jorge

The hospitalization rate is slowly going down

February 6, 2021 by jorge

Although the hospitalization rate in Houston is still high, it’s pretty clear now that we passed a peak and it’s going down. This means my predictions were, fortunately, wrong.

The daily new positive cases are still pretty high.

What’s going on here? I wish I could tell you that people are being more careful, but my small, careful outings into the community reinforce my bleak view of human nature. We, as a whole, are not being particularly careful.

As a refresher, why do I care about/harp on hospitalizations so much? For two reasons: The first is because they are predictive of deaths, so less hospitalizations means less people will die. The second is because, after a threshold, they have compounding bad effects. If the hospitals get saturated enough, people die of milder cases because they can’t get help, and people with other conditions die as well. And the healthcare personnel get sick and/or burn out, further compounding the problem

What IS going on, then? It’s a combination of things. I credit two factors. Both of them are, or would be (if I got it right), genuinely good news.

Factor one, and this one isn’t a guess: we are relentlessly getting better at figuring out who is in danger of immediate serious complications^* or not, and admitting people who truly need it to the hospital, letting milder cases play out at home.

Factor two, an educated guess: the vaccines are having an effect already. The vaccination strategy is fractured and frustrating, leaving people who want them scrambling for shots. However, the vaccines do work, and we’ve -haphazardly, but correctly- prioritized the most vulnerable, and the healthcare personnel who are both necessary, and at risk of being vectors. Aside: under the same logic, please vaccinate the teachers NOW.

As of this writing, 11 doses of vaccine per 100 people have been distributed in the US. This doesn’t mean 11% of people are fully vaccinated, because the vaccines require two doses. Most people have one shot in them, which confers some protection, and a few already have both.

In Harris County, approximately 9% of the eligible population already has received one shot (click on the image below to access an interactive dashboard). If this was a randomly-distributed 9% of the population, the impact would be minimal; but it’s a selected group of vulnerable individuals, with the most likelihood of getting very sick or dying. This likely makes a big difference in hospitalization rates, so I expect them to continue falling.

(Note that the doses allocated/shipped above are for all of Texas, and the Harris County data is in the small popup)

The strategy of allocating vaccines to vulnerable people first poses an ethical question, which is should I get one if I’m not at immediate risk but somehow get on a list? I agree with this NY Times opinion piece: get it. On one hand, gauging your own risk is not that easy. On the other, getting more people vaccinated is a net good. On the gripping hand, there are no guarantees it’ll go to someone who needs it more than you do. Barring genuine counter-indications, we should all get vaccinated if at all possible. You’re arguing timing in marginalia.

Finally, I leave you with a counterargument to one of the dumbest antivaxx conspiracy theories.

Image result for you're not worth microchipping change my mind — The technology doesn’t exist, and you’re not that interesting anyway.

*By “immediate serious complications” I mean things that will kill you quickly. All available evidence on long-term complications (“long-haul COVID”) shows that they can be very, very bad, and we have -so far- no way of knowing who’ll have or not. If you prance about the community without a mask on and act carelessly, thinking the doctors will save you if necessary, you’re a selfish dumbass.

COVID-19 Houston update

January 23, 2021 by jorge

(source, as always: https://www.tmc.edu/coronavirus-updates/tmc-daily-new-covid-19-hospitalizations/)

We seem to have reached a plateau in terms of hospitalizations. I hope this is true, but I suspect it may be a mirage. I expected more of a spike after Christmas. Its absence leads me to wonder if we’re yet to see the full downstream effects of all that travel and all those family reunions. I hope I’m wrong, but I fear this is a momentary respite and we’ll start climbing again.

ICUs remain full but we still have overflow capacity, so if you need medical attention you’re likely to get it. This is much better than the alternative, but -as always- I want to remind you that being in an ICU gasping for air is a horrible, horrible experience that causes PTSD and years of nightmares. Avoid getting this disease. Wear your mask; wash your hands; keep social distancing.

Houston schools remain open, and remain an obvious transmission/dissemination route for the virus. See their own dashboard, and note the stark upwards trend in confirmed cases in HISD. There’s spikes, yes. But around the 2 in “Jan 2021”, the curve bends upwards.

This isn’t surprising. Keeping kids – with their triple-whammy of less self-discipline, and being packed indoors for hours, and a tendency to have less-symptomatic infections, makes for a simmering infectious mess.

Parents’ selfishness doesn’t help. My own kids’ elementary school recently saw a spike thanks to people who sent their symptomatic kids there every day for a week. Even the day they took them to get COVID-tested, and the next day while they awaited results. Bone-headed, at least, and deeply irresponsible and uncaring at worst. And -without any data to back it up- I suspect this problem repeats itself all over Houston, the state, and the country.

Schools are unlikely to be closed any time soon. They provide food security and a stable environment to many children, and enable adults to go to work. I may wish things were different, and I do, but it’s reality. There are no approved vaccines for children, either, so this will be a concern for a while.

Therefore,

Please get all adults in your house vaccinated as soon as possible. Don’t second-guess whether someone else would be better off, etc. Just do it.
Please be extremely careful, to avoid passing this virus to your children, and indirectly killing someone else’s parents or grandparents
Avoid all social situations unless they’re properly socially distanced and masked
And even then, avoid them if you can
Don’t go out to dinner, much less indoors
Work remotely if you are able to
Virtual-school your children if you are able to
Don’t fly places on vacation

We need everyone’s best efforts so we may all survive, and thrive despite the virus. Things will eventually go back to normal, but we can’t rush it.

I still dislike AT&T, but I use it again

January 1, 2021 by jorge

A long time ago, I abandoned AT&T for the crimson-tinted waters of Comcast. They offered faster service, and they didn’t directly try to hoist a fraudulent contract on me.

However, AT&T came around about a year ago with an offer a nerd can’t resist. Symmetrical, uncapped, 1 Gbps residential fiber-to-the-home. No strings attached. No requirements to use their video anything. Just sweet, sweet, fast data.

I called Comcast to ask if they could do anything beyond the meager 30 Mbps upload (or so; I forget) I had. They couldn’t. I switched.

The service, it pains me to say, has worked phenomenally well so far. It regularly speedtests at 930 mbps up and down. It’s been a lifesaver in these “work and learn from home” times.

Their mandatory home gateway doesn’t have a true bridge mode and is therefore hot garbage, but I can live with it. Other than that, I like it, and it works as advertised.

They even sent me a letter recently saying HBO Max was now included with the fiber service, so I ditched my independent subscription.

Haven’t had to talk to a rep, which makes me happy.

With HBO, taxes, fees, etc. and without asking for special discounts, I pay about $90/mo. I find it very well worth it.

The UK COVID variant

January 1, 2021 by jorge

Happy new year! The new COVID “British” variant (UK B117 strain) isn’t, of course, British. It was merely detected there first. See this thread by Dr. Gurdasani on Twitter, and the linked report from the Imperial College London: https://twitter.com/dgurdasani1/status/1344774555718590464

The TL;DR version is that this virus is
A. Apparently more transmissible in children, and
B. Has a higher R_t (the coefficient of transmission in real-time^*).

