Marco.org

Jacqui Cheng at Ars Technica wrote an article on why RSS is bad for you:

The first time I went without RSS in August, I simply went around to three or so of what I consider to be the best sites to get the latest news from. I combined that with my usual e-mail communications … and my regular scans of Twitter in order to figure out what was going on during the day. It was stress-free, and I never felt like I was missing anything—I knew that if something truly important or controversial blew up, I’d hear about it instantly via Twitter and our loyal readers.

The next day when I loaded up my feeds, there were literally thousands of items piled up from the day before. … And when I ended up sifting through them all, I realized that I hadn’t missed a single story doing things the “old fashioned” way—rather, by following all these feeds, I was instead seeing hundreds of iterations on the same handful of stories. And I was wasting time going through them all day long.

RSS is a great tool that’s very easy to misuse. And if you’re subscribing to any feeds that post more than about 10 items per day, you’re probably misusing it. I don’t mean that you’re using it in a way it wasn’t intended — rather, you’re using it in a way that’s not good for you.¹

You should be able to go on a disconnected vacation for three days, come back, and be able to skim most of your RSS-item titles reasonably without just giving up and marking all as read. You shouldn’t come back to hundreds or thousands of unread articles.

What Jacqui did in RSS’ absence is always helpful: letting other people filter popular news sites for you. This is critical to sane RSS usage so you don’t need to subscribe to the frontpage feeds of high-volume blogs (Engadget, Lifehacker), aggregators (Reddit, Hacker News), or general news sites (The New York Times, CNN). If you like those sites, either browse them “manually” without RSS whenever you feel like it, or just wait for people to link to it from Twitter, Tumblr, Facebook, or Google’s social network of the year. And let this role inform your following decisions on these networks.

But if that’s the only way you get your news, you’ll only ever see the most popular articles. If you’re trying to get away from the “echo chamber”, that’s going to make your problem worse.

RSS is best for following a large number of infrequently updated sites: sites that you’d never remember to check every day because they only post occasionally, and that your social-network friends won’t reliably find or link to.

I currently subscribe to 100 feeds. This morning, I woke up to 6 unread items: one each from 6 of my feeds. Granted, it’s a Sunday on a holiday weekend, so this is a pretty low-activity day. On high-activity days, I usually wake up to about 25 items.

I don’t use an RSS app on the desktop anymore: I just use the Google Reader site. I can check it whenever I want, but nothing’s in my Dock collecting red badges to distract me every few minutes.²

This setup works well. I can follow tons of low-traffic sites and keep my reading list more diverse than if I relied only on social links, but other people ensure that I never miss anything great on the high-volume sites.

This post is extremely geeky. If you’re not into computer minutia, feel free to skip it.

My current Mac Pro era, started in February 2008, is over. My flings with MacBook Airs are also over for the foreseeable future.

I’ve switched back to a 15” MacBook Pro full-time, right when the MacBook Air is convincing many geeks like me to dump their 15” laptops for an 11” or 13” Air, possibly with an iMac at their desk. Many will think that this is strange timing, so it might be interesting to those as geeky as me why I chose this.

Why the Mac Pro before?

When I bought the Mac Pro in 2008:

• The biggest laptop hard drive was 250 GB, and laptop hard drives were much slower than desktop hard drives.
• Disk performance was the primary performance bottleneck. On laptops, there wasn’t much you could do about it. The best way to get good disk performance was to use a desktop and stripe multiple fast hard drives in a RAID array.
• MacBook Pros only supported up to 4 GB of RAM, while the Mac Pro supported four times as much. (High-capacity modules came out later that made it actually support eight times as much.)
• Midrange MacBook Pros had about a quarter of the CPU power as midrange Mac Pros.
• Midrange Mac Pros cost about $900 more than midrange MacBook Pros.

These conditions are very different now, primarily because of the biggest advance in computer performance since cheap RAM in the late ’90s: SSDs.

SSDs did more to close the performance gap between laptops and desktops than I think anyone in 2008 would have imagined:

• An SSD is more beneficial to general performance than any hard drive or RAID setup. For most users, it’s even better than adding RAM.
• Desktops and laptops use the same SSDs, with the same prices and capacities, at the same speeds.

