What's Going On?

So I need to get my own iPhone/iPod touch device for development purposes. I’m not going to shackle myself to AT&T just yet, and with the iPhone refresh imminent (http://buyersguide.macrumors.com/#iPhone), I’d be a fool to buy one right now rather than wait till June.

Fine, so that narrows it’s down to the iPod touch, of which there have been three generations (http://en.wikipedia.org/wiki/IPod_Touch); of course, only the newest third generation is for sale through Apple, but I can always get a refurbished second generation (http://store.apple.com/us/product/FB533LL/A?mco=MTA4MjY3NjQ) or used second generation (http://shop.ebay.com/?_from=R40&_trksid=p3907.m38.l1313&_nkw=ipod+touch&_sacat=See-All-Categories) for a little bit cheaper (but not much, really, less than $50 on average). Given that this is just for me to develop on for the next couple of months, of course I looking to spend as little as possible on this stopgap.

Basically, the second generation is equivalent in terms of performance to the iPhone 3G, while the third generation is equivalent to the iPhone 3GS. Specifically the processor speed, ram, and graphics chips line up almost exactly (from http://en.wikipedia.org/wiki/List_of_iPhone_OS_devices):

1. CPU (iPhone 3G: 412 MHz, Touch 2nd gen: 532 MHz), ram (both: 128 MB), and GPU (both: PowerVR MBX Lite 3D)
2. CPU (both: 600 MHz), ram (both: 256 MB), and graphics chip (both: PowerVR SGX535)

One thing I really like is the idea of having something roughly equivalent to a first gen iPhone, which defines the baseline for performance, which would mean that I should get a refurb or used second gen Touch.

Fortunately, I remembered one other detail which completely changes the picture. Consider the following facts:

1. OpenGL ES 1.1 and OpenGL ES 2.0 are mutually incompatible. That is, apps built against 2.0 CANNOT run on devices that only support 1.1. (http://www.engadget.com/2009/06/10/iphone-3g-s-supports-opengl-es-2-0-but-3g-only-supports-1-1/)
2. The second gen Touch ONLY supports 1.1, while the third gen Touch supports 1.1 and 2.0. (http://toucharcade.com/2009/06/20/iphone-3g-s-vs-ipod-touch-cpu-opengl-es-benchmarks/)
3. “[OpenGL ES 2.0] was also chosen for WebGL, OpenGL for browsers.” (http://en.wikipedia.org/wiki/OpenGL_ES)

In short, if you want 2.0, it has to be the third gen iPhone, and Point #3 above makes me want 2.0. Done, decision made.

Decided to put in some more time on Mozilla Weave 1.0 minimal server after this attempt, but a little less ambitious: getting it working locally on Mac OS X (10.5.8), where I have full admin rights. While the instructions are _technically_ sufficient, there’s a lot more that can be said to make it more straightforward.

By far the most frustrating and poorly documented part are permissions issues. The instructions below are pretty crude and insecure, but it’s just for practice. So here’s what I did:

1. Download the minimal server, and read the README.
2. Move the whole “weave_minimal” directory into your DocumentRoot; for me, it was at “/Library/WebServer/Documents”. Note that you might have to `sudo` to get it in there. It also doesn’t hurt to rename it to something that doesn’t begin with “weave”, to minimize weird path conflicts down the line. E.g.: `sudo mv weave_minimal /Library/WebServer/Documents/minimal-weave-1.0`
3. For good measure, change the ownership and groupship to root and admin, respectively, and give them full permissions: `chown -R root minimal-weave-1.0`, `chgrp -R admin minimal-weave-1.0`, `chmod -R 775 minimal-weave-1.0`
4. Modify your “httpd.conf” (in /etc/apache2/httpd.conf) by adding the following line at the end: “Alias /weave /Library/WebServer/Documents/minimal/index.php”
5. Restart your local Apache instance by going into System Preferences > Sharing > Web Sharing (check to enable)
6. Per the README, hit the URL to initialize the sqlite database: http://localhost/weave/1.0/blah/info/collection
7. You should get a pop-up that axks for authentication info (username and password). You MUST enter “blah” for username, and the SOMETHING (anything) for password (it CANNOT be blank).
8. You still won’t be able to authenticate, but if you look at your minimal-weave-1.0 directory, you should now see a new file called “weave_db”.
9. If you’re this far, congrats, you’ve finished the most annoying parts! Now we create a new user; thankfully somebody in the comments tipped me off to this, otherwise I could have wasted hours. The “create_user” script in the installation is actually a PHP commandline script, and MUST be run this way on the commandline: `/usr/bin/php create_user` (you probably need to sudo).
10. Now you’re done, and can install the Weave client from the project page. Some last notes during client installation: in Firefox Preferences > Privacy, you MUST remember some kind of history, otherwise the Weave preferences tab will be greyed out (this was very confusing, and I hope it’s clarified in the next version of the client).