While the R_t for the OG COVID is 0.92 IN LOCKDOWN^** (i.e., on average, one person gives it to 0.92 other people, so the number of infected tends to go down, IN LOCKDOWN)… the UK B117 strain’s R_t is 1.45. IN LOCKDOWN. Which means exponential spread, and (given time), everyone gets sick. Even in lockdown (see footnote ^** again). Collapsed hospitals, etc. Which means we would need much, much, much more stringent measures to control it.

Before your eyes glaze over, this just means that people give the new variant to each other much more than the original one.

It has already been detected in Colorado, with no discernible connection to the UK. It’s a good bet that wherever you are, it’s also already there. We didn’t know how to look for it, or that we should. Now that we will start looking for it, we’ll probably find it everywhere. And since it’s more contagious, it may even become the dominant strain.

For the nerds in the audience, the mechanism for increased transmissibility is still unknown. There is a mutation in the spike protein (yes, the same one targeted by the vaccine) that makes it bind tighter to human cellular receptors. This is a reasonable hypothesis as to why it’s more contagious, but we don’t actually KNOW yet. We don’t know whether the vaccine is equally effective against this strain, either. The people who make the vaccine think it’ll be effective. How much, we don’t know yet.

Therefore:

Get vaccinated as soon as you can (I’m scheduled for Tuesday, thank my employer!)
Keep social distancing measures in place. Masks. Masks. Masks.- For the love of everything that is sane, avoid parties, enclosed bars and restaurants, gatherings, etc.
Yes, this means avoid “outdoors” fully enclosed spaces. That’s dumb as rocks.
In fact, avoid restaurants and bars other than carryout, period.
In case you don’t think I’m serious about the above: I’M AN INVESTOR IN TWO DIFFERENT SMALL RESTAURANTS. Sucks to be me.
Seriously consider switching the kids to distance learning. I KNOW this is a heartbreaking decision for a lot of families, and many have incredible difficulties re: work. Also, kids don’t learn as well. I know. I know. Do it anyway if you can.
The vaccines are not yet approved for use in children; they won’t be for weeks-to-months. Do you want your kids to catch a lethal disease, or one that will give them lifelong disabilities? Of course you don’t.
Do all of this even AFTER getting both doses of the vaccine. We need to know more before we let our guard down.
Yes, I’ve been fantasizing about traveling and going to dinner once properly vaccinated. No can do, for now.

* R₀ is an inherent property of the virus. R_tis the transmission coefficient in real-time, as observed. R_t can and does change over time. We don’t actually know R₀ for either variant, although there’s a proposed range for the OG COVID. It stands to reason that UKB117 R₀ > OG R₀, but we don’t know yet.

** The UK’s lockdown measures SUCK, because they include in-person school… and this variant seems to affect children more. Chicken or egg problem there: children are more exposed, so maybe that’s why they’ve been more affected? Hopefully actual distancing might help.

*** Originally posted on Facebook. The formatting was improved here.

**** Sorry, this came out a lot longer and ramblier than expected.

Relative Citation Metrics

February 12, 2020 by jorge

This is the first post in what I hope will be an ongoing series in scientometrics.

Today’s topic is an excellent paper -and idea- by Hutchins, Yuan, Anderson, and Santangelo called the Relative Citation Ratio (RCR).

The RCR is part of a new wave of citation metrics that aims (broadly speaking) to correct for the well-known shortcomings of the Journal Impact Factor (JIF). The idea behind the RCR, and proprietary equivalents like Elsevier’s Field-Weighted Citation Impact (FWCI), standardized via Snowball Metrics, is to adjust citation counts for the size of the field. In other words, publishing in JAMIA (impact factor of around 4.3 at the time of this writing) has a much-smaller built-in audience than publishing in JAMA (impact factor=51 and change). How do we account for audience size when measuring how impactful/important a paper is?

The answer that the RCR and FWCI propose is to measure the size of the field, and then tell us how much -or little- a paper is cited compared to the average of the field. In other words, if my paper is cited twice as much as other papers in the same field, my paper’s score is 2.0. With this in hand, I can compare the relative impact of papers across fields.

Even better, the “field” is computed dynamically, at least in the RCR’s case, by looking at papers co-cited with the one we’re focusing on. So you can assume that someone, somewhere, with expertise decided that these ideas were worth meshing together.

Overall, I think this is great, and much better than looking at “raw” citations. However, I also see weaknesses in it. They stem both from the methodology and from the general idea.

The latter is easier: If I publish in the New England Journal of Medicine and get 100 citations, it still means I published in the freaking NEJM and got 100 people to cite my work! Sure, it may be so-so compared to other NEJM articles, but it still influenced other people who then went and used my idea. I struggle with thinking that this is “the same” as a putative paper with 2 citations in the Journal of Obscure Armadillo Studies. In other words, using relative citation metrics we’ve found a way to compare apples and oranges “fairly”, but we may be comparing them by something akin to how much they weight; meaningful under some circumstances, but not really speaking to inherent value. If I have 300 grams of oranges, and you have 300 grams of apples, which fruit is better?

The methodological weaknesses are (IMO) from the lack of transparency. You can’t really compute RCR and similar measures without a large dataset, a citation network, and a computer. As such, it’s opaque, and works as an oracle – whatever it says, goes, and it’s hard to check the work. You just will have to trust it. This is especially painful because there’s very little publicly-available citation data, and most of it is behind expensive paywalls at Elsevier, Clarivate, or others. Further, just like any other citation-based metric, it will change as new citations appear… so yesterday’s paper can’t be judged until a couple of years have passed, and today’s RCR will be different tomorrow. We should timestamp these things.

With all that said, it’s a better impact metric than raw citations or JIF, and for looking at scientific productivity broadly -for the 30,000 foot perspective- it’s definitely better. The importance of removing the effect of field size can’t be underestimated. In essence, it’s got “context” built in, and for committees or executives looking at quality of scientific output across a program, department, or institution it should prove extremely useful.

For judging a single paper or author… there’s no replacement for careful, qualitative peer-review.

RCRs, and a free, open -albeit somewhat small- citation database, are available at the NLM’s new, excellent iCite site/tool.

OctoMac

February 9, 2016 by jorge

I like 3D printing. It’s an interesting, if fiddly, hobby. Quite fun as long as you don’t actually need to print anything, because things go wrong All. The. Time. The filament snaps and your printhead goes on, obliviously, for hours. Data communications break down. The hot end that melts the plastic clogs, and no more feed. There’s no end to the amounts of disaster that can happen.

Since 3D printing is SLOW and error-prone, monitoring it remotely is very desirable. Hence, Octoprint – brilliant open-source software that can control your printer and expose what’s going on through a web interface. A lot of people control their printers through Octoprint running on a Raspberry Pi, so much so that the default way to get Octoprint is as a Debian-based distribution for the little computer called “OctoPi”. I wanted to run Octoprint, but I had no Raspberry Pi.

What I did have, however, was a decommissioned Core2Duo iMac from 2006. It was awesome back then – 256 MB of VRAM, 3 GB of RAM. I was living the life, I tell you.

The latest Mac OS that will barely limp along on this thing is 10.7 (“Lion”), 5 versions ago (we’re on 10.11 “El Capitan” now), and it works poorly. While I could probably get Octoprint to run on Lion, the lack of security patches on a system that’s meant to go on the Internet gave me pause. So I decided to install a modern, useful, resource-light operating system on it: Linux Mint 17.3

This guide will teach you how to turn your old iMac into a useful Octoprint controller. Thanks to @vafarmboy for the idea of writing this up!

