Marco.org

I’m : a programmer, writer, podcaster, geek, and coffee enthusiast.

Going to WWDC? Want an iCal export of your Favorites from Apple’s Sessions & Labs page?

I made WWDCal tonight to do exactly that.

Let me know if it works (or doesn’t) for you: instapaper@marco.org. Thanks.

Microsoft didn’t lead the way to the iPad

John Gruber ended a recent Linked List post with this comment:

Microsoft led the way to the iPad, they just happen to have nothing on the market or even on the horizon that competes with it.

(The first part was probably sarcasm, but let’s go along with it, because Paul Thurrott seems to truly believe it.)

He was responding to Paul Thurrott’s iPad article, which included:

And of course most Apple people don’t even know that Microsoft and its partners had been innovating in this market for a decade already anyway.

By this, he’s referring to Microsoft’s Tablet PC hardware and software, UMPCs, and Windows netbooks.

In this case, Thurrott is giving Microsoft far too much credit.

Microsoft’s efforts have always wedged their PC operating system and PC-class hardware into tablet-sized products.

The iPad is the result of expanding a smartphone OS and smartphone hardware into a tablet-sized product.

The PC-to-tablet approach creates devices that are feature-packed but bulky and clunky. PC hardware needs to be downsized, sacrificing battery life, performance, thinness, lightness, simplicity, and quality. PC software needs to be downsized, sacrificing usability and practicality, in an effort to fit itself into the relatively constrained screen space. The resulting products are versatile enough to do more computing tasks than smartphones, but tend not to excel at much, if anything.

The smartphone-to-tablet approach creates devices that are feature-light but relatively ultra-portable and highly usable. Smartphone hardware can be upsized, achieving massive battery life with small, low-power, inexpensive hardware. Smartphone software needs to be upsized, which might make inefficient use of the relatively large screen space at worst, but never feels cramped. The resulting products are less versatile, but tend to excel at more of the smaller number of tasks that they perform.

The two approaches share very little hardware innovation and almost no software innovation.

Everything Microsoft has ever released in these device classes has used the PC-to-tablet approach. They launch new tablet or ultraportable products every five years with different expectations, yet the market responds the same lukewarm way each time, because each product is the same set of tradeoffs with a different sticker on the front.

Microsoft didn’t lead the way to the smartphone-to-tablet approach, and they didn’t lead the way to the iPad: they led the way down their own path that got them somewhere completely different, irrelevant, and unsuccessful.

My WWDC tips, after only having gone once, especially for iPhone developers

These are pretty limited in scope and wisdom compared to tips from the veterans. But here’s what I learned from going to WWDC last year, and going to the Tech Talk in New York last winter (which was like a condensed WWDC, and which I’d strongly recommend if you get the chance).

Bring a workable laptop. By that, I mean a laptop on which you can do development work, because WWDC sessions will inspire you to try new things immediately. If you’re choosing between the thin-and-light or the workhorse, pick the workhorse. If you’re considering going iPad- or iPhone-only, I strongly advise against it.

Sync your iPhone to that laptop. If you can only sync to your Mac Pro back home in New York, you’ll be pretty disappointed that you can’t install the brand new OS seed announced in the keynote until a week later. And on the off chance that the new iPhone is available immediately there, you’ll want to be able to buy one and set it up.

Bring an extra iPhone battery. You’ll use your iPhone a lot. You can plug your laptop into power strips strapped to the chair legs during most of the sessions (although a unibattery MBP’s long life is very handy), but you won’t have many chances to recharge your iPhone during the day. (Another tip: Sit near the chair legs with the power strips strapped to them.)

Go alone. Non-developer spouses will be bored senseless. Bring them to SXSW instead, where the sessions are an afterthought and most people skip most of them. But at WWDC, attendees actually attend. (See next tip.) Unless your spouse is another developer with a WWDC ticket, take this trip alone, for his/her sake. (The same applies to non-developer friends and coworkers.)