The other gaps have narrowed as well: today, midrange Mac Pros have twice the RAM capacity and about twice the CPU performance.¹ And the price gap has widened: the midrange Mac Pro costs about $1500 more than the midrange MacBook Pro.²

The Mac Pro is due for an update, but I realized recently that I didn’t even care what it was: laptops have become so awesome in the last few years that I no longer needed a desktop.

More recently, two factors have made this even more clear: today’s mobile quad-core Sandy Bridge CPUs are incredibly fast and surpass many desktop CPUs in performance, and with Thunderbolt, laptops now have a full-speed external expansion bus.

Why not a desktop and a laptop?

Since I never stopped buying laptops since getting the Mac Pro, I’ve been a multiple-computer user since 2008. Trust me, it sucks.³

Yes, Dropbox helps a lot. But multi-computer usage still sucks for so many reasons, because most files, applications, and settings don’t sync. And if your laptop has a smaller capacity or lower performance than your desktop (as it probably does), it’s going to be less useful than you need, more often than you think.

The economics aren’t great, either. Having two computers means at least two hard drives or SSDs to buy and keep backed up, two sets of RAM to maximize, two AppleCares, and two licenses for some apps.

Why not a MacBook Air?

The MacBook Air is great. The 11” is extremely small, and the 13” packs a pretty decent amount of power and screen space into its size. An Air is a great choice for nearly everyone.

But I’m more demanding of my hardware than most people. If a laptop is going to be my primary computer, I need more CPU power, storage space, and screen resolution than the 13” Air offers.

And the Air’s reduced size and weight aren’t as important to me anymore. I work at home now, and when I go on quick day trips, I hardly ever bring a laptop, preferring an iPhone and optional iPad instead. When I do bring my laptop somewhere, such as when traveling, I want more power and screen space than the Air offers so I can get work done. And I usually travel by car, where size, weight, and tray-table compatibility are unimportant.

Why buy a 15” now?

It’s strange timing. The current 15” MacBook Pro was released in February, and there are rumors (including my own speculation, twice) that the next 15” MacBook Pro update is going to be a major revamp with an Air-like shape, lower weight, and possibly even a Retina display.

But if that’s true, the new 15” will probably also be Air-like in its internal storage options: SSD-only, with a maximum capacity of maybe 512 GB, and no standard 2.5” drive bays.

That’s going to be great when the internal SSD capacity grows large enough to hold all of my stuff, but that’s unlikely to happen in the first version: I need about 1 TB of internal space to be comfortable today, and that grows over time.

Rather than eagerly anticipating the next 15” revision, I took this opportunity to buy what’s likely to be the last 15” MacBook Pro that can hold two standard 2.5” drives.

I won’t need that kind of third-party internal expandability forever, but I still do today.

So what did I get?

I got the fastest 15” configuration with the high-res antiglare (matte) screen, which was available refurbished for $2199 ($400 below retail). It arrived in perfect condition, with no cosmetic imperfections or bad pixels.

The savings from getting it refurbished helped me deck it out with OWC upgrades:

That’s 8 GB of RAM, a 480 GB SSD,⁴ and the stock 750 GB hard drive in the Superdrive bay with the Data Doubler. For my rare optical-drive needs, I got this CD/DVD/Blu-ray burner, which is portable and bus-powered even though I don’t anticipate taking it anywhere.⁵

The new MacBook Pro’s front edge is much less sharp than the edge of the 2009 model, and I took a macro photo that looks like a butt to confirm:

Left: 2009 MacBook Pro. Right: 2011 MacBook Pro.

It’s much more comfortable to type on the 2011 model, with one big exception: the 2011 model runs much hotter.

I need more time to form a concrete opinion, but it seems so far that this CPU’s awesome performance definitely comes at a cost of increased heat and reduced battery life. For my intended usage as a desktop most of the time, that’s an acceptable tradeoff.

I’ve been using it full-time for a few days. So far, it’s an awesome system, it’s as fast as my Mac Pro was (and even faster in some ways), and it feels great to be back on one computer again.

A quick tip I just found after too much digging in Google:

Deleting an email address from Mail.app’s autocomplete list can easily be done from the Window, Previous Recipients menu. Nearly every search for autocomplete adjustments says this.

But I wanted to do the opposite: add addresses to it, since I rely so heavily on autocomplete when composing mail, and my new clean installation started with an empty list. I still can’t find any way to bulk-add — I’d love to just autocomplete every recipient’s address from every message in my Inbox or folders.