What you’ll need:

An old Intel Mac (some parts of this guide are iMac-specific) running some old version of OS X; mine used Lion, but anything 10.5 or greater should work.
A blank writable DVD
Making sure that you can afford to lose whatever information you have on that Mac.
That Mac should be somehow connected to the Internet.
A familiarity with system administration tasks on the Mac.
Read THE ENTIRE GUIDE BEFORE DOING ANYTHING.
Really, read the entire thing.
Chutzpah

First steps

Boot into OS X. You’ll need to stash away a few Apple files, and modify the partitioning scheme of your computer. Again, you might -and as far as I know, WILL- lose everything on it, so back it up if you care about the contents.

Obtaining the iSight firmware

You’ll probably want to use the iMac’s built-in iSight, if it has one, to monitor the progress of your prints. The iSight is internally connected to the Mac as a USB device, but it runs proprietary firmware. That firmware is not available anywhere BUT an existing installation of Mac OS X. Hence, grab the following file and stash it away on a USB drive or cloud service to use later:

/System/Library/Extensions/IOUSBFamily.kext\
/Contents/PlugIns/AppleUSBVideoSupport.kext\
/Contents/MacOS/AppleUSBVideoSupport

(note that that is a SINGLE path spec – one file that you need to grab, broken into three lines for readability.)

If everything goes according to plan, you should be able to mount the Mac partition from Linux and just copy the file… but you never know, so please do yourself a favor and back it up.

Download Linux Mint

I highly recommend a very lightweight window manager (we’re talking 10-year old hardware, here) and a 32-bit version of the OS. While my Core2Duo can run 64-bit Linux, your Mac may not be able to; and 32-bit software tends to use less RAM than 64-bit versions of the same. We won’t be doing anything that requires a huge address space. Remember that the original runs on a Raspberry Pi 🙂

So, please download Linux Mint, Xfce desktop, 32-bit edition. It is remarkably clean-looking, elegant, and performant, especially if you remember I’m running it on 10 year old hardware. It makes for a pretty good desktop.

Remember when Safari looked like this? At least it’s not brushed metal.

Create installation media

My old iMac can’t boot reliably from a USB device. You have to go old-school – burn a DVD of the thing. Disk Utility can do it for you, if you remember the Good Old Times™.

Yes, I had a blank DVD lying around.

I downloaded Linux Mint anew JUST FOR YOU.

All bow before the now-forgotten but formerly glorious SuperDrive

My “Super”Drive buzzes like crazy while burning a disk, but still works.

A few minutes later, we have a Linux Mint install DVD

Installing a boot manager

You’ll need something to allow you to choose which operating system to boot. Apple has a basic one that you access by holding down alt/option while booting. It doesn’t know about Linux, and is pretty dumb. We can do better. rEFInd is the gold standard for Intel Mac boot managers. It presents a friendly UI, and knows how to handle different Linux distros, Windows, and Mac OS.

rEFInd screen

So go to the page (I know, SourceForge), download the latest rEFInd binary (0.10.2 as of this writing) on your victim Mac, unzip it if your Mac doesn’t auto-do it for you, fire up a terminal, cd to the proper directory, and install it. Follow the onscreen prompts; rEFInd will try to elevate its own privileges via sudo.

cd ~/Downloads/refind-bin-0.10.2
./refind-install

Yes, my old iMac is called iCheese. iMac/iCheese – get it? Hoo boy, I kill myself sometimes.

Note that I did this on the Mac where I already have installed rEFInd and Linux, so it found the existing Linux partition and upgraded my old rEFInd install. The procedure for installing a new version is the same.

rEFInd is obsessively documented by its author. If you have any doubts or issues, go read the documentation. And donate a few bucks if you can swing it. He’s done an awesome job.

Linuxing

Reboot

Make sure your shiny Linux DVD is inserted in the SuperDrive, and reboot the system. rEFInd should come up, and it should see the inserted volume as a generic “whole disk volume” and offer it as a viable boot option.

The DVD is helpfully identified by a DVD icon. Your rEFInd screen will not look identical to this.

Since it’s booting a legacy volume, rEFInd assumes it’s booting Windows, and shows the wrong logo. It isn’t.

Much better! Yes, we’re actually booting Linux Mint.

A few minutes later (remember, we’re talking 10-year old hardware and reading from a DVD here), you’ll be presented with a nice Linux desktop.

The Linux Mint live environment

See the “Install Linux Mint” icon in the top left corner? Guess what we’re going to do.

Start the installer. It will want you to be connected to the Internet. My old iMac has a busted Ethernet port, and the live installer doesn’t know how to handle Apple’s wireless hardware. Too bad – I can’t slipstream updates and language packs into the install, but maybe you can; it doesn’t matter too much, as we will update the system later anyway.

Make VERY SURE you install alongside other operating systems if you want to keep you Mac partition!

I can’t guide you through the install beyond this, as my computer is already partitioned and running Mint. One of Mint’s quirks is that it wants to store data and operating system files on separate partitions, so you can wipe and reinstall the operating system on large upgrades without touching your data. LET IT DO THIS. You’ll thank me later.

See you on the other side.

Linux Mint configuration

Once you reboot after the installation, Linux Mint should be an option on your rEFInd menu. Boot into it.

rEFInd booting Linux Mint. It’ll take a few minutes.

Make sure your Linux Mint system is connected to the Internet, select a spiffy wallpaper if you want, and open a terminal. Do the classic “update system” Debian Linux incantation to make sure you’re on the latest version of everything.

sudo apt-get update && sudo apt-get dist-upgrade

Reboot, and grab the AppleUSBVideoSupport file I told you to stash earlier. Place it on your spiffy desktop. If everything went well, you could also copy it from your Mac partition (open the file manager, click on the name of your Mac partition to mount it, navigate to /System/Library/… as described above and copy the file to your desktop).

Open a terminal and install the isight firmware tools.

sudo apt-get install isight-firmware-tools

The package will want to know the location of your AppleUSBVideoSupport file. Aren’t you glad you saved it? It should be at

/home/[your username]/Desktop/AppleUSBVideoSupport

and your system should be able to extract the iSight firmware from it. It’ll tell you if it succeeds.

The AppleUSBVideoSupport file on my spiffy Mint desktop

It looks like a Java file. It isn’t.

You could now trash the file forever, but who knows when you’ll need Apple’s proprietary firmware again. Stash it someplace safe.

Octoprint

Octoprint is written in Python and runs on top of a fairly standard Python web software stack. However, it’ll need access to some hardware, and if you want it to run as a service instead of invoking it by hand every time, we’ll need to do a bit of work.

To get Octoprint itself, we’ll need git. To set it up as a robust web service, we’ll install it in its own isolated Python environment. To do that, we’ll need Python’s virtualenv module. To install Python packages within that environment, we’ll need pip. Some of OctoPrint’s dependencies need to be compiled against the Python development kit.