See as many sessions as possible. This isn’t SXSW. Most sessions are extremely valuable, especially because a lot of information is revealed, often in casual remarks by the engineers, that isn’t available anywhere else. Videos of each session are available after the conference, but they usually only include the slideshow and audio, and critically, not the often-invaluable Q&A at the end of each session. If it’s important to you, go to the session in person.

Don’t skip any timeslots. I don’t care how tired or hung over you are — this is why you’re there. You can sleep later.

Take notes. I never took notes on anything in school. I was that smartass who would just stare at the board with nothing on my desk and remember the important stuff. But trust me: take notes here. Make a “WWDC” folder full of text files. Start a new file for each session, and keep that text-editor window (TextMate for me) as the only app running during the sessions. Take notes liberally. You don’t need to copy down the code samples — they’ll be posted somewhere later — but you may want to note the key method names. The importance and volume of note-taking is why you shouldn’t go iPhone- or iPad-only.

When you don’t have any must-see sessions in a timeslot, go to one that you’re only vaguely interested in. You’ll learn a lot, and they’re often a gateway to new things that you didn’t know you were interested in. App developer? Go to a game session. Or vice versa. Know nothing about Core Data, internationalization, accessibility, or Instruments? Now’s your chance.

Go to sessions about topics that you think you’ve already mastered. You haven’t. I still go to sessions on UITableView, UIViewController, UIWebView, etc., despite thinking I know these inside and out, because I learn something new each time — about how to use something, or how something behaves behind the scenes, or why to use something — that greatly benefits my app and can make the price of admission worthwhile alone.

Go to the UI-design sessions. These, by far, are always the ones whose notes I consult the most afterward, and that I come out of thinking, “WWDC is worth every penny.”

Find good coffee. (Intended for black-coffee drinkers. If you put crap in your coffee, you’ll probably like it from anywhere.) The Blue Bottle Cafe, 66 Mint St., is best. But because it’s a gourmet, slow-process boutique, it takes more time than you probably have before the morning session. The best fast option is the Peet’s at 370 4th St., which is a small step above Starbucks in black-coffee quality. The brownish liquid substance in those giant metal urns in Moscone is not coffee. Do not attempt to drink it.

Bring some questions for the Labs. Last year, I had just finished a major release before WWDC, so I couldn’t think of anything to ask the Labs people except one question that they couldn’t answer. (I later found that the answer was to use “Delays content touches” on my UITableView.)

Promote your stuff with your wardrobe. It’s OK to wear your own T-shirt. (I wished I had one, but I forgot to get one made in time. And I forgot again this year.) But it’s not usually a business-card-passing setting, like those awful meetings with Business People™ where everyone stands up first and passes their cards around the table. It’s not like that at all. (I consider this a feature, not a bug.)

Most importantly, absorb as much as you can, meet as many fellow developers as you can, and have fun.

Need an iCal export of your Favorite sessions from Apple’s scheduling site? I made this.

That experience taught me a lot about what really matters in programming. It is not about solving puzzles and being the brightest kid in the class. It is about realizing that the complexity of software dwarfs even the most brilliant human; that cleverness cannot win. The only weapons we have are simplicity and convention.

Jonathan Edwards

Preview.fm: An experiment for fast browsing of full albums.

I buy full albums, not singles. I listen to the complete albums, and I don’t use shuffle. My iTunes is sorted by “Album by Year”. I like albums.

But as this increasingly becomes a minority opinion, music storefronts like iTunes and Amazon MP3 are encouraged to build their interfaces and priorities around hit singles. As a result, whenever I discover a new band and browse their albums to decide which to buy, the storefront interfaces often work against me, making it difficult to quickly find a band’s albums and navigate between a bunch of them for preview and comparison.

So I made this.