This is admittedly a very ghetto and unprofessional installation guide (PLEASE PLEASE PLEASE don’t set up your instance like to for _anything_ other than playing around, on a non-production machine), but it was just an experiment to learn what’s _supposed_ to happen during an install. Now I can try and get this thing working on Dreamhost. We’ll see.

The 1.0 release yesterday seemed like a great time to try Weave again. Following the big bold suggestion at the top of https://wiki.mozilla.org/Labs/Weave/Sync/1.0/Setup, I hit up http://tobyelliott.wordpress.com/2009/09/11/weave-minimal-server/ and tried to install my own instance. Alas, Dreamhost bested me. Too tired to fully explain why, but quickly:

1. Dreamhost doesn’t allow “Alias” directives in .htaccess (either that, or you can only have Alias directives in httpd.conf/apache.conf, in a VHost block; whatever, the bottom line is you can’t use it).
2. You somehow need to map some path to a PHP file. That is, something like “host.com/weave/a/b/c” is actually supposed to be a request to “host.com/weave/install/dir/index.php” with a bunch of extra URI stuff after it (in this example, “/a/b/c”). Don’t worry, this sentence doesn’t make sense to me either.
3. Apache rewrite rules are confusing enough, but factor in PHP processing, and it’s more than I can figure out right now.

Damned shame, from the instructions it sounded like it would take 10 minutes, max. <sigh/> Will try again another day. Gods, I hope somebody else figures it out by then.

Doug’s “Super Remove Dead Track v2.1″ works like a charm: http://dougscripts.com/itunes/scripts/ss.php?sp=removedeadsuper

Installation takes about 20 seconds. Note that if your music collection is large (mine clocked in at 120 GB…), and you have a lot of dead links (I had around 10-15 GB), it could take up to an hour.

For the record, I am hurtin’. I have never gotten so excited about a product launch before, and now I know the bitter sting of disappointment.

Now that I’ve had a little bit of time to calm down (and stop <sob/>’ing), I just can’t get away from this: there is really nothing revolutionary about this. It’s literally a bigass iPhone. Let me count the ways:

1. The physical layout is exactly the same: volume controls on the right, power button on the top right, home button on the bottom, headphone jack on the upper left. They really did nothing more than expand it.
2. The home screen is nothing more than a scaled up version of the iPhone home screen. It’s 4 icons by something; I still haven’t checked out the new 3.2 SDK since it’s Snow Leopard only), with a dock/quicklaunch at the bottom (6 versus 4! Ooooo…).
3. No multitasking??? Hopefully it’s just around the corner like the rumors say (like it’s been for like a year now). It _really_ doesn’t make sense, given that the official reason before was battery life; this thing is supposed to clock 10 hours of usage, and a month of standby!
4. Virtual frakking keyboard, only. This is what’s killing me: there’s so much potential to revolutionize data entry with this form factor, and it just feels squandered. How can you use two thumbs to reach across a 9.7 inch diagonal slab??? Putting it in your lap is super-duper lame. Basically, it’s as efficient to actually write something (even an email) with the iPad as it is with the iPhone. <sob>What a waste</sob>.