Let’s start by installing the necessary packages and their dependencies.

sudo apt-get install git python-dev python-virtualenv python-pip

Now we will create a Linux user that will “own” Octoprint.

sudo useradd octoprint

This user will host the Octoprint sourcecode, and a Python virtual environment containing the actually-running instance of Octoprint. This user will also need to be able to talk to serial (and USB) ports. Therefore, it needs to be added to some groups.

sudo usermod -a -G tty octoprint
sudo usermod -a -G dialout octoprint

Become the user octoprint and switch to its home directory.

sudo su octoprint
cd /home/octoprint

Clone the Octoprint git repository.

git clone https://github.com/foosel/OctoPrint.git

Create a Python virtual env that will hold the running OctoPrint server.

virtualenv server

At this point, /home/octoprint will contain two directories, OctoPrint and server.
The octoprint user's home folder

Enter the Python virtual environment to do the rest of the setup.

source ~/server/bin/activate

You’ll be rewarded with (server) pre-pended to your Bash prompt as a reminder that you’re operating inside the server virtual environment.

OctoPrint takes care of installing its own dependencies. We just need to execute the setup script. cd into the source directory and run it.

cd /home/octoprint/OctoPrint
python setup.py install

If everything goes well, the last line of output should read “Finished processing dependencies for OctoPrint…”. Otherwise, you’ll need to do some troubleshooting.

Great success

We’re not done yet with the installation, but we can run OctoPrint and check that it’s working at this moment. Try executing

octoprint

at the command line. Then open Firefox on your Linux machine and browse to http://localhost:5000

Even greater success

At this point, OctoPrint will work fine, but it still needs to be started by hand. Enjoy your success, but close Firefox, go to the terminal window and hit Control-C to stop the OctoPrint server. We’re not done yet.

Setting up the service

If you’re setting up a computer as a print server, you probably want the print server to start up when the computer boots.

Become your user with admin rights again (NOT the user octoprint), and copy the init scripts included with the Octoprint source code to their proper places. You also need to make the init script executable.

sudo cp /home/octoprint/OctoPrint/scripts/octoprint.init \
/etc/init.d/octoprint
sudo cp /home/octoprint/OctoPrint/scripts/octoprint.default \
/etc/default/octoprint
sudo chmod 755 /etc/init.d/octoprint

You’ll need to edit /etc/default/octoprint to reflect some of our configuration choices. In particular, make sure that the user and binary locations are set correctly. Use sudo nano /etc/default/octoprint to work on it. Note that the DAEMON line is commented originally. You need to uncomment it, so the new lines look like this:

OCTOPRINT_USER=octoprint
DAEMON=/home/octoprint/server/bin/octoprint

The updated OctoPrint defaults file

Finally, add the OctoPrint daemon to the system startup.

sudo update-rc.d octoprint defaults

Congratulations, you have a working OctoPrint server! Reboot and browse to your linux system’s ip address, port 5000 to see OctoPrint running in all its glory.

Octoprint is running at http://192.168.76.133:5000 on my home network. You address may vary.

Webcam setup

COMING LATER – the webcam works fine under Linux, but I haven’t been able to get it working with Octoprint itself.

It’s a watch.

May 24, 2015 by jorge

Yes, I got an Apple Watch. Sorry, Watch.

And people want to know. What is it? How is it? Is it any good? Is it worth the money? All reasonable questions. I’ll answer them as best I can. Once again, there’s already way-too-much written about the Watch all over the internet. I’m not going to rehash professional reviews, that do it much better than I ever could. I’m just here to give my opinion.

In that vein, I’m going to completely gloss over lots of functionality. I don’t have bluetooth headphones, for example, so I don’t care about the Watch’s ability to play music standalone. I’m never far enough from my phone that I will care about the Watch’s independence. I’m fine with it as a strict accessory to the phone. I don’t use Maps for directions (although the Maps on the wrist are kind of cute). This is strictly based on what I’ve actually used and done with the watch.

What is it?

It’s a watch.

No, seriously. It’s a watch.

In 2015, these things on the right aren’t “Smartphones”, or even “iPhones”. They are phones. This is what I think of when I think of “phone”, or say “let me grab my phone”. I don’t mean a rotary-dialed, black, heavy, wired monstrosity that my kids have never seen. Those are antiques. A phone, in this day and age, is a pocketable computer with a mostly-glass face, ubiquitously connected to the Internet, which happens to enable voice communications several different ways. VoLTE, VoIP (FaceTime, Skype, Google Hangouts, Facebook messenger, etc.), GSM calls, it doesn’t matter any more. It allows me to communicate with people, in a ton of different ways. One of them happens to be backwards-compatible with earlier wireless voice communication technology. It won’t even be around for a lot longer.

Phone numbers are weird and antiquated. Soon we’ll have to resort to explaining historical background to tell the young’uns WHY phones have numbers at all. “Well, it was like an IP address…”

If I could go back in time and tell tech-loving 20-year-old me in 1994 that I’d be carrying around a 64-bit pocket UNIX computer with a 1920×1080 color screen, based on the NeXT OS, permanently connected to the net, at speeds MUCH FASTER than the best Ethernet could accomplish then, with AN ENTIRE GIG of RAM, and a battery that lasts all day… I don’t know. I would’ve punched myself for lying, I guess. It is just too good. Then I’d tell 1994-me that we call these things “phones” nowadays, and act all cool like it’s no big deal. And 1994-me would probably try to steal my iPhone, morals be damned, and then go all sad because it couldn’t get a signal, and what the $#!^ is the strange port on the bottom, and WHY DO YOU TIME TRAVEL WITHOUT BRINGING THE CHARGER… But I digress.

The Watch -terrible name- is a watch, in the sense that the glass-and-aluminum slabs we carry around are phones. It’s just a circa-2015 attempt at making a watch, by the most successful tech company in the history of the planet (believe that, 1994-me!).

And it’s beautiful.

The pictures of the thing, even those meticulously-photographed or rendered 360º views on the Website, don’t do it a lick of justice. There, it looks thick, unwieldy, and the proportions seem a bit off. In person, those concerns disappear. It’s smaller than it looks in the pictures; or, at least, it’s the right size for my wrist, at 42 mm (the larger of the two versions).

Watch with our “beautiful” work carpet in the background.

The observant reader will notice that I got the stainless steel “black” model. I won’t go into details. It wasn’t a rational decision*, but I believe it’s the best-looking of the entire bunch, by far. Feel free to disagree. That’s the whole point of the large number of offerings and combinations.

Also, I’m not becoming an WristModel any time soon, not without a lot of shaving and trimming, and a shirt from somewhere fancier than CostCo.

I guess that the difference lies in looking at it from a normal “looking at my wrist” distance, versus the close-up pictures that a small product like this demands. You don’t look at it from a couple of centimeters away, normally. You look at it from 30-50 cm away. At that distance, it looks great.

How is it?

Well made. Insanely well made.  is making this manufacturing prowess an everyday thing, so we take it for granted. It’s even better made than the iPhone 6+, which is a ridiculous statement. The Stainless Steel Watch is heavy, much heavier than it looks, and has that dense “quality” feeling you get when closing the door on a large Mercedes-Benz. Thunk. It’s a bit clinical, but it exudes Quality. Like many other Designs, then, if they were made of surgical steel.

People complain that the software is complicated and unfriendly. I disagree. It takes a little bit of prodding around and poking on the UI, but it becomes second nature in a few hours. Now I just use it without thinking about it. Yes, there’s a couple of things that still aren’t 100% intuitive, and the extra button on the side could be used for something else; I don’t think I’ve ever pressed it to accomplish something other than turning the watch on. It doesn’t bother me that it’s there, being useless. I suspect it’ll come into its own when more people have Watches and communications becomes a more central function…. or it’ll be reassigned as a “home” (or possibly “Siri”) button in WatchOS v2.0 and we’ll be better off for it.