But I did find the autocomplete file’s location so I could migrate it over from my old computer’s disk:

~/Library/Application Support/AddressBook/MailRecents-v4.abcdmr

I quit Mail, copied that to the new computer (mine was 954 KB!), overwriting the stock one (only 49 KB), and relaunched Mail. All of my old autocomplete entries were back with no apparent side effects.

Here’s another solution to a problem I just had, hopefully to save future Google searchers some time. How to fix this:

• Open iChat’s Preferences.
• Under Accounts, select Server Settings.
• Try unchecking “Use SSL”, or changing the Port to 443, or both.
• Try taking iChat online again. You may not need to — if it’s this problem, it should take you online immediately.
• Take iChat back offline, go back to Preferences, Accounts, Server Settings, and click “Use Default Settings” (which changes it back to SSL and port 5190).
• Try to go online again and it should work.

It seems that Lion’s networking is still a bit finicky under 10.7.1. (This also might be the AOL servers’ fault, but probably not.)

Ben Brooks noticed and blogged about how Instapaper’s social features, introduced earlier this year, are minimal:

There’s just a list of articles that people you chose to follow decided that they liked. All without knowing who, or if, anybody will ever see that they liked that article.

It’s a fascinatingly private social system.

That was exactly the idea, and I’m very happy to see it perceived that way.

Social features are tricky. Social dynamics in real life are complex, so every social mechanic we construct or word choice we make will carry unintended baggage, connotations, and ambiguities. (“Like”. “Friend”. “Follower”.)

Social networks also need to address difficult issues with identity, privacy, harassment, spam, and information overload.

These systems require a lot of time and money to develop, maintain, and support. And when they’re ready, they need to compete with all of the other social networks for people’s time and attention.

With Instapaper’s following system, I wanted to deal with as little of the difficult baggage as possible, even if it meant omitting some of the “sticky” social dynamics that can significantly boost user counts and engagement. The result is a very small social feature-set that piggybacks on other established social networks: Twitter, Facebook, and email.

There are no public usernames, avatars, or profile pages. Nobody’s quitting Facebook for Instapaper. Companies aren’t rushing to establish an Instapaper following. No newly engaged couples have rushed to update their statuses on Instapaper.

To label each story in the interface with the person it came from, Instapaper just uses the label from however you found them. If you found me by searching people you follow on Twitter, I’ll be “marcoarment”. If you found me through Facebook, I’ll be “Marco Arment”. And if you found me by email address, I’ll be “me@marco.org”. Instapaper could cross-reference these, but it doesn’t.¹

There are no notifications whatsoever for following and unfollowing. Nobody can tell who follows them, how many people do, or even if anyone does. In addition to removing the emotional rollercoaster of follower counts and unfollows, this may actually increase following activity: if people realize that others won’t know when they follow or unfollow, they may feel more comfortable doing so. (I sure do.)

In short, I want to leave the social features to the social networks. I want to use them to make Instapaper better, not try to make Instapaper replace them. They can deal with all of the baggage and reap all of the rewards. I’m not interested in that game.

Instapaper takes advantage of your social networks to let you easily share what you’re reading and give you recommendations when you want them (and only then), but remains a quiet escape from the social networks when you just want to read.

One of the uncomfortable and mildly offensive quirks about the Microsoft culture is that they never acknowledge competitors. This was particularly obvious this week: in all of the Metro videos and developer talks that I’ve seen so far, I don’t think they’ve said “Apple” or “iPad” once.

Instead, they painfully dance around to avoid it, equivocating and genericizing even the most iPad-specific references with “other tablets”, “some competitors”, “you may have seen”, “out there”.

They know they’re talking about the iPad. Everyone in the auditorium knows they’re talking about the iPad. All of us on the internet know they’re talking about the iPad.

This awkward avoidance betrays a lack of confidence in their innovation and an internal culture of severe denial — two deep-rooted traits that Microsoft is famous for. So when the Microsoft people speak like this, it’s not serving them well: to everyone outside of Microsoft, it’s painfully obvious that they’re either delusional or trying very poorly to bullshit us.

And I’m honestly not sure which it is.

With so many projects, if the customer is willing to go without a small subset of the functionality they think they need, it can save a massive amount of effort, cost, and complexity and result in a much more elegant, hassle-free solution that makes them much happier in the long run.

This option isn’t always feasible. Sometimes, “needs” really are needs. But often, people’s “needs” are much more flexible than they think.