Other thoughts:

1. The price is pretty freakin’ amazing. On that, Apple certainly did deliver. Kudos for the 3G plans as well; thank the gods there’s no contract madness. Oh, and unlocked, of course. In terms of commercial viability, I think this point alone makes up for all my complaints above, and this thing will sell.
2. No camera? That’s kinda crazy; I’m assuming that it’s going to get one very soon.
3. I wonder if this is something that’s supposed to replace laptops for the average non-technical user (like Mom). But, see point #4 above.
4. I really don’t care about the name; it sounds fine to me. I mean, c’mon, is “iPod” really that good of a name? One of my first impressions of the name was that it sounded like something out of a bad scifi movie: “The iPods are here… to eeeeeaaaattttt yyyooooouuuurrrrrrr BBBBRRRRAAAAAIIIIIIIIINNNNNNNNNSSSSSS!”
5. Arg, I really want to try out the iPad simulator, but then I need to finally clean my machine and get Snow Leopard. @apple Hurry up and refresh the Macbook Pros please!
6. That dock thing looks retarded. Also, who the hell is going to carry around a keyboard??? Wtf????? I really don’t understand this one.
7. 4:3 aspect ratio (1024 x 768)? Huh? I guess it’s easier for reading? That really sucks for watching movies though. By the way, the iPhone is 3:2 (480 x 320), so I’m wondering how auto-scaling for existing iPhone apps is going to work.

Now, the million dollar (well, $499 to $829 question): will I get one?

1. Yes, I probably will.
2. No, I will not get it right away; I just don’t trust Apple’s Rev A stuff (they don’t exactly have a great history), and I’m calling there’s going to be key feature upgrades (camera, please?) between the first and second gen, like there was with the iPhone.
3. Yes, the above might be a damned lie and I’ll just get it within a month or two of it coming out. Reason: point #7.
4. Yes, I can see how this will be the device I spend the most time with. Browsing and watching movies is by far my dominant task, so curling up in bed for hours with this thing seems pretty cool. BUT how do I put my own movies on it? Huh, need to think about this one. Mobile VLC???
5. No, there’s no way this will replace my laptop. But that’s because I need to write stuff and program. _Maybe_ it’s good enough for casual mail and spreadsheets.
6. No, there’s no way this will replace an iPhone. This one’s obvious, right?
7. No, I am not getting one because it’s awesome (thought it seems like if you use it, you fall in love with it pretty quickly), but because I want some of the App Store madness that will probably happen. The gold rush probably won’t be quite as crazy as the iPhone gold rush, but it’s still going to be a hell of a lot of fun (and profitable???) developing for it. The simulator is really no substitute. My experience with iPhone development is that UI and convenience issues that are obscure and abstract in the simulator just scream out at you when it’s actually on the device. And, think of the kinds of games you could make… <drool/>
8. Yes, I will probably get the 3G, even with the $129 increase in price. Seriously, no computer is worth anything if you can’t get online, and this thing is supposed to be mobile, right?

<sigh/> Time for me to get back to the real world and figure out real problems.

This sounds like a fun side project: the Analytics X Competition. It’s basically a competition in the Netflix format, where it’s your job to attempt to predict homocides in Philadelphia. You can use any datasets you can find, and the organizer provides a form to submit predictions and gets scores. There are month winners ($20) and an end-of-the-year winner ($100); obviously, nobody’s in it for the money.

For the hell of it, I created an account, and will work on it randomly when I’m bored. Initial (empty) project page here: http://dev.thmttch.net/analyticsx

This is great stuff, for when you don’t feel like doing anything, but still want to engage your brain.

1. Marc Pachter: The art of the interview (http://www.ted.com/talks/marc_pachter_the_art_of_the_interview.html)
2. Hans Rosling: Let my dataset change your mindset (http://www.ted.com/talks/hans_rosling_at_state.html)

For those of you who don’t know, TED Talks are pretty good stuff. I particularly liked these, which was recommended by this post (http://blog.ted.com/2009/12/10_tedtalks_to.php):

1. Magnus Larsson: Turning sand dunes into architecture (http://video.ted.com/talks/podcast/MagnusLarsson_2009G.mp4)
2. Itay Talgam: Lead like the great conductors (http://video.ted.com/talks/podcast/ItayTalgam_2009G.mp4)
3. Mary Roach: 10 things you didn’t know about orgasm (http://www.ted.com/talks/mary_roach_10_things_you_didn_t_know_about_orgasm.html)
4. Malcolm Gladwell: Malcolm Gladwell on spaghetti sauce (http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce.html)

Steve Holt!