Which mobile phone carrier?

I have 4 lines with AT&T and am less than pleased. While data speeds are excellent, any call longer than 20 minutes will drop at least once. Coverage at my house is poor. AT&T’s voice quality is poor and has been for years.

My other major objection is their unlock policy. While AT&T will finally unlock iPhones after their contract is up, that still requires that I keep a 2-year old phone around for international travel. While exploring my options, I put together this TCO (total cost of ownership) comparison chart of the major US carriers (AT&T, Verizon, Sprint, T-Mobile) and a few resellers.

Mobile Phone Plans

The larger PDF also compares plans from Cricket, Virgin, and Ting. Pricing is shown for each plan with 1, 2, 3, or 4 phones.

Notes:

  • The One Time Net value assumes the value of an iPhone 5 in two years is $250. If you purchased an iPhone 4 two years ago, that’s about what your handset is worth today.
  • Ting with 3,000 minutes looks outlandish, compared to 1200-1,600 for everyone other plan. That’s because Ting and Virgin have no nights-and-weekends, or mobile-to-mobile. I looked at my past 12-months phone bills to determine how many minutes we’d need.
  • An excellent related post, Which iPhone 5 for a Global Traveller

Are charter schools good public policy?

From the paper How the worlds best performing schools come out on top by McKinsey:

Though the best charter schools demonstrated significant improvements in student outcomes were possible, and certain chains of charter schools showed that reliable models could consistently deliver improvements in a succession of schools, in the aggregate, the results of the charter schools did not significantly outperform those of other schools. The National Assessment of Educational Progress (NAEP) went so far as to suggest that students in charter schools slightly underperformed their counterparts in public schools, even after allowing for student background (Exhibit 4).

The most compelling argument against charter schools is that they don’t deliver better outcomes. Even after cherry picking the best (or as often, wealthiest) students out of the public school system.

They also measured the effect of simple fixes like throwing money at the problem (in the form of teachers) to reduce student teacher ratios.

Education spending versus student performance

Reduced student teacher ratios didn’t help. Other attempts at decentralizing educational policy ranged from ineffective to disastrous. So what does wok?

The highlights of the study are that, “To improve instruction, these high-performing school systems consistently do three things well:”

  • They get the right people to become teachers (the quality of an education system cannot exceed the quality of its teachers).
  • They develop those people into effective instructors (the only way to improve outcomes is to improve instruction).
  • They put in place systems and targeted support to ensure that every child is able to benefit from excellent instruction (the only way for the system to reach the highest performance is to raise the standard of every student).

This paper highlights findings I’ve read elsewhere–Finland and Singapore didn’t create the best educational systems in the world by segregating the students. They focused on improving the quality of instruction for every student. The difference between the best performing students in Finland and any OECD nation are comparable. The top systems are better because of how far up they bring their lowest performing students.

Our National Debt

This graph makes it easy to see what’s driving our increasing national debt.

causes of national debt

Being able to visualize the causes makes it easier to understand what a proper solution would be. Ending the Bush tax cuts as soon as economic conditions allow would have the biggest impact. Ending the wars is the next biggest fiscal gain.

Money Mischief, By Milton Friedman

In the first chapter Friedman compares using gold as the basis for money to using stone discs:

For a century and more, the civilized world regarded as a concrete manifestation of its wealth a metal dug from deep in the ground, refined at great labor, transported great distances, and buried again in elaborate vaults deep in the ground. Is this one practice really more rational than the other?

Then he spends all but the final chapter detailing the pain and suffering inflicted upon societies using silver or bi-metal standards. I felt like much of the data (some in works with no citations) was cherry picked or tortured to fit his conclusions. If his conclusions are sound, and the evidence is so clear, then why does he need to work so hard to prove them? I nearly gave up the book. But the the last chapter turned out to be quite good, containing this little pearl:

It is natural for individuals to generalize from their personal experience, to believe that what is true for them is true for the community. I believe that that confusion is at the bottom the most widely held economic fallacies-whether about money or as an example just discussed, or about other economic or social phenomena.