In that sense, it’s a typical Apple Rev A product. Wonderfully made, with some obvious glaring shortcomings that will be fixed in Rev B and make me look like a fool for buying a Rev A one. Oh well. Someone had to.

The real complaint, I think, is discoverability. The Watch is not immediately obvious. If you’re used to breaking ALL THE POTS in videogames to see what’s inside, or poking around a computer to figure out how to do things on it, you’ll have no problems with it. It’s easy. If you expect to be held by the hand and guided around, you might have trouble. Force Touch, where you touch harder to bring out different functionality, is only obvious because everyone’s written about it. It’s like a right mouse click. If you grew up with PCs with two or three-button mice, it’s obvious. If you’ve never used a mouse before, there’s no reason you’ll intuitively understand that the buttons do different things. The same applies here. You have to be willing to experiment. If you do, the Watch will become second nature in no time.

Ok, ok, but how do you USE it?

You look at it.

That’s most of the use cases. Twist your wrist to look at your watch, and the time’s there. Also a number of other small pieces of information, depending on how you’ve configured it. The outside temperature. Your activity level so far for the day. Your next appointment/meeting. How much battery it has left. The date. These are called “complications”, in a nod to traditional watchmaking. Sorry, I meant horological timepiece craftsmanship. The watch world is obnoxiously pompous.

When you get a notification, the watch will by default “tap” you on the wrist, sound a gentle tone, and you can… look at it. You can tap on the notification to dismiss it, or reply (if it’s a message) using canned replies or Siri. It works pretty well.

You can also set the notifications to display a general message (such as “Message from John”) instead of the actual content, for privacy reasons. Then you tap on the notification to see the contents. I’ve set it up this way. More taps, but less information leakage at inopportune moments.

As to the “taps” courtesy of the “taptic engine”… I think there’s something wrong either with me or my Timepiece. I feel them more like vibrations than taps. I’ll ask Support at some point.

You can also press the “Digital Crown” (if ’s guilty of anything, it’s too much deference to Horological Bouse De Vache) to bring you to the sorta-homescreen, with app icons. It’s easy to navigate around by swiping and tapping on apps. Here’s where you’ll also learn about First Class and Second Class Citizenship.

First Class Citizenship is bestowed by  and  alone, on its own apps, built using an unpublished and not-yet-available SDK. When you tap on an app, it opens immediately and works well. Second Class Citizens are apps written by unwashed third-party developers on ground not as hallowed as Cupertino’s. They are actually remote displays for code running on the phone.

When you tap on a third-party application, something like this happens:

Watch: “Hey, Phone, wake up! The user wants to run an app!”
[Time passes]
Phone: “I guess that could be ok. Which app this time?”
[Time passes]
Watch: “Dunno… Let’s say CalcBot.”
[Time passes]
Phone: “Ok. Let me look it up.”
[Time passes]
Phone: “I found something. It seems to have a bunch of code inside it meant for you. Let me execute it now.”
[Time passes]
Phone: “Hey, I got something for you! Would you mind displaying this?”
[Time passes]
Watch: “Huh? You talking to me? Display what? Oh, A SCREEN? For the user? We are still doing that? Ok, I guess.”

The look you’ll see the vast majority of the time you spend interacting with third-party apps

Have you tried to stare at your watch, which, annoyingly, has a pretty aggressive screen-off timeout to save battery, for 15 (“fast” apps) – 45 seconds (“slow” apps) while holding up your arm? It’s not pretty. This takes FOREVER. The screen turns off in the middle, so you have to flick your wrist back and forth. Or tap the watch. Several times. It’s foolish. It’s incomplete.

I could complain about it more, but let’s say third-party apps are unusable right now. Native apps are coming at some point in the future, much like on the original iPhone. Today, they might have been better off launching without them. It’s that bad.

Fortunately, the first-party apps are quite good, and enhance the Watch’s purpose of Being A 21st Century Watch. The usual Stalwarts are there: Stocks, Weather, Messages, Calendar, Mail, Music, plus a few Watch-specific ones. The central, and most obvious one, is… Watch. When in the home screen, press the Crown to center on Watch, then again to open it. In other words, no matter where you are, pressing the Crown enough times will get you to the Watchface, which is a perfectly sensible design.

That the watch is meant to run mostly as a Watch is also obvious when you notice that Glances, quick views accessible by swiping up from the bottom of the screen, only work inside the Watch app. These are somehow quick, and even the third party ones are quick…ish, so they sort-of-work.

The Digital Crown

A nicely-balanced scroll wheel for your wrist. Sometimes I wish it were just a tiny bit heavier. Sometimes I wish it clicked when turning. It works fine.

Unlike apparently everyone else in the world, I have 0 issues scrolling content with my finger on the screen. This was supposed to be A Big Deal. It isn’t. Maybe I’m too used to multitouch, or maybe it’s that the wheel doesn’t provide a complete interaction model. You’ll still have to tap, slide fingers, Force Touch, etc.

That said, in the few places where the Digital Crown is the only to interact (such as choosing complications, or selecting a color for the second hand on the Utility face), it makes perfect sense and works great.

I don’t know; maybe I was expecting The Second Coming of Engelbart, and the revelation of a completely new interaction model. The crown Just Works, which is both impressive and slightly boring.

The fitness tracker

The fitness tracker is pretty standard, as these things go, but on one hand improved by the great screen, and on the other hand too simplistic for true fitness buffs. For me, it’s more than I need and a lot better than my previous fitness tracker, a FitBit One. The One was quite good but I kept losing it. I had to buy it again. I own three now, and they are… somewhere.

The Watch’s main strength for me is being a LOT more in-your-face than a FitBit. It will tap you on the wrist once an hour and tell you to stand up and walk around, which I do. It will sit in a corner of your watch face, if you want, being slightly judgmental. Well, yes, you stood up… but have you EXERCISED? Have you? Even then, the default goals are modest: stand up at least once an hour 12 times a day; burn 500 calories “while moving” (even a sedate walking pace seems to qualify); and exercise 30 minutes a day, which is recommended but tougher for office-bound people like me.

My friend the elliptical has been lying to me.

The Watch knows your age, height, weight, kind of exercise you’re doing, how much you’re moving, and instant heart rate. The latter is sampled intermittently, unless you open the “Heart Rate” glance, which monitors it constantly. All these should be able to power a fairly detailed model of how many calories you’re burning. Which can lead to some very nasty discoveries… like that your favorite elliptical lies. Blatantly.

I suspect that the Watch lies, too, but since it knows more about you and the software is written by software engineers, I’ll assume its estimates are better until proven otherwise.

I can’t speak to the running trackers. I don’t run. The elliptical and “general standing-around-and-puttering” seem to work fine.

The UI for the tracker (captured from my iPhone below, but the Watch’s is very similar) has a brilliant piece of data visualization. Each band represents one of the goals: Outermost Red is Move, Middle Green is Exercise, and Inner Cyan is Stand. As you can see below, I exercised a lot yesterday, but not today. The bands wrap around visibly when you pass your goals, which I found surprisingly encouraging.