Pros:

Cons:

Disclosure: The iTunes links are affiliate links, and I will receive 5% of the sale price if you buy an album from Preview.fm. So buy a $9.99 album and I’ll get 50 glorious cents. Buy 20 of them and I can buy my own album! It’s like those terrible “Get 10 CDs free!” music clubs from middle school, but in reverse. (Some of my 10 free CDs in middle school: Trio, White Town, Erasure. Remember any of yours?)

[Flash 10 is required to watch video.]

What do other people do on planes?

iPhone multitasking and background updating

One of my most common feature requests is for Instapaper to periodically download articles in the background. A lot of people forget to launch the app to let it download content before going underground or boarding a plane.

I’ve already received multiple emails from people who are excited for iOS 4’s multitasking because they can’t wait for this to finally stop being an issue, because they think Instapaper will be able to download articles periodically in the background.

It’s painful to respond, crushing their hopes, to tell them that the iOS multitasking system doesn’t allow me to do that.

By naming these features “multitasking”, Apple has set customers’ expectations to include what apps can do in a traditional computer multitasking environment.

It’s going to mislead people into expecting such behavior from apps, but we can’t actually deliver most of it.

Some people will notice that no apps can do these things and properly focus their disappointment on Apple. But many others will only notice the shortcomings in one particular app that they need to do the “impossible” and blame that application, leading to dissatisfaction and negative thoughts about the app.1

As long as iOS “multitasking” can do much less than traditional multitasking — which will probably always be the case — this is going to be an issue.

Proposed solution: A new multitasking type

The addition of one more multitasking service would solve this issue for a lot of application types: a periodic network request. Here’s how I would do it:

Executing the request “whenever it deems that it should” is important. iOS can decide, for instance:

This would allow Instapaper to download updates in the background, and would also greatly benefit RSS readers, Twitter clients, chat programs, weather and news widgets, and a huge number of other applications that currently can’t get much benefit from iOS’ multitasking.


  1. I saw this effect when people would email me to ask if I could please remove the scroll-to-top feature when they accidentally tapped the status bar. They didn’t realize it was a system-wide gesture that I couldn’t disable. I had to get creative↩︎

He likes toffee cookies, exceptional coffee, new socks, key lime pie, pants, and Trappist ales.

Tiff, capturing me in one sentence 

People or things who wish me a happy birthday every year in an automated or semi-automated fashion

I’d like to extend a special thanks to Paul Kafasis, John Gruber, Maja Henderson, Patrick Gibson, Neven Mrgan, Christa Mrgan, Jim Ray, Jessie Char, Adam Lisagor, Roxana Altamirano, Scott Simpson, and Guy English for inviting me to what became an impromptu birthday dinner tonight. It was a blast, and it’s great to have such awesome friends to celebrate with, especially since I had to be out of town and away from my awesome wife.


Photo by Adam Lisagor.

Thanks, guys.

How I write and time-manage

I was asked a few times by some very nice people at WWDC this week how I manage my time between Instapaper and Tumblr, and how I write essays here that occasionally make sense. If you’ll forgive my auto-back-patting, here’s the answer I gave, and I think it’s worth sharing with you because you can do the same sort of thing for your benefit.

At SXSW 2009, John Gruber and Merlin Mann gave a spectacular talk about producing high-quality creative work. Listen to it. Really. Stop here and read this later after you’ve listened.

Ready? OK.

In this talk, Gruber says that when he’s writing Daring Fireball, he’s picturing his ideal reader — a copy of himself — and conceptually writing just for him. With everything he writes, he’s writing to and for that one ideal reader, not trying to boost his SEO for target phrases or appeal to an ever broadening demographic.

Well, I do the same thing, except that my ideal reader is John Gruber. That works pretty well, in the sense that I try to reach a high enough quality standard to match my perception of his.

This really does work to improve my writing. Imagine if every PC manufacturer pretended like Steve Jobs was going to look at their new laptop case design before it went to market. I think they’d try a bit harder.