Apple’s customers are often the sort of people willing to make these tradeoffs, because that’s how most of Apple’s products are designed: if you can compromise on some of the features and capabilities you think you need, you can get a product that works better and makes you happier with far less aggravation. And for most people, the benefits will outweigh the missing features.

This culture of compromise has been cultivated by Apple’s relentless pace of forcing progress and killing legacy support. Apple’s implicit message is simple: “We know what’s best. If you do things our way, everything will work very well and you’ll be happy. If you don’t like it, that’s fine with us.”

People who aren’t willing or able to compromise on their needs regularly are much more likely to be Windows customers. The Windows message is much more palatable to corporate buyers, committees, middlemen, and people who don’t like to be told what’s best for them: “You can do whatever you want, and we’ll attempt to glue it together. It won’t always work very well, and you might not like the results, but we will do exactly what you asked for.”

For many (or even most) of Apple’s products, the forced compromises are critical to their greatness. There’s no way to achieve delightful, well-designed, deeply integrated platforms with highly polished software and very few problems, but that are also generous with access rights, open to deep modification by middlemen or end users, and able to run on a large, diverse, and unregulated pool of hardware.

For Windows 8 to succeed, especially for its tablets to compete well against the iPad, it will need to adopt a more Apple-like confidence to say “no” to its customers when they ask for everything under the sun. And, surprisingly, they might be doing exactly that.

One of the reasons Metro is interesting to people like me who usually ignore Microsoft is that it’s full of very un-Microsoft-like decisions, generally for the better.

The question isn’t whether Metro will be good: it probably will be. And that’s a huge accomplishment for Microsoft that they should be commended for.

But how will their customers react?

Will Metro be meaningfully adopted by PC users? Or will it be a layer that most users disable immediately or use briefly and then forget about, like Mac OS X’s Dashboard, in which case they’ll deride the Metro-only tablets as “useless” and keep using Windows like they always have?

How will Windows developers react to the restrictive Metro environment after being able to do almost whatever they wanted for nearly two decades?

And how well will the many groups within Microsoft, a company famous for infighting, adopt a fundamental shift toward simplicity and restraint when so many of them seem to be going in the other direction?

There’s trouble in my laptop paradise. I ended that with this caveat:

It’s much more comfortable to type on the 2011 model, with one big exception: the 2011 model runs much hotter.

I need more time to form a concrete opinion, but it seems so far that this CPU’s awesome performance definitely comes at a cost of increased heat and reduced battery life. For my intended usage as a desktop most of the time, that’s an acceptable tradeoff.

This turned out to be worse than I thought it would be, and today, the last day of my return period, I decided to return it and exchange it for a slower model.

The problem

Not only does the 2011 15” MacBook Pro run much hotter than previous generations, but as a result, the fans very frequently kick up to full speed (6200 RPM), greatly annoying anyone in the area who’s not wearing headphones.

I’ve done a lot of Google research and impromptu Twitter surveys, and this seems like a very common problem, exacerbated by a number of factors:

• The quad-core i7 CPUs in the 15” run very hot, with a TDP of 45W, 10W higher than the previous dual-cores. Many people speculate that the 15” case and fan design just isn’t designed for CPUs that need this much cooling regularly.
• The Core i7 CPU can use Intel’s Turbo Boost to dramatically overclock one CPU core to handle single-threaded loads whenever power and heat allow. It therefore hits its maximum TDP often when just one core is being maxed out, such as when Spotlight or Backblaze index new files, or when Time Machine is backing up.
• The 2.2 and 2.3 GHz 15” models come with an AMD Radeon HD 6750M GPU, which adds significant heat. Modern 15” MacBook Pros dynamically switch between the low-powered Intel integrated GPU and the high-powered (and hot) AMD GPU as needed by applications, but the AMD GPU will always be used when an external monitor is connected — the Intel integrated GPU cannot output video over the Thunderbolt port in the 15”. (Oddly, it can in the 13” MacBook Pro and both MacBook Airs.)
• It generates so much heat that it can’t be used in “clamshell mode” (closed, but with an external keyboard, mouse, and monitor attached), which has reduced airflow, without risking overheating, kernel panics, and possible long-term damage. I can’t see why Apple even enabled clamshell mode on it.
• Some people report issues with the battery draining while playing a game. The 15” MacBook Pro comes with an 85W power adapter, but when stressing the CPU and GPU (such as when playing a game), the laptop draws about 93W — I don’t know enough about the electronics here to say for sure, but an insufficient power adapter capacity sounds like a plausible reason why the battery might drain under high computational load. If so, this is an embarrassing oversight by Apple.
• The thermal paste seems to be very inconsistently applied, or there’s some other common manufacturing defect with the cooling system that varies per laptop. Many, many people reported having these issues and exchanging the laptop for a new one that shows far fewer issues. Many dedicated tinkerers have also reported success after reapplying the thermal compound between the CPU and heatsink, but that requires a very deep, warranty-voiding disassembly that I’m not comfortable doing. (I also haven’t owned thermal paste since 2001.)