The Camera Remote

This turned out to be surprisingly good. It’s like a Selfie Stick, with two important improvements: 1. It doesn’t look douchey (i.e. you can set down the phone somewhere, no need to use the stupid stick), and 2. it lets you use the rear camera. The one that actually takes GOOD pictures. You stare at your watch screen, and can trigger a photo instantly or with a 3-second delay. The watch screen is not a very good framing viewfinder, due to its shape and size, but it does the job just fine.

The Watch Faces

You can use any Watch face you want, as long as it’s provided by Apple. For now. Maybe. Or maybe not. No one seems to know.

They are all beautiful in their own way, some weird, some very weird, some cute, and some utilitarian and very watch-y. I find myself drawn to the Chronograph face, which you see above. Yes, Mickey Mouse is there, and he’s perfectly animated, and taps his foot adorably every second, and is perhaps too cute.

You can customize the faces plenty. Add or remove numbers and marks; change the colors of certain elements; add, move, or remove complications in different parts of the UI. Some of the faces do things. Mickey taps his foot; the Astronomy face can zoom in and out of the solar system and show you the state of astronomical bodies at different points in time; the Motion face will show you a pulsing, glowing jellyfish swimming in a meaningless void (I did say weird, didn’t I)?

To take some of the sting out of the lack of third-party faces, you can save faces customized the way you want them. So you can have two different Chronographs, for example, one with a white background, exercise tracker, battery capacity, and a different one with a blue background, world clock, upcoming appointments, and weather.

There’s enough choice there to personalize the Watch pretty deeply, within the Sandbox. However, If you don’t like any of the Defaults, or don’t think you can tune one to your liking, this product isn’t for you. Much like on a 19th century mechanical watch, you can’t change the face[s] it comes with.

Some reports claim that third-party complications may be coming. That would be great. I’d love third-party watch faces, but I’m not going to hold my breath waiting for them.

Is it any good?

Yes, it’s quite good, if you take it at face value. It’s a watch with enhanced functionality. It’s a luxury product first, I suspect, which just so happens to be able to do more by talking to your iPhone. Infuriating shortcomings? Third-party app performance, which is unacceptable. Other things that fit under “is it any good?”

Battery life

Yes, you’ll charge it every night. That said, I’ve made it to the end of the day with between 20 and 50% charge left, depending on usage. I’ve not ran out yet. I don’t think nightly charging is a big deal. Every night, I plug my iPhone into its charger. Now, next to it, I put the watch on its magnetic dongle. Takes an extra five seconds.

Fit and finish

As I said above, insanely well made. The materials feel top-notch, and it looks built to ridiculously tight tolerances. The links in the bracelet are all IDENTICAL. The latch mechanism locks into place solidly.

Originality

I’m of two minds on this one. It’s a product made by the largest consumer electronics company that has ever existed. It’s expensive. As such, it gets -100 coolness points immediately. Pebble did it first. I wore a Kickstarter Edition Pebble for a while, until the display died and its band cracked. Android Wear did it before, too, and does some things better. Hell, Microsoft and Timex made a smartwatch of sorts. In 1994.

That said, there are some much-welcome innovations to Watchland in the Watch. The changeable bands – so easy even a klutz like me can do it, and they lock into place seamlessly. I got a sports band, for sportsing. It’s “cheap” (meaning it costs $49, which will buy several perfectly serviceable digital watches), but it feels great.

For sportsing!

The Stainless Steel link bracelet – you can add and remove links using only your fingers, but they stay solidly in place. It’s brilliant. It comes with a little velvety pouch to store your unused links. It feels like a true luxury product. Which it is, so it’s good that it fulfills that part of the bargain.

Usefulness

I look less at my phone. Sometimes a lot less. And I’m not fiddling with the Watch constantly, either. I’m just not suffering from missed notification anxiety as much.

Siri on the watch is surprisingly useful, and has picked up a couple of tricks – it auto-detected that I text my wife in Spanish, and expects Spanish there, but English when texting my coworkers. Very nice, and something sorely lacking from the phone implementation if you’re multilingual. This, and this alone, has made it tremendously useful for me. I can use Siri now. I really couldn’t before.

It’s a narrow use-case, but it’s mine.

Otherwise, the standard smartwatch advantages apply. I can glance discreetly at my wrist and keep up with what’s going on. I can even send some replies. It reduces notification anxiety. For that, you have to customize which notifications you want on the watch. It’s a little laborious, but it’s a one-time thing. I don’t want to be notified of every email I get. I don’t want most notifications the phone produces, in fact.

Filtering notifications down has the nice secondary effect that the watch becomes your “important notification” place. And, as such, the signal-to-noise ratio is better than the phone’s. Yes, you can tweak the phone’s notification settings too, but doing it this way lets you have a two-tiered notification system. Important stuff -> wrist. Less important stuff -> phone. Less important even -> no notification.

Is it worth the money?

No, and yes.

The Watch doesn’t exactly sit at the bottom of Maslow’s Hierarchy of Needs. In other words, don’t forgo anything you need to get an Watch. It is not going to improve your life, except incrementally, and even then only if you already have a nice life. A computer or a smartphone improve your life tremendously, giving you access to gobs of information and entertainment on demand. The Watch? No. It’s a “nice to have”, definitely not a “need to have”.

Do I need the Watch? No. Does anyone? I can’t imagine someone who NEEDS one, but maybe I lack imagination.

That said, if your retirement savings are growing, you’re not forgoing things in life you enjoy, you’re not getting in debt to buy the thing, and you enjoy the newest and shiniest gadget (I certainly do!), then go for it.

If you’ve been ogling nice watches for a while, which I have, but never committed to buying one, the Watch is a different take on the genre. A wonderfully-built, slightly clinical take; it’s missing a bit of warmth. But it’s a watch that makes a lot more sense to me, in 2015, than Tag Heuer et al. It’s a luxury watch for the 21st century.

By the way, let’s not pretend the Aluminum Watch Sport is cheap, either. Cheap-er than the Stainless Steel or Gold ones, yes. Cheap? Not at all, not when the Moto 360 is cheaper and looks great, or you can get a beautiful Seiko or Citizen for less.

In the end. it is a very nice gadget.  is probably the only company in the world that’s earned so much of my trust I’m willing to get up at 2 AM to preorder, sight-unseen, an expensive new gadget I’ve never touched. There’s a reason for that trust; the Watch exceeded my expectations in quality and beauty.

I’m probably keeping mine. I like it. I’m just under no illusions that it’s necessary. Not yet. Maybe not ever.

*At least I wasn’t crazy enough to get a gold “Watch Edition” (Worst. Name. Ever.), so I’ve got that goin’ for me. Which is nice.

Lies, damn lies, and AT&T reps

December 5, 2014 by jorge

This comment by a FB friend on my post about leaving AT&T made me think. That wasn’t the first time I heard lies from AT&T reps, especially retention reps. If I have to guess, the basic explanation of why this happens is pretty obvious. (Warning: complete speculation forthcoming).

Retention reps are paid to retain customers.

They are probably given at-risk pay, as in “minimum wage plus bonuses to make a less-bad salary if you retain X customers/time period” and/or are fired if they don’t meet a certain quota.