Managing my time is trickier. The short answer is that I cherry-pick: Instapaper is a collection of fairly simple things. I don’t do anything for it that requires massive amounts of time, because I simply don’t have enough time to do that. It usually ends up taking 4-8 hours per week, which fits easily into a few evenings or a single weekend day.

This is where Merlin’s influence comes in. He’s great at reminding me, via his excellent talks and interviews, that it’s easier to be highly productive when you only have a limited time window in which to do so. It’s easier to perform an amazing, in-the-zone, four-hour block of work on a Friday night if I’m leaving town the next day and I know that it’s is the only chance I’ll get all week to work on Instapaper.

And, similar to the way I envision writing to Gruber, it’s useful to picture Merlin looking over my shoulder when I’m slacking off and browsing the internet instead of working.

Merlin voice in my head: “Is that really a good use of your time? What did you make today?”

And that’s about it. It’s remarkably effective.

(Hi, John and Merlin. I hope this isn’t creepy.)

Employing tricks like needless pagination, auto-refreshing (see Salon.com), misleading headlines, and the like is cheating. You didn’t earn those pageviews, you tricked people into giving them to you. And then you look at shit like popups, popunders, double underlined links, Snap previews, Tynt scripts, and so on, and it’s pretty clear how hostile it all is. It’s nothing but money-grabbing. If you’ve got it set up so bad that your readers are employing things like ad blockers and Safari’s Reader, you fucked up. You did something wrong. You overestimated how much your readers are willing to tolerate.

Richard Dunlop-Walters 

Everyone applauds the hyper-connectedness we’re experiencing today. The truth is we can’t really leverage it in a very meaningful way. There are a chosen few that get proper attention, the rest just end up in a sort of long tail of human connections. They’re relegated to an almost trivial status – only acknowledged as a scored point on your ‘friends’ or ‘followers’ tally.

Richard Ziade: Growing Old Friends

The coffee setup, 2010.

I own a lot of ways to make coffee, but I’ve settled into a pattern of using primarily these for home preparation.

It’s a way to make one cup at a time, in less time than it takes to use a normal drip pot, with much easier cleanup, and with a much better cup as a result.

Pictured, left to right:

To brew a cup, I:

Comments

Why John Gruber doesn’t have comments on his site:

Now that [Daring Fireball] has achieved a modicum of popularity, however, what I tend to get instead aren’t queries or complaints about the lack of comments, but rather demands that I add them — demands from entitled people who see that I’ve built something very nice that draws much attention, and who believe they have a right to share in it.

They don’t.

Bijan Sabet likes comments:

For the blogs I frequent, the comments are 99.9% respectful, entertaining, informative and rewarding. Occasionally there is a bad apple …but that’s just life. Same is true on my site.

My experiences with comments haven’t been as positive. Blogs with good comments do exist, like Bijan’s and many of the small tech and VC blogs that I assume he reads, but they’re unusual.

I’m fiercely independent, to a fault. I dislike relying on anyone for anything, and I’m not a very good “team player”. I don’t see my writing as a collaborative effort, and I don’t see my site as a community in which I need to enable internal discussion via comments.

I also disagree with the widespread notion that comments are “discussion”, or that they form a “community”. Discussion and communities require mechanics such as listening and following up that are rarely present in comments.

A blog post is a one-to-many broadcast. Comments are the opposite: many-to-one feedback. A true discussion medium would encourage more communication between the commenters, forming larger quantities of many-to-many interactions and de-emphasizing the role of the blog post’s author. In practice, that rarely happens.

If comments are behaving as many-to-one feedback, there’s minimal value to showing them to the world, because the world largely doesn’t read them. But the act of showing them to the world — your world, not the commenters’ — creates a setting in which commenters are encouraged to behave negatively.1

We already have a widespread many-to-one feedback medium that avoids this: email. So that’s the feedback system that I allow on my site. Anyone can email me, and I will read it.