Put simply, it sounds like the 2011 15” MacBook Pro has a lot of manufacturing inconsistencies and defects that result in improper cooling. But it also sounds like Apple wedged far too much power draw and heat into an enclosure that’s not designed to properly cool or power these demands.

It’s not like Apple to release products like this.

But if I can get one working within my tolerances, it’s still the perfect computer for me right now. So I’m trying.

The attempted solution

Tonight, after requesting the return on my 2.3, I ordered a new, slower model that’s likely to run cooler with less fan noise. It’s the base-model 15”, but with an antiglare (matte) display added:

• AMD Radeon HD 6490M: It’s manufactured on the same 40nm process as the 6750M, but has far less computational power, so I’m guessing this will result in much lower power consumption and heat.
• 2.0 GHz quad-core i7: This is the only CPU available with the 6490M. While I don’t expect it to be dramatically cooler-running than the 2.3 since it’s specced with the same 45W TDP, it does have a lower clock and lower Turbo Boost speeds, so I can probably expect an approximately proportional heat reduction.

So I’ll see what happens. I’m not sure what I’ll do if it doesn’t work out. This generation of 15” MacBook Pros has a lot of problems, and Apple really needs next year’s cooler-running Ivy Bridge processors as soon as possible.

After reading Ryan McCarthy’s post at Reuters about Business Insider, I wanted to run some numbers.

After all, they’ve linked to nearly every significant article I’ve written for the last few years, often automatically by scraping Techmeme.

When humans have been involved, they’ve rewritten my titles to be more inflammatory and attract more clicks, which irritates me more than how much their cluttered, ad-overloaded site “design” buries the link to my article:

Continue reading at Business Insider »

But what offends me even more than rewriting my titles and burying my links is how their layout so strongly implies that I’m a Business Insider writer and I endorse my name and writing being splattered all over their site:

Click here to continue reading at Henry’s blog…

Why wouldn’t I want to be associated with Business Insider? It has nearly everything that offends me as a web reader and writer: linkbait headlines, more ads than content, more sharing buttons than original words, top-list “slideshows” that make readers click for every item and defraud advertisers into thinking that their pageviews are legitimate, Tynt messing with copy and paste, Vibrant Media’s double-green-underline ads, generic images slapped next to each post (often poorly Photoshopped®), and tabloid coverage of every rumor and inflammatory non-event so they can fight all of the other tabloids for Google’s pennies.

Any promise of helpful traffic referrals to the original authors is misleading: the traffic from Business Insider is minimal. Here’s a selection of referral sources and their referred visit counts from January 1, 2009 through today, as measured by Google Analytics:

• StumbleUpon: 731,939 (mostly to one-off funny things like this)
• Daring Fireball: 516,705
• Google searches: 391,743
• Hacker News: 366,123
• Techmeme: 65,834
• The Brooks Review: 15,333
• Shawn Blanc: 13,930
• furbo.org: 12,760
• My wife’s personal blog: 9,729
• Business Insider: 8,891

That’s 8,891 hits that I wouldn’t have otherwise had. But over a span of 995 days during which they linked to my site many times and republished many of my articles, that’s nothing, especially for a site that claims to get 12 million unique visitors per month.

And, if given the choice, I wouldn’t trade 8,891 hits for my name and writing being used on their site like this.

It’s the same game famously played by the Huffington Post and many other “aggregators”:

• Reprint popular news articles and blog posts en masse, copying all relevant details so nobody needs to read the original post.
• Bury the source links so almost nobody will find them, and if they do, they’ve already read all of the important details from your reprint.
• Tell the real authors that you’re helping them build their name recognition and sending traffic their way.