They are probably also given some kind of “ethics training”, (at a guess) online and using a slide deck you click through in terminal boredom that says “lying is bad, mmmkay”. They can probably be fired if someone reviews the call, notices the lie, and… if the lie doesn’t result in a retention.

Because a blatant lie that results in a retention is unlikely to be reviewed. The customer’s happy, the rep’s happy, and the company’s happy. Who’s wasting time reviewing that? Even if it is, it’s probably given a lot of parsing/wiggle room.

One of geekdom’s favorite Futurama lines

The most spectacular lie from an AT&T rep ever was when I tried to cancel my UVerse service. I liked it, but Comcast offered me faster service for less money (I’m noticing a trend here). FWIW, Comcast’s reps suck, too, but at least the service has been solid.

Anyway. I called AT&T to cancel Uverse. Got predictably shunted to retentions. Explained the situation. The rep was understanding, sympathetic, and we agreed there was nothing they could offer that would make me stay. Account closed.

Then I got an email stating that they were giving me a promotional rate in return for agreeing to a two-year contract extension with an ETF.

I don’t think I’ve ever been that angry at a single email.

Fortunately, the rep that took my very irate next call canceled the account immediately, and there was no mention of an ETF. No one ever apologized for the lies, though.

AT&T doesn’t get it

December 5, 2014 by jorge

I’ve been a pretty happy AT&T customer for many years. They’ve never been the cheapest (some MVNO, as of this writing), they’ve never been famous for their network (Verizon, probably), but for me they worked reliably day after day after day. Billing was arcane, but reasonably predictable. As long as I didn’t have to meddle with the account. Which I did. But I’m getting ahead of myself.

I could get an AT&T signal pretty much anywhere, except the Washington DC metro, and speeds were fast, and in general things worked.

However, every time we visited family in Chile I had to either pre-pay $30 for 300 MB of international data (which I can consume in minutes), $60 for a gig (still little, too much money), or borrow/bum/buy a local SIM, which is a hassle. A local SIM gives me a local number for a couple of weeks, and renders me unable to use my USA one. It’s therefore suboptimal, because I want to be able to get my texts and calls. So when T-Mobile started offering free international roaming (it’s not LTE, but apparently it works fine), I started paying attention.

Changing carriers is a hassle of course: hours on the phone; paying two contracts for at least a few days, if you want to keep your all-important number; perhaps, paying ETFs. Your phones may not work on another carrier’s network.

I don’t travel internationally that much, I told myself. Between 1 and 3 times a year. I can deal with the SIM swaps.

Then I discovered that T-Mobile had an agreement with my employer that would give me a discount on every line. AT&T just gives me a discount on the main line.

Then I noticed that my contract with AT&T had expired and there were no ETFs to be concerned about.

Then I discovered that T-Mobile has an actual, for-real, unlimited LTE data plan, and I started to salivate. There is one caveat: you can’t use over 5 Gb of tethering a month, which isn’t a problem for me. It sounds fair, because the T-Mobile price is decent and it’s an obvious “Please don’t replace your home internet connection with our cheap unlimited data” move.

So I decided to try and spent 15 min on the phone, and a few days later I had three T-Mobile SIMs in front of me. I own my phones and they are unlocked, so I simply swapped my iPhone 6 Plus‘ card and I had a T-Mobile phone. Of course, it worked. That’s a given.

Of the three SIMs, one’s unlimited; the second, for my wife, with 3 GB/month. She’s never used that much. If she goes over that… the connection slows down, but it doesn’t reach into our wallet. The third, for my dad’s sporadic visits to the US, 1 GB/month. Should be enough, but if he needs more, it’ll work. Just more slowly.

All of this costs less than my current plan on AT&T.

I quickly noticed that T-Mobile’s signal wasn’t as pervasive as AT&T’s. Much has been written about this, too. The short story is that T-Mob’s signal can’t penetrate structures as well as AT&T’s. Armed with the imperfect meter that is “the signal strength dots on the iPhone”, I walked and drove around and noticed that it fluctuated a lot more than AT&T. Notably, in my office I get 0 or 1 “bars” of T-Mobile service, and 5 of AT&T. This doesn’t matter as much, because of T-Mob’s WiFi calling. Basically, it routes calls over the Internet when connected to WiFi. It just works™, even on our corporate network.

When the signal’s good, the network is fast. This was only with 4/5 bars. Houston is a premier market for T-Mobile, and they have their best LTE here. It shows.

This is basically as good as I’ve ever measured on AT&T.

Even when you don’t hit their full signal, the network is fast enough for normal use: downloading email, web browsing, the all-important social networks, etc.

Did I mention that I get all of this for less money than AT&T charged me?

So, T-Mobile SIMs in hand, I call AT&T to see if they are willing to match some of T-Mobile’s terms. Not all; I didn’t think they would, but they do have a better network, and that’s worth something.

And that’s when I ran into a tremendously interesting phenomenon. AT&T has absolutely no clue of where things are headed. Zero. Zip. Zilch. Nada.

I talked to a very pleasant customer service representative. He lied only a little (there was no such thing as an unlimited LTE plan on T-Mobile, he said, despite the fact that it’s on their webpage, I bought it, and I have a SIM labeled as such). I’m not saying he accused me of lying, just of being misinformed.

I guess the “of unlimited high-speed” part is a hallucination, then.

I owed them money for the iPhone 6+ on installments, he said. No, I don’t; I bought it on AT&T Next, true, because they wouldn’t sell it retail at launch. I paid it off entirely shortly thereafter and promptly unlocked it.

No matter, a few lies won’t deter me. Is AT&T willing to throw me an international data bone remotely comparable to T-Mobile’s ($0 for unlimited, albeit slow 3G roaming)? No. I can pay $30/month for a ludicrous 300 MB, or $∞/kb for pay-as-you-go roaming. I may talk to the International people and see if they offer me a better rate, though. Why the ^&%$# do I have to do that, AT&T? Waste my time going from department to department, begging for a handout?

Are they willing to give me a larger data allowance for less money to come closer to the “unlimited” line I care about?

No.

What I didn’t expect, and what shocked me, was that they tried to persuade me to lower my bill by switching to a lower data plan. “You only use 3 GB/month on average“, the rep said. You don’t need a big plan. I can lower your bill by lowering your data allowance.

Silly, silly rep. I’m fortunate and privileged enough. I don’t care about $10 a month for the plan. I also know statistics. An average is a bad way of looking at this. I use a lot of data when I travel nationally, and I don’t want to pay $10/gig in overages. Because I’ve been known to download Linux distros over the phone.

I care about the game-changing potential of unlimited data everywhere.

I want to give the phone to my kids on a car trip and let them use Netflix for hours on end. I want to stream music, and Hulu+, and Amazon video, and one day HBO, as much as I want. I want to download apps with impunity. I want to FaceTime/Skype/Hangout/whatever is fashionable this month until I’m blue in the face.

The reason, clueless AT&T, I use little data is because you charge me insane amounts of money if I go over the line. So I keep myself in check, and use as little as I can by default instead of as much as I buy. I don’t have the time, patience, or inclination to micromanage data use. I’ve got a lot more interesting things to do with my time. I’m therefore unlikely to stream Netflix regularly, or to really put the phone and your nice network through its paces.

