Blogging

It’s not often that I disagree with Ryan Block or John Gruber — these guys are always on top of their game — but I have to disagree when it comes to the recent SIM-Unlock question.

Don’t know it? It was when, during the Q&A portion of the recent Apple announcement regarding the iPhone SDK and the roadmap for the deployment of third-party apps, that Ryan Block asked Steve Jobs:

Will SIM unlock software be considered software not allowed in the app store?

As the live blogging reported, the answer was “(pause)…Yes. (Laughter)”.

Block blogged this later:

But even knowing the answer would be a resounding no, asking whether SIM unlock software would be allowed can be construed as a statement — as well as and an opportunity for Apple to address the millions of potential customers who’d like to be able to use an iPhone on a non-prescribed carrier. (source)

Gruber said this in support:

Just because the answer is obvious doesn’t mean it wasn’t a fair question. I don’t have a problem with Apple serving as a gatekeeper with approval over all apps, but if that’s the role they want, their policies should be explicit. (source)

Here’s the problem, guys: These weren’t the questions asked. What got asked was “will you circulate SIM unlock apps?” and the answer was an obvious “No.” The failure here is that Question 1 was somehow meant to imply Question 2 but only Question 1 can be answered, ergo, it was a waste of time.

Or as Kramer might say, why don’t you just tell me the question you want answered?

Here’s two alternatives:

• “Who is defining which apps are allowable and which are not for the app store?”
• “What is Apple’s policy on third-party applications that compete with AT&T services?”

In film school, one of the first rules of interviewing I ever got was to remove questions that can be answered with a simple Yes or No. Jobs was lobbed an easy “No” — which is exactly how he fielded it — that really no one would argue with, and the purpose of the question was so opaque that we now have two blog entries clarifying it. “Because” is not an acceptable answer, so if the answer must be a reason or an explanation, then you’ve got a winner.

Instead, we got the equivalent of “will you allow software that voids the user’s warranty?” Excellent work there, Professor Layton.

Recently, the use of the Nokia 770 as a simple drawing tablet struck me as a great use of the recently discounted gadget. I would draw up some ideas, and use Tumblr’s image-upload-via-email feature coupled with Sketch’s send-in-email command to publish them. Of course, I ran into a problem — the images didn’t appear with my posts.

After several days of poking around, I tracked down the problem: any PNG file generated from Sketch is given a “.sketch.png” extension. Always. And once a file has this extension, it is assigned a “sketch/png” MIME-type, not “image/png” which would be the default for PNG images, or even “application/x-sketch-png” which would still be unofficial, but at least follow standard MIME-type conventions for application-specific names.

Just to note: MIMEs were designed to be very flexible. Collisions with file extensions happen, where .XLS doesn’t mean Excel Spreadsheet but means eXtensible Library Sequence or some other custom format (although this has improved in recent years now that we’ve moved past DOS and the dot-three file extension). But that’s why MIMEs came to be, because there’s no guarantee that “sketch/png” is even the same kind of data as “image/png”. In fact, based on the naming convention, it’s explicitly denying that it’s an image.

As far as I can tell, they mistakenly assigned these types because they assumed that those final three letters were unique to an image file, and that even if the MIME is unrecognized by other systems, they would naturally just know it’s an image by its name. Basically, Nokia never left the land of DOS, even though they’re rooted in Linux and the device is an “Internet Tablet.”

The worst part is that there’s really no reason for Sketch to do this. Sketch does not save any application specific metadata into its files — they are really just vanilla PNG images. In fact, the only reason this would be done is so the 770 itself would know the difference between a plain PNG and one created by Sketch. That way, if I selected “Open” from the file browser, it would choose Sketch over the standard Image viewer which would only display the image and not let me edit it.

But hold on a second — why wouldn’t I want to draw on existing photos? That sounds like fun! Why wouldn’t I simply want to browse my sketches either? Sketch is only an editor — I can’t freely browse everything in it, plus if I accidentally make a modification to it, I’m always dismissing a Save dialog when I close it. And by its own logic (since you cannot edit dot-extensions on the device) it could easily sort the two apart if needs be using the extra “.sketch” it adds to the file name.