Business Insider once asked me if they could “syndicate” all of my blog posts automatically and give me an official byline on their site in exchange for — you guessed it — links back to the articles on my site.¹ I politely declined, because they’ve effectively done this for years without my consent, and it’s not doing me any favors.

I wonder if they’ll reprint this one. (Update: They did, auto-scraped from Techmeme, but deleted it later.)

Business Insider’s mass replication of my writing is the only downside that has ever made me reconsider my Creative Commons license. If they’ve had any beneficial effect whatsoever, I haven’t noticed.

This talk by John Gruber and Merlin Mann at SXSW 2009 came up again from John’s link to a fan video, and it was a great reminder to redownload the audio and listen again.

It has become much more famous than I imagine John or Merlin expected at the time, but I’m particularly fond of it because of my experience in the audience, which I should share before I become too uptight about my blog appearing unprofessional or embarrassing myself in front of my famous internet friends. I almost didn’t post this even now, and while that usually means that I really should post it, it sometimes correctly means that I shouldn’t. But oh well.

(If you don’t care about my feelings or the 5by5 crew, this will be boring. It might be as interesting as hearing about someone’s hazy recollection of a nonsensical dream. You’ve been warned.)

Tumblr sent my wife and me to Austin for SXSW 2009 to help throw a party. We flew in two hours before this talk, but it was the only thing that interested me on the entire SXSW schedule, so I rushed and made it there.

I still remember how nervous-sick my stomach felt at the restaurant with the Tumblr people beforehand. I was worried that I wouldn’t make it on time. I was going to meet these people in person whose work I admired on the internet, and I had never crossed those worlds before (and I’m not the smoothest person socially). I didn’t have my own badge and had to fraudulently use someone else’s, so I was worried about getting “caught”, whatever that would mean. I had never attended a conference before, so I had no idea how it worked, how tight security was, what it would be like inside the convention center, or where to go once I got there.

But I made it in by strategically ensuring that the badge was always flopped over or tucked under my arm slightly so nobody would notice that I wasn’t John Maloney, not realizing how lax security is at conferences (especially SXSW).

A few minutes before the talk started, I noticed that the guy sitting next to me looked a lot like Lonelysandwich. I was pretty confident that it was him — the guy’s iconic. It took my last remaining nerve to ask cautiously, “Excuse me… are you, uh, Adam?”

“Oh hey, are you Marco?”

What. How. What. (I still don’t know.)

So we got to talking, and then John and Merlin walked in to do the talk. They came by to greet Adam.

Merlin saw me and exploded with Instapaper praise. I was speechless. This guy’s a big deal on the internet, and I was worried about how awkwardly I’d introduce myself and how I’d probably put my foot in my mouth, and here he was recognizing me and introducing himself.

Meeting John was even more dumbfounding. He saw me, said nothing, and then told Adam, “Don’t let him leave.”

Tumblr was big by then and Instapaper was already more than a year old. I knew these guys knew about both, but I never expected to be personally recognized. It was an unforgettable experience already.

Then I saw the talk, during which they referenced my stuff twice. (Mind further blown.) And the talk, as we know now, was not only great but timeless and universal.

It was the first conference session I’d ever attended.

For the following year, I thought all conference sessions were that good, and I was so intimidated that I declined invitations to speak at conferences myself because I thought I had to achieve that level of quality.¹

It wasn’t until I got my own badge for SXSW 2010 and attended multiple sessions there that I realized how bad most of them are, so I was finally comfortable agreeing to my first talk.²

In addition to inspiring me to be a better writer and inadvertently killing my conference-presentation confidence for a year, this famous little 2009 SXSW session leveled my juvenile notion of celebrity. After the talk, since I wasn’t allowed to leave, I was introduced to many more great people famous for their blog, software, humor, or music,³ and it went similarly well with all of them.

Among people who are well-known to subsets of internet geeks, nobody’s walking around with entourages or bodyguards. Nobody’s forced to wear hats and sunglasses outside or avoid shopping in public because they’ll get mobbed with fans and paparazzi. In the wise words of Ted Dziuba, “We all recognize that it’s just the internet. At the end of the day you still go outside and nobody knows who you are.”

The internet “celebrities” I was so nervous to meet are now my friends. I see them a few times each year at conferences. It turns out that we’re all just regular people who like similar things and are in the same little circle of interest.

So next time you’re at a geeky conference and have an opportunity to meet someone whose work you admire, just go up and introduce yourself, because they’re just a regular person, they never get “recognized” during the other 360 days each year, and they’ll probably really appreciate it.