I guess I could buy a 200 Gb/month plan at… (checking website) no, oops, you don’t offer one. 100 gb seems to be the top tier, at $375/month as of this writing. I may be somewhat price insensitive, but I’m not an idiot, either. I’m not paying that. Paying AT&T’s overage charges if my plan is too small, on the other hand, is stupidly expensive too. ($10 or 15/gig? Why am I even required to know this??)

I want to do all these things without managing my usage. I probably won’t use that much data, for now. A few gigabytes a month. But maybe one month I’ll shoot a full-length movie on the phone and want to upload it in HD to Youtube, and guess what? I CAN NOW.

T-Mobile seems to be embracing being a “dumb pipe”, and by doing that they got:

A new customer with pristine credit, three lines, and that sets the bill on auto-pay.
Me to buy their highest tier of service… which I’d never do on AT&T.
For me to feel that I can finally use my phone to the fullest, without counting bytes here and there. That makes me happy.
You know what I do with companies that make me happy? I stick with them. See: Apple.

It’s an amazing and utterly obvious idea. Here, we’ll give you the connectivity you need to really use the amazing tool that a modern smartphone is. No, you don’t need to meter your usage. Just go. Enjoy. We’ll give you a connection; you can just use it. In most of the world. We’ll never try to bankrupt you. That’s a game-changing promise. And it came with a smaller bill.

So today I called T-Mobile again and ported my AT&T numbers over, ending a several-years-long relationship. I’ll miss the five-bars-in-most-places rock-solid network.

I won’t miss the lying reps, the nickel-and-diming, or the byte-counting. I definitely won’t miss those.

The iPhone Six Plus

September 29, 2014 by jorge

If you want a formal, professional, detailed review of Apple’s new shiny just head to my favorite tech haunt, Ars Technica. This is a personal account.

A long, long time ago my wife dragged me to the Apple Store to check out the new, just released iPhone. On the day it was released. It was a madhouse.

Until that moment, I didn’t understand the hype around the iPhone. I thought it would never amount to much. After all, the Treos and Windows CE did everything (and more) Apple promised.

That’s why I’m not a pundit. I’ve been wrong before, but this one was a doozy.

The next day, we bought two iPhones. Since then, I’ve owned every single iteration of the device, plus a few Android phones in between. And when the 6 and 6+ were announced, I ordered a 6+. A happy stint owning a Galaxy S4 taught me that I like big screens (…and I can not lie), and I like high-DPI screens.

As a photography aficionado, the optical image stabilization on the 6+’s camera was the icing on the cake. I couldn’t not have it.

Let’s get this out of the way first: the thing is HUGE.

Yes, much larger than an iPhone 5/5s, and plenty larger than my old Galaxy S4.

How large? Here’s how an iPhone 5 swims in a 6+’s silicone case.

Many critics have said that Steve Jobs would’ve never allowed this monstrosity to be built, and they are probably right. But as much as I admired Steve’s drive, passion, and genius, it’s time to move on, and Apple seems to be saying the same thing. It’s not Steve’s Apple any more.

Steve’s Apple is still alive and well, though, in other senses. Fit and finish is better than ever before. The thing’s built beautifully. No, I haven’t bent it. Yet.

In the hand, the Six Plus feels BIG. I can hold it comfortably, but not-quite-use it with the hand holding it, except for a couple of taps. I have big hands, but my fingers are stubby. As the gringos say, Your Mileage May Vary. Except for a couple of icons at the bottom of the screen, one-handed use for me is Right Out.

That’s a heavy price to pay, as iOS is clearly designed to be used one-handed. The Back buttons at the top left in most apps assume you can reach them. iOS’s launcher fills in its screens religiously from the top left. You can’t skip spaces, and you can’t place stuff arbitrarily, the way most Android launchers let you. This means you can’t put your most needed apps at the bottom of the screen, within reach of your thumb.

Apple has addressed this with a half-assed compromise called “Reachability”. Much virtual ink has been spilled on this feature. Tap (don’t press) the Home button twice, and the screen slides down until (on the +) about only 40% is showing.

Suffice it to say, my stubby right thumb STILL can’t reach the top left corner of the screen with Reachability, not without physically painful contortions.

So one-handed use, at least for me, has been sacrificed on the altar of the 5.5″ screen.

It is a pretty big sacrifice. It means no whipping out the phone if you are carrying something in one hand. It means no checking the weather absentmindedly. When you’re using the 6+, it wants your attention.

This big sacrifice, however, is paid back. The screen itself is fantastic: tack-sharp, colorful, and it finally seems to be Right There on the surface of the phone. See my ZTE Open musings for the exact opposite.

The ZTE Open and its 1.5 kilometers between “glass” and display.

I won’t bore you with technical details…

Ok, that’s a lie. I love those technical details. I rejoice in technical details.

Apple did something weird with the 6+, probably for power and cost reasons. The phone reports that its screen is 2208 x 1242 pixels, and that’s what screenshots show.

Click for full-size version

The physical screen itself, however, is 1920 x 1080 pixels. The phone renders the image internally at 2208 x 1242, and then downsamples it on the fly to 1920 x 1080. You’d think this is a bad idea. I thought it was a bad idea. Jaggies. Fuzzy text. We KNOW that screens need to be pixel-perfect to look good.

In practice, it looks perfect. The physical resolution is so high, that your eyes can’t make out the details and rendering tricks. Mine can’t. This isn’t iPhone 4/5 or iPad Retina Display quality, where you can sort-of make out the pixels if you squint or hold it close to your face. This display has no discernible pixels, full stop.

If knowing that you’re being fed fake downsampled tricksy pixels through your eyeballs bothers you on some fundamental level, I sympathize. It would’ve bothered me 10 years ago. I’ve moved on. The screen looks fantastic.

And that screen looking fantastic is one reason sacrificing one-handedness is not bad. I read on my phone. I read A LOT. Kindle books. Email. Web forums. Email. News. Email. Reviews of things. Email. Text messages. Email. More screen is better for reading all day. There’s no question about it. Samsung taught me that.

More screen is also photographic bliss. Yes, holding the phone in “camera” position is weirder due to the larger size, rounded edges, and power button just across Volume + (or “shutter” as many of us think). The price, however, gets you easier framing, better reviewing, and more detail. I’ll take the tradeoff. Taking pics with the 6+ just feels better. More on the camera in a bit.

The final thing regarding one-handed use is that you lose it… but you gain much, much, much better two-handed use. I can fly tapping out emails two-thumbed on this thing. The larger screen targets and easier hold with two hands make it much better.

Much virtual ink has also been spilled around the 6 and 6+’s cameras. While the all-time best camera on a phone crown probably goes to some Nokia model, this is clearly the best camera Apple’s ever put on a phone.

For a beautiful, artistic take on the improvements go watch Austin Mann’s video review.

The Optical Image Stabilization? It just works. I haven’t been able to get the flash to fire in auto mode at all (I checked it. It works). The camera seems to always get enough light. The dual-phase contrast detection “Focus Pixels” make focusing instantaneous. Noise is low, focus is quick, composing and reviewing are a delight.

Kids looking at penguins at the Moody Gardens aquarium – taken with the 6+

They say the best camera is the one you have with you. The iPhone cameras have been getting better and better – this is just another evolutionary step, but more than ever before it makes me want to leave the SLR at home. Apple, how about RAW support some day?

For now, I’m keeping the 5.5″ monster. Big, bold, and beautiful. Here’s to the new Apple.