The real problem here is that Nokia chose to create a end-user distinction that wasn’t really necessary in the first place. After all, they didn’t name the app “Photos” (there’s no camera lens/app in the device). Let all images be Images — photos or hand-drawn — and let us decide if we want to start from a blank slate, or add LOLCaptions on our pets or friends.

That way, you don’t have to make up non-standard extensions like “sketch/png” in the first place.

By the way, if you haven’t tried using Tumblr, I highly suggest you do. Not only is the product great, but the support ain’t bad either. :)

I like Radar.net. I really do. But I’m kinda torn over the whole service.

On one hand, Radar has more flexible group control and treats pictures like forums. That’s really cool, because it’s become clear that Flickr is not really optimized for this sort of behavior. Let’s say you post 40 photos from vacation, and you get a comment on several of them. You’ll have to visit a special page or two or use the RSS feed to find out because those photos may no longer be on your main page, and they don’t get marked up in any special way. But with Radar, those photos pop to the top, sort by most recently commented on. Instantly, you know when there’s a conversation happening. That’s what they’re optimizing for, and it’s perfect for mobile picture blogging. You can get a number of different alerts, from IM notifications (making it a bit like Twitter for MMS, as some people have been calling it) to email alerts that can be sent daily, weekly or even every-other-day.

On the other hand, Radar is private. Like, no-peeking private. No one can get in and play around with the site’s capabilities unless they create yet another online account. Certain modes of invitation involves some arcane-feeling “invite codes.” All the activity is hidden, and extra effort is needed to convince people to join up. And since it’s totally private, certain features like RSS go missing. If it’s a mobile picture-blogging experience they’re after, it’s not quite complete, since regular blogging has a very public face to it, and folks who do start private blogs leveraged the publicity of the public users to lure their friends into the habit. There’s really no lead to follow when first browsing the site, and it makes it a harder sell than it needs to be.

I think of Flickr as a community of photographers, not a discussion system. I’m happy when people comment on my photos, but I don’t expect it to be anything more than some feedback about the photo itself. It’s not a message to me, it’s a message for the next viewer. Even today, I will find comments or notes on old photos that I’ve never seen before — It’s just not visible enough or real-time enough. In the end, I use it as an online iPhoto.

I don’t want to take my lovely 10 MP DSLR photos and cram in a bunch of 1 or 2 MP flashless cameraphone shots. They’re just two different things to me, and I desperately want to keep them separate. Radar is almost exactly what I wanted for this, but the problem is, will anyone show up?

(Current score: 2 of 11 invites accepted)

Interesting. Just got a notification that an Uncabled user named “nareman” — who used killerspm@runbox.com as an email address — was created two days ago and just recently posted to very automated-looking spam comments to the blog.

Normally I’d brush this off a little, but the whole thing feels like someone’s testing out a spambot that will handle email account verifications. Perhaps there’s already been one, and this is just my own first encounter.

Whatever the answer is, it helped remind me that I need to go back and check if the CAPTCHA module for Drupal was finally updated for 5.0. (Just checked and it has been. Updating soon.) I hope that the ability to login using credentials from other Drupal-powered sites will make the account requirement on this site a little easier to swallow, but then again, it may just be another attack vector down the road where spammers will set up their own Drupal sites, register their spambots, and auto-login that way.

Sigh. The battle continues….

Derek, who runs the amazing 5thirtyone.com has been talking about IMified for some time now, and I finally decided to give it a whirl. IMified is a chat buddy — something that isn’t immediately clear on the home page — that you add to your IM service of choice, and it basically gives you command line-like access into a number of features. Notes, reminders and to-dos are built in, but blogging and even Twitter posting can be enabled. This post, in fact, was written inside IMified (and explains the lack of line breaks, since I’m still learning the system). It’s an awesome little system, you don’t need a password at all (since that’s all managed through your IM service) and you can get up and running using it in a matter of seconds. Very nice.

Linux users are the OS equivalent of the early Napster or Kazaa users, the kind of guys who would waste hours online downloading illegal songs and risk getting all sorts of malware just to avoid spending 10 bucks on a CD.
— Fake Steve Jobs