œphemærata

Weaselly.

Apple, via The Verge, via Gruber:

We stopped supporting CarrierIQ with iOS 5 in most of our products and will remove it completely in a future software update. With any diagnostic data sent to Apple, customers must actively opt-in to share this information, and if they do, the data is sent in an anonymous and encrypted form and does not include any personal information. We never recorded keystrokes, messages or any other personal information for diagnostic data and have no plans to ever do so.

Gruber:

“Most of our products” is weaselly, though.

There’s other weaselling going on here, though. What constitutes “supporting CarrierIQ”? Who is “we”? One could argue that a CarrierIQ daemon running on an iPhone sending data to CarrierIQ’s servers doesn’t constitute Apple collecting that data. I’d like to know, comprehensively:

  • What data collection software (CarrierIQ or otherwise) was part of each iOS version and device model (and carrier, if that’s a relevant variable).
  • What data was collected, and under what cricumstances.
  • What data was transmitted, and to who, and under what circumstances.
  • What facilities exist for users to opt-out of part or all of that data collection.

If what I though Apple was doing previously remains the case—collecting anonymous usage data and uploading it through iTunes when the device sync—then I don’t have a problem. But I wonder what was being collected and transmitted, six months ago when nobody had iOS 5.

Dozenfold Backups.

This last year has been the most careful I’ve ever been with backups: my home network has a Time Machine volume, and I have a Carbonite subscription. So most things are in three places, and my Simplenote and Dropbox stuff in in more. This week, I’m taking it up several notches, maybe mostly because of listening to John Siracusa.

I’ll make two changes.

First, I’ll put my Notational Velocity notes folder inside my Dropbox folder. 

Second, I’ve procured an external drive to SuperDuper! my iMac’s hard drive to, nightly. 

This means that:

  • For large video files (not backed up by Time Machine or Carbonite), I’ll have two copies, one on my hard drive, and another on my external dopplegänger. There will be up to a 24 hour delay between the file arriving on my HD and being backed up. Many of these I have DVD originals for, too.
  • For my iTunes library (backed up by Time Machine but not Carbonite), I’ll have three copies. Time Machine may lag behind by an hour, the dupe by a day. Technically most of these I have originals for in the form of CDs.
  • For most other files, including my photos and apps, I’ll have four copies (HD, dupe, Time Machine, Carbonite). Note that anything backed up by Time Machine or Carbonite is also versioned. Lag time will vary between an hour and a day.
  • For anything on Dropbox, I could have up to eight copies. I use Dropbox on my iPhone, iPad and sometimes my work laptop. Any given file could thus be on the DropBox server (lag time: seconds), any of four computing devices (lag time: seconds for my home desktop, variable based on launching the DropBox app for the others), and once pushed down to my home hard drive, backed up to Time Machine, Carbonite, and my duplicate drive. 
  • Lastly, anything in SimpleNote (which I use on both iOS devices and, via Notational Velocity, on my home desktop) will exist first on the device I entered it on, then the SimpleNote servers, then it will get pushed to Notational Velocity on my desktop which will write it to my Dropbox folder, which will get backed up to the Dropbox servers. Opening Simplenote or Dropbox on the iOS devices will pull down the updated notes, and my Dropbox folder on my home machine gets the trio of Time Machine, Carbonite, and SuperDuper. So if I write something pithy in SimpleNote on my iPhone, then launch DropBox and SimpleNote everywhere I can, that note will exist in twelve locations, four of them with version control. This means that based on my typical usage patterns for SimpleNote, I will have the most fault-tolerant shopping lists ever! 
(The image is from The Big Picture)
Yay for democracy and all, but what really interests me is the bilingual blackboard in the background. I guess that if you have left-to-right language A (here, English), and right-to-left language B (here, Arabic), you can have a nice bilingual tabulation by putting language A in a column on the left and language B in a column on the right. Native readers of either language will see the tabulation as reading the way they would expect, with an extra column “at the end”. I’d like to see more signs and stuff from the Sudan to see how bilingual conventions have evolved around two languages with different reading directions. 

(The image is from The Big Picture)

Yay for democracy and all, but what really interests me is the bilingual blackboard in the background. I guess that if you have left-to-right language A (here, English), and right-to-left language B (here, Arabic), you can have a nice bilingual tabulation by putting language A in a column on the left and language B in a column on the right. Native readers of either language will see the tabulation as reading the way they would expect, with an extra column “at the end”. I’d like to see more signs and stuff from the Sudan to see how bilingual conventions have evolved around two languages with different reading directions. 

merlin:

“Thorstenson Finlandson” (The Hudsucker Proxy, 1994)

I’ve often felt isolated in my love of The Hudsucker Proxy. Which is a shame, because it’s a fun update on the classic screwball comedy. Great script and fantastic production design. Plus, duh, Paul Newman (“Sure, sure, kid.”). THP was maybe the third thing I ever purchased on VHS.

And, luckily, like the The Big Lebowski, this is a Coener whose image has been burnished over the years.