Those who truly want to start a discussion usually have their own blogs, so they can write their commentary to their audience. If it’s a Tumblr reblog, I’ll see it and read it. If it’s an external link and they email me with the link, or they make a corresponding Twitter post mentioning “marcoarment” or “Marco Arment” or a URL containing “marco.org” or any short URL resolving to something that contains “marco.org”, I’ll see it and read it.

I don’t make it difficult to give me feedback.

What’s not possible is reaching my audience, on my site, without my permission.

Given that this site represents me, and I’ve earned an audience over a very long time of people who generously allow me to take tiny slices of their attention on a regular basis, I don’t think that tightly controlling its content is unfair.


  1. If the emails I get are any indication, I’m doing the world a favor by not enabling comments.

    Plenty of sites get good commments, but it’s not the common case. I can’t identify any general metrics on whether you’ll have good comments, but writing opinion pieces to an audience of tech people is definitely not a recipe for civil, intelligent discussion in traditional commenting systems. ↩︎

How to add files to Subversion containing “@” in their names

This doesn’t work:

$ svn add icon@2x.png
svn: warning: 'icon' not found

This does:

$ svn add icon@2x.png@
A  (bin)  icon@2x.png

You must append another at-sign (“@”) to any filename that contains one that isn’t intended to be interpreted as a revision specifier.

I suspect this has been a first-time issue for a lot of Subversion users this week.

Lucius Kwok’s blog has been great recently.

My low-contrast iOS 4 wallpaper

The built-in iPhone OS (now iOS) 4.0 wallpapers are all too high-contrast for my taste. They impair my ability to quickly distinguish app icons’ shapes and names. (Honestly, I really dislike this feature. But, like hit-song ringtones, I know Apple had to do it to fit into the market.)

But I liked having a bit of texture. So I found a basic black leather texture graphic1, decreased the contrast, darkened it even further, cropped it to an ideal section, and made this:


(It looks even darker on the phone.)

If you like it, download the wallpaper image.


  1. I apologize to the creator of this texture: I can’t find it again to properly attribute it. It’s been sitting in a folder for a long time. ↩︎

tiff-i-brought-that-back-from-illinoise-marco

Rambling about Verizon

I use my MiFi1 every day on my Metro-North train commute. It’s a semi-hostile environment for cellular signals: there’s a rock cliff adjacent to the railway for many portions of the trip, and the last 10 minutes are spent in a tunnel.2 But many parts of the trip are above ground with no obvious obstructions, so data service should be fine.

It’s not. Not even close. It’s spotty at best: sometimes I get full EVDO signal strength, and sometimes I get no reception at all. The MiFi switches between EVDO and the old 1X network constantly. More often than not, in this corridor, I’m on the 1X network.

When traveling, I find EVDO coverage to be good, but not great. I still fall back to 1X in less-populated areas. EVDO definitely has not been universally deployed on Verizon.

To the best of my knowledge, Verizon doesn’t make any phones that can’t use 1X, even though EVDO has been widely deployed since 2005. Everything needs to be able to fall back to 1X because it’s still not universal.3

To expect Apple to wait until LTE to make a Verizon-compatible iPhone, therefore, seems unlikely. LTE hasn’t been deployed anywhere yet, but the current promise is that it’s going to start being available late this year. Based on the deployment of EVDO, when might it be practical to own an LTE-only phone? 2016?

Even an optimistic and probably unrealistic estimate of 2013 is far too long of a wait. Apple’s quickly losing ground to Android today for only one reason: the majority of mobile-phone owners in the U.S. (the majority of people in the U.S.) choose their network first and their phone second. The phone selection is nearly an afterthought. They go to the nearest red phone store when their two-year contracts expire and pick out their next phone from whatever they see in the store.

Half of those people are going to stores that don’t sell the iPhone. And when they get there, they’re barraged with giant Droid banners depicting big globes of iPhone-like “app” icons and an ad campaign stating that Droid phones do everything that iPhones do. Store employees are rewarded with contests and bonuses for selling as many Droids as possible, and Verizon is pushing them strongly with two-for-$99 sales.

