoption of connection creation tools, just like content being discoverable is driving the current wave of adoption of content creation tools.

Three thought experiments, just for "fun."

• Short term: This weekend Six Apart's David Recordon is hosting the social graph FOO Camp; what are the odds that a hack comes out of there to explore relationship strength and/or missing relationships based on this API.  (Hey, Steve -- you're friends with Jerry on these three services, why not on this one?)
• Medium term: How long until this is connected explicitly into my Google profile. It would be trivial for them to ask me to claim sippey.com as my own using OpenID. Search, advertising and all of their social stuff could be customized based on this information.
• Long term: social relationships as an influencer of credit scores?

gmail, iphone and ajax

This is pretty much the definition of a first world problem, but I know I'm not the only one who thinks the new Gmail UI for the iPhone isn't all its cracked up to be. They moved basic transactions like reading and archiving messages into asynchronous calls, and it creates this incredibly disjointed and sluggish user experience. Case in point, archiving a message.

• Steps to reproduce: view a message, touch the archive button.
• Expected result: user is returned to the message list, with that message removed from the list.
• Actual result: user is returned to the message list, with that message still in the message list. After several seconds a banner message appears stating that "The conversation has been marked as read." And then, after several more seconds (longer depending on the speed of your connection, on EDGE I've seen this take at least 10 seconds), another banner message apperas stating "The conversation has been archived."

Look, I'm a big fan of asynchronous JavaScript and XML. Just like I'm a big fan of HTML and CSS. And heck, HTTP for that matter. But the team made the technology they used the lede of the story, which leads me to believe that the requirement wasn't "make Gmail faster for iPhone users, especially on EDGE," but something like "port UI to AJAX."

On the plus side, IMAP setup is now much easier, so maybe I'll switch to that.

software design as performance art

Kai Krause answers John Brockman's question What Have You Changed Your Mind About with an interesting take on the nature of software design.

I used to think "Software Design" is an art form.

I now believe that I was half-right: it is indeed an art, but it has a rather short half-life: Software is merely a performance art!

A momentary flash of brilliance, doomed to be overtaken by the next wave, or maybe even by its own sequel. Eaten alive by its successors. And time...

This is not to denigrate the genre of performance art: anamorphic sidewalk chalk drawings, Goldsworthy pebble piles or Norwegian carved-ice-hotels are admirable feats of human ingenuity, but they all share that ephemeral time limit: the first rain, wind or heat will dissolve the beauty, and the artist must be well aware of its fleeting glory.

(That link there? To the Goldsworthy images? That's what they call "added value.")

sharing is caring

Paul Kedrosky: "Hell, to adapt a cliche, hath no fury like that of a software user whose software newly works the way the documentation says it will."

the missing piece

I'm very interested in today's announcement re. Amazon's SimpleDB. This is, of course, the missing piece in the Amazon cloud services story -- storage, processing power, and now a hosted DB service. It's simple: there are domains, in domains there are items, and each item can have 256 attributes, and each attribute can be 1,024 bytes in length. Automatic indexing, no up front schema definition. I'm sure there are tons of startups that have been waiting for this, because now "Developers can run their applications in Amazon EC2 and store their data objects in Amazon S3. Amazon SimpleDB can then be used to query the object metadata from within the application in Amazon EC2 and return pointers to the objects stored in Amazon S3."

Some less than educated questions that come to mind:

• Is this Dynamo, or Dynamo backed?
• How long until there are libraries for abstracting SimpleDB in PHP, Perl, Ruby, Python, etc.?
• Could you / would you need to run a memcached service in front of SimpleDB in an EC2 instance? I can see wanting to do that in order to reduce the cost of i/o and processing time on SimpleDB (it's more expensive there than in EC2).
• SimpleDB is obviously very far away from traditional RDBMS' like, say, MySQL. If developers build apps with SimpleDB on the backend, there's potentially some interesting lock-in risk for them (opportunity on the Amazon side) if someone doesn't develop a SimpleDB equivalent that you can run on your own boxen.
• As a part-time data geek, I'm already imagining some very interesting ways I could use this to analyze large data sets that are basically a massive set of name / value pairs.

In sum, [this is good].

the note could not be saved

Small Macintosh OS X Leopard hint, first in a series of one, because I usually don't do this kind of thing. If you have a local mail folder named "notes" you need to rename it before Mail will let you save a new note you create.  Otherwise you will get an error message that reads "The note could not be saved." May legions of Leopard users find this blog post through the wonders of Google and save themselves just a little bit of time.