Mentioned because:

  1. I think this scene’s a riot
  2. I long ago adopted Thorstenson’s name for the occasional fake account-creating identity (thanks, 1password!)
  3. I like that I’ve moved Thorstenson’s office to our local Arby’s.

Dude, this movie was one of the first things I bought on laserdisc, that’s how much I love it. 

michaelhoney:

stewf:

RE-BO by Bootleg Objects
“Beocenter 1400” – designed for Bang&Olufsen in 1973 by Jacob Jensen – embodies the epitome of Scandinavian design at the top of its historical relevance. Formal language, relation of nature and technology, and formalism in interface design all conjure up the spirit of modernism. In the series of the Bootleg Objects, the BO.02 is a representative of the era of the music cassette. However, the cassette slot now houses a smart card reader. Further, a DVD-drive is hidden behind a previously unused groove in the front panel, and a 16:9 TFT display has joined the object on the sly. The legendary slider control formerly used to control the radio tuning now becomes both a display and controller for a whole slew of functions. Consequently, instead of “tuning” the label now reads “anything”.


And it’s a “Non-functional design dummy”, which for twelve thousand euros is a shame. I suppose I like the idea of a genre of sculpture that is essentially mockups of consumer electronics concepts, but the interesting bit of the design isn’t just the external fixtures; it’s how they interact with the working device. Because design isn’t how it looks; it’s how it works. So: the “ANYTHING” slider is cute, but what does it do? How do the designers propose to solve the (interesting) problem of setting keyboard-ish information (which the “P2P” and “WWW” controls imply the need for) using only ’70s-style stereo sliders. Without being really annoying.
This — like the Coke-powered mobile phone that was making the rounds a few months back as a similar “design dummy” — is just an idea. Ideas are cheap and easy. Making good ideas work is the interesting part.

michaelhoney:

stewf:

RE-BO by Bootleg Objects

“Beocenter 1400” – designed for Bang&Olufsen in 1973 by Jacob Jensen – embodies the epitome of Scandinavian design at the top of its historical relevance. Formal language, relation of nature and technology, and formalism in interface design all conjure up the spirit of modernism. In the series of the Bootleg Objects, the BO.02 is a representative of the era of the music cassette. However, the cassette slot now houses a smart card reader. Further, a DVD-drive is hidden behind a previously unused groove in the front panel, and a 16:9 TFT display has joined the object on the sly. The legendary slider control formerly used to control the radio tuning now becomes both a display and controller for a whole slew of functions. Consequently, instead of “tuning” the label now reads “anything”.

And it’s a “Non-functional design dummy”, which for twelve thousand euros is a shame. I suppose I like the idea of a genre of sculpture that is essentially mockups of consumer electronics concepts, but the interesting bit of the design isn’t just the external fixtures; it’s how they interact with the working device. Because design isn’t how it looks; it’s how it works. So: the “ANYTHING” slider is cute, but what does it do? How do the designers propose to solve the (interesting) problem of setting keyboard-ish information (which the “P2P” and “WWW” controls imply the need for) using only ’70s-style stereo sliders. Without being really annoying.

This — like the Coke-powered mobile phone that was making the rounds a few months back as a similar “design dummy” — is just an idea. Ideas are cheap and easy. Making good ideas work is the interesting part.

Artists make me stabby.

This: http://www.petermiller.info/polaroid.html is a cool piece of art. Neat idea, neat art. Well done, Peter Miller.

But let’s look at the penultimate sentence of Miller’s description of the piece:

By separating one-step photography into two irreducible steps of photochemical photography: exposure and chemical process, I disrupt the authority and authenticity inherent to Polaroid imaging. 

This is the kind of self-important crap that gives art a bad name. Previous to reading this, I was totally unaware that there was any “authority and authenticity” to be found in instant photos. Immediacy, maybe — I can buy there being immediacy, but I don’t see how an instant photo (Polaroid or otherwise) is inherently more authentic or authoritative than any other unretouched film photography. And how is what the author did disruptive? How is his process less authoritative, less authentic? He’s elaborately staged a photo, but he could have constructed an array of 90 instant cameras and had them fire off simultaneously to achieve the same effect, without his vaunted two-step reduction.  

It’s the last sentence where he goes totally off into la-la land:

I am able to disorient an audience’s preconceived understanding of how the world around them works.

Are you, now? I’m willing to bet that essentially nobody is disoriented by this piece. Certainly my worldview isn’t anchored by preconceptions about how instant photography works. 

Here’s the thing: Art doesn’t have to be profound to be good art. It doesn’t have to “say” something or “challenge” something or “disorient” or “confront” or “mean” anything. It can just be good art — something that looks interesting and is executed skillfully. These are some of my favorite pieces of art:

Do they say anything? Do they make assertions or challenge anything? I don’t think so. Within the context of the art word of times they were created, perhaps they did. And that’s interesting and valuable, but ultimately I don’t think it’s necessarily inherent to the art. Miller’s piece is interesting, and nicely done. And I think that’s all art needs to be; dressing it up with grandiose statements about how it challenges preconceptions or disrupts paradigms only cheapens the work itself. It seems symptomatic of an insular subculture where artists aren’t so much concerned with the work they do but with how it is perceived; a subculture that’s essentially ignorant of how the world at large will appreciate and regard their art. 