Verizon has reached a powerful point in their marketing: for Verizon customers curious about the iPhone, Droid is close enough. Close enough is powerful, and Apple is rapidly losing ground to it4.

Droid isn’t actually “close enough” to the iPhone in most important ways, but in marketing and customer perception, it doesn’t matter. Apple can’t win this fight on quality and overall experience because most of these customers have never owned iPhones. They don’t know what they’re missing. They just know what the Verizon marketing told them: Droid phones are pretty much like the iPhone. When they encounter all of Android’s rough edges, they assume that all smartphones are like that, and grow to generally dislike using them. (Much like the computer market.)

Apple only has two options: either give up the mass market in the U.S. and accept slow growth in U.S. sales while Android undermines the iPhone’s base, or launch a CDMA 1X+EVDO Verizon iPhone very soon.

How soon is difficult to say: the sooner, the better. Technology is almost certainly not the limiting factor here — I imagine a CDMA iPhone 4 is ready to go, or close to it, today — so it’s a matter of other influences, such as product timing with the GSM iPhones, supply chain delays, or contractual issues with the carriers.

Every month without a Verizon iPhone costs Apple dearly in marketshare and mindshare. The longer they wait, the lower their chances to reclaim the difference.

I’m guessing a CDMA Verizon iPhone will be available within 6 months.

If it isn’t, I might need to start learning Java.


  1. I wasn’t using it during the keynote. It stayed on the fake desk in my hotel room all week so I wouldn’t need to buy the hotel Wi-Fi. ↩︎

  2. Verizon is the only carrier that provides any cellular service in this tunnel, and they only started doing so in the spring of 2009. But service ends as soon as the train reaches Grand Central, at which point I can get strong AT&T 3G service. So when I’m waiting for the train to depart in the evenings, there’s no Verizon service. I can’t win with any single carrier. ↩︎

  3. Preliminary research seems to indicate that EVDO really is data-only and voice calls are placed on 1X. Still, even this data-only MiFi can’t afford to be EVDO-only. ↩︎

  4. To be clear, Apple is losing ground to “Droid”, not Android. Very few people are choosing Android as a platform first, then finding the carrier that has the best Android phones. Yes, I know you chose it because you were angry with AT&T or you dislike Apple’s policies, but you’re probably a geek like me, and most normal people have absolutely no idea what Android is and wouldn’t care if you told them. ↩︎

My life is a series of Apple lines.

By Tiff. That’s about right.

Quick iPhone 4 impressions

The screen is incredible. It looks like a glowing printed page, as if someone took one of those big vectorized iPhone display stand-ups that they put in the store windows, and shrunk it down.

The screen’s contrast ratio is extremely high. I’m probably going to need to change Instapaper’s Dark Mode color scheme on the iPhone 4 to be closer to what I use on the iPad.

Some common visual techniques, like the 1-pixel white shadow used to emboss header and footer text in grouped tableviews, look strange and a bit harsh at this resolution.

It’s very apparent when an app hasn’t updated its graphics at 2X yet. Honestly, it looks terrible. Developers: Update your artwork ASAP.

The camera is much sharper than the 3GS camera. I suspect this is going to wipe out point-and-shoots even further from the iPhone-owning demographic. I may have second-guessed the S90 purchase if I had this iPhone camera 8 months ago.

The entire interface is fast. It wouldn’t surprise me at all if this A4 runs at the same clock speed as the iPad’s.

The ring/vibrate switch is extremely different. It now has much less travel distance and is much harder to push. This makes it more difficult to tell what state it’s in by feel alone, but should result in fewer accidental toggles when it’s in your pocket.