And you should really listen to that talk.

I’m sorry I doubted TextMate 2’s future. With your announcement that it is indeed tangibly progressing, it’s a good time to revisit your former statement that it would be a free upgrade to all TextMate 1 customers, which I think you regret.

Please withdraw the free-update offer. We will not hold you to it.

Really. Please. I’m a TextMate customer and I’ve been using the same license since 2006.¹ This is an application that I use every day to do most or all of my job. And I bet many of your customers will say the same thing.

Given the value that we get out of TextMate, it’s already grossly underpriced. Please let us give you more money.

You promised the free update in a different era, probably expecting different circumstances. Things have changed.

If the free-update offer still stands when TextMate 2 ships, I will not take you up on it. I’m buying TextMate 2 as a new customer at full price. And I bet many other developers will gladly do the same.

TheNextWeb has a good overview of the facts from today’s Amazon event.

This will probably have wide implications, and I’ll definitely have more to say in the future, especially once I’ve actually used these things. (I’ve preordered a Kindle Touch 3G and a Kindle Fire to review here and test with Instapaper.)

The new e-ink Kindles

The e-ink Kindle prices are misleading: the quoted prices on e-ink Kindles are all with “Special Offers” (a euphemism for ads, how “special”). Add $30-40 to the price of each e-ink Kindle to get one without ads.

In my Nook Simple Touch review, I speculated (and hoped) that Amazon would drop the big, ugly keyboard from the next Kindle, and I’m very glad that they did. (I also hoped they’d drop the audio components. They did, but only in the low-end button model.)

The cheapest one, $79 with ads or $109 civilized, retains the 5-way controller from the Kindle 3 (which old video-game nerds like me probably know as a “D-pad”) and looks a lot like older Sony Readers. Navigation is going to be just as clunky as previous Kindles, so I don’t think this is worth getting for most people.

The Kindle Touch, $99 with ads or $139 with self-respect, has an infrared touch screen, just like the Nook Simple Touch. I have high hopes for this one: this is probably going to be the e-ink reader worth owning for the next year or two.

There’s a 3G model for $50 more. That’s a relatively big price premium for a feature that most people will never use, but it’s effectively prepaid: like previous 3G Kindles, you’re not charged for data in the U.S. (They usually have different arrangements that I don’t know enough about in other countries.) So if you do a lot of traveling or are often away from Wi-Fi networks, and you’d like to get new Kindle content wherever you are (especially useful for newspapers, magazines, and Instapaper), the 3G model is probably worth getting.

I’ll post a full review of the Kindle Touch when I receive mine.

The Kindle Fire tablet

This is going to be interesting.

The Kindle Fire lines up with most of my speculation (and everyone else’s): it’s a 7” Android tablet, but Android has been buried under Amazon’s completely custom interface, and it runs all of Amazon’s content stores: Kindle books, music, video, and Android apps. The only surprise is the $199 price, which is lower than almost any of us guessed.

The Fire will be the first Android-powered tablet to sell in meaningful volume, but whether it’s a viable platform on which to sell high-quality paid apps remains to be seen. The Amazon Appstore for Android has been hostile to developers so far.

It’s definitely going to compete with the iPad for some customers, but I doubt it’s going to make a significant dent. It’s probably a high-end Kindle, not a low-end iPad. But this will almost certainly hinder the already negligible sales of other Android tablets.

What we’ll see with the iPad depends on why people buy iPads. My theory is that there’s an iPad market, not a “tablet market” — that people want the iPad and specifically seek it out without comparing it to other tablets.

A “tablet market” suggests that people first decide they want a tablet, then they comparison-shop and choose the one that best fits their needs and budget, like buying a dishwasher. I don’t think we’ve seen any plausible evidence that a meaningful number of customers think of tablets generically like that.

But if anything’s going to prove me wrong, it’s the Kindle Fire.

Time will tell.

Blatant self-promotion

If you’re ordering one of the new Kindles from Amazon’s U.S. store, I’d appreciate if you’d use one of my affiliate links:

• Kindle (the basic one with the buttons — I don’t think this is worth getting)
• Kindle Touch (probably best for reading books)
• Kindle Touch 3G (probably best for reading books if you travel)
• Kindle Fire (wildcard — too early to say, might be best for everything)

Thanks.