Here’s how I think the best art happens. An artist — of any kind — thinks of an project that interests them. Then they work as hard as they possibly can to make it as good (however they understand “good” in the context of their own work) as they possibly can. And then they present it without commentary, letting it speak for itself. I think that’s true for writing, music, filmmaking, painting, sculpture, furniture, photography, poetry … anything we call art.

Hackerdom, tarnished.

Around about 1991, when I was finishing university and wondering what to do with my life, I spent a lot of time reading the New Hacker’s Dictionary (the book incarnation of the Jargon File), feeling at once both inspired (“There’s a whole subculture of people that I identify with”) and mournful (“I don’t know any of them! And I suspect I wouldn’t fit in even if I did!”). But the basic sensibilities of the book, particularly when it comes to unixish technologies, have stuck with me.

Just now, I was Googling for a better way to turn off Emacs’ default backup behavior when using Git (or, as it turns out, in general). I found it here, in the generally helpful emacs wiki. But was struck by several classic hacker affectations that, I’m sure, I once would have found endearing, but now I just found trite and annoying. To wit:

"Myanmar, Independence Day" At the top of the page (when I visited the site, January 4 2009), underneath the nav stripe. Presumably there’s a different message for each day. But why? First, I don’t need the emacs wiki to tell me interesting factoids about today’s date: that’s just noise. (It’s also noise to spiders indexing the site, so it’s actually a functional defect). Second, Myanmar isn’t very independent, what with being under the control of a pretty brutal and oppressive regime. Third, even though this bit of fluff is date-related, it’s not presented as such — unless you twig to what’s being done, it seems totally incongruous. I know what’s happening — someone decided it would be fun to have their templating system pull an interesting date-factoid for each day. Whee. But that kind of stunt — like .fortune files and random email signatures stopped being amusing and started becoming an attention drain. If I want to know interesting shit about what happened on January 4, I can look that up. This is just noise, and it’s noise of a “Look at the exciting creative things we hackers love to do” flavor. Editing, folks: that’s where creativity gets real. Editing.

The smug tone. “This is primitive and boring”. “Civilized people…” I’m just tired of it; tired of people making disparaging assertions about other folks over things that are, really, really pointless. I’ve let emacs do it’s tilde-appending thing to me for close to twenty years; apparently that means I’m not civilized. Again: editing. You want to be informative? It’s had to do that while being judgmental.

The affected quirkiness. “…out of the box in Emacsen…” I had to think about this before I realized that “Emacsen” wasn’t the name of some variant of emacs; it was just the aren’t-we-wacky irregular plural of “Emacs”. Worse than that, it’s grammatically singular, because it should agree in number with the the later pronoun in the sentence “..in Emacsen with its…”. So not only is it obscure for no reason, it’s wrong!. “Emacs installations” is better.

The total plaintextness. Folks, an hour spend building a CSS file for the site would go a long way. You don’t have to change any of your HTML. It’s not hard, and graphic design is a part of interface design, and interface is part of software. Make your software better and lose a bit of the “plain text is best” attitude.

All of these things seem to be part of (I now realize) an attitude of smug, judgmental quirkiness that seems very dated. For me, at least, some of the trappings of hacker culture have lost their shine as I got older.

(also, what’s up with the “Cette page est …” at the top of the page? What kind of hacker builds localization like this — with manually inserted links? Geez)

The Urge To Rewrite

Conventional wisdom is — in the software engineering world — that if you have an application in production, and you get a wild desire to rewrite it from scratch, then you shouldn’t do it. Never, ever ever. You should refactor to improve the application incrementally. Ground-up rewrites are never worth it — they inevitably take much longer than you originally estimate; they prevent you from maintaining your application that’s in production; and they introduce piles of new bugs (or, even more likely, reintroduce old ones).

But.

I wondering if there’s a point where, however profound the practical arguments against a rewrite are, there’s a psychological upside. You get to a point where you’re so sick of the old codebase that it takes you twice (or ten times) as long to do anything with it, just because the disgusting cruftiness of it pisses you off so much. But a rewrite would be all shiny and new and let you do bold, heroic things and be vastly more productive. You’re keen and excited agin because of the spiffy new architecture you’ve invented. Or whatever. 

So: could it be that, while a rewrite is a horrible choice from an operational standpoint, from a motivational one it can be excellent. Especially for one-coder shops.

I ponder this because it’s more interesting than the changes to the crufty in-production application I have to make…

(I maintain a fairly hacktackular mod_perl/Mason application and I’m really, seriously thinking of rewriting the whole thing in Rails to keep my interest levels up. Mmmm, Rails.)