The $29 “Bumper” is hilarious. I think Apple is really in the iPhone-case business, and the iPhone is just an attempt to sell us something to encase. It’s day one, everyone’s freaked out about breaking their new all-glass iPhones, and they’ll sell you a “bumper” case — the only case available for the iPhone 4 in the Apple store today — for $29. Looking at this tiny piece of rubber, you really have to admire Apple for having the balls to charge $29 for it. And they’re going to sell a ton of them. The best part? It comes in a lot of colors, but today, only black is available. So if you’re really set on a teal or pink one, but are paranoid about breaking your new iPhone, you’re probably going to end up buying two of them.

But I have a feeling we’re not going to do that. I have a feeling that, because this would cost a lot of money, we’re going to let it ride. … In another ten years or so there will be another deep-sea drilling blowout, and we’ll express outrage that the necessary lessons weren’t learned after this blowout. But we won’t do what it would take to actually prevent the next one.

The Economist

Patent pending.

NEW iPhone 4 antenna booster! Just $29.99. Really works!*

* As tested compared to licking your hands and squeezing really hard in a weird way that most people are unlikely to do accidentally at the same time that their hands are wet and they’re transferring a lot of data. Results not guaranteed.


I should clarify: this post is satire. I don’t actually use my phone with tape on it — I just put it on there for this photo. It’s a joke. Even though I always hold my phone like this, I haven’t been able to reproduce the supposed problem. Hell, Gizmodo broke theirs trying to squeeze it in just the right way. I really don’t think this is a significant issue.

A LOT of people have mistakenly thought that I was actually using my phone this way and suggesting this as a real solution. Actually, it probably would work. But I don’t need it, and you probably don’t, either.

In most elevators, at least in any built or installed since the early nineties, the door-close button doesn’t work. It is there mainly to make you think it works. (It does work if, say, a fireman needs to take control. But you need a key, and a fire, to do that.) Once you know this, it can be illuminating to watch people compulsively press the door-close button. That the door eventually closes reinforces their belief in the button’s power. It’s a little like prayer. Elevator design is rooted in deception—to disguise not only the bare fact of the box hanging by ropes but also the tethering of tenants to a system over which they have no command.

Nick Paumgarten: Up and Then Down. (Told you so, everyone who has tried to convince me that our elevators’ door-close buttons did anything.)

Cameron Hunt: The Spot

OK, now I’m a believer. Via Daniel Jalkut, who got me with this: “The one-finger Pause My Network™ feature in iPhone 4 is awesome!”

Glass MacBook Pro screens are heavier

I finally got to weigh my “anti-glare” matte 15” MacBook Pro against my friend’s nearly-identical 15” with the default glass screen on the same scale.

To be fair, the components weren’t identical. He has the newest model with the higher-resolution screen, and I have the summer-2009 model (the first unibody 15” model available with a matte screen). And I replaced my disk with an X25-M, which reduced the weight by 0.03 lbs.

But that’s still a significant difference. It doesn’t make it a MacBook Air (3.0 lbs.), but keep in mind that the original white plastic 13” MacBook was 5.0 lbs. (Today’s 13” aluminum MacBook Pro is 4.5 lbs.)

AT&T’s data network has actually, finally, improved in Manhattan

AT&T claims that the New York City network has been significantly upgraded. In initial tests, this appears to be the case.

In March, I posted this speed-test graphic to show how bad the data service was in Manhattan.

Now, I’ve updated it. Keep in mind that this is relatively unscientific, as it’s only one data point, and I moved from an iPhone 3GS to an iPhone 4 since the original test:

It’s a huge improvement. (Unless I’m touching the antenna gap. Yeah… I was wrong about that. It’s a real problem.)

I’ve also spent the evening using the data connection around midtown and the Metro-North corridor casually, and the difference is like night and day. This isn’t to say that they just turned it on today, but there has definitely been a very recent change — probably within the last few weeks, at most — that made a huge difference. I haven’t noticed it before today because I’ve trained myself never to use the AT&T data network in Manhattan if I can help it.

I hope my limited impressions and tests reflect city-wide network quality. And I hope it stays this way.