The first and last chapters of the book are good. I think that only economists that drink from the same Kool-aid pitcher as Milton Friedman will enjoy the rest of the book.

Adventures in legacy software

Today I opened M.Y.O.B. accounting files from 1999-2001. The files cannot be upgraded because they experienced some data corruption. The only way to access the files is to run the versions of M.Y.O.B. that they were created with.

The solution was to run Mac OS 10.6 inside a VM, which unfortunately, is a violation of the Apple EULA for Mac OS 10.6. Oops. The reason for using 10.6 is that it’s the last version of Mac OS X with support for Rosetta, the PowerPC emulator that lets Mac OS X apps written for PowerPC run on newer Macs with Intel CPUs.

Rosetta allows AccountEdge (the versions of MYOB written for Mac OS X) to run, but one of the files I wanted to access was from 1999, and it requires MYOB Accounting Plus v9, which is a classic application. For that, I turned to SheepShaver, a Mac OS Classic emulator, running Mac OS 9 within the 10.6 VM.

In summary, the solution was to run two emulators within a VM. With that, I’m able to run every version of MYOB within a single VM.

I also have a VM that runs AppleWorks, for those rare occasions when I stumble across a very old file that I would like to upgrade to the latest version of AppleWorks, so that I can further upgrade it to something that’ll run on Mac OS X Lion.

Also, since I couldn’t find one online, I also created a table of the MYOB versions, their marketing names, release numbers, and year of release.

Name Version Release DB ver. Year Company
MYOB 5 5.0.8 5 1994 Best! Ware
MYOB 6 6.0.1 1996 Best! Ware
MYOB 7 7.0.3 7 1996 Best! Ware
MYOB Plus 8 2.0.5 2 1999 Best! Ware
MYOB Accounting Plus 9 3.0.4 3.5 1999 MYOB LImited
AccountEdge 1 4.7.0 2001 MYOB Limited
AccountEdge 2 5.5.3 2001 MYOB Limited
AccountEdge 3 6.5.3 2002 MYOB Limited
AccountEdge 2004 4 7.5.0 2003 MYOB Technology Pty Ltd
AccountEdge 2008 8 12.0.6 2008
AccountEdge Pro 2012 12 16.1.4C 2012 Acclivity Group LLC

 

Mobile phones in Costa Rica

To use an iPhone in Costa Rica, one must have an unlocked phone. When we travelled in February, my iPhone 4S was not yet unlocked so I took a jailbroken and  unlocked iPhone 3. When we arrived, I dropped in a prepaid Kolbi SIM, set the data APN to ‘kolbi3g‘ and was off and running. I found the APN on the flexispy list. I used the TetherMe app to share the 3G connection with my iPad.

During the two weeks trip, I had only one issue when my phone stopped working. I just needed to reboot the phone and unlock the SIM. Besides an iPhone 3 being slow, everything worked flawlessly for the rest of the trip.

Notes to self:

  • do bring along the bluetooth iPad keyboard
  • February is the right time to go. The change from soggy grey Seattle days is welcome, and the 9 Seatte residents we met in C.R. agreed.

Scale is hard: gmail edition

Sometime on or about August 8th, IMAP access to Gmail stopped working for me. I don’t depend on Gmail so I ignored the error messages for 6 days. Then I simulated an IMAP session and was able to authenticate to Gmail, but after that, all IMAP commands timed out. So I looked to the Google Apps Dashboard which reported that everything was fine. Lies. Then I found the Gmail Known Issues page where I found this:

We are aware of an issue where users are receiving an error that ‘”imap.gmail.com” is not responding’ when using IMAP on their computers, mobile devices, or tablets. We are currently working on resolving this issue.

In the meantime, you can sign in to Gmail through a web browser. For Android and iOS users, you can also download the Gmail application in the interim.

Then I found the Gmail Forum within the Google Product Forums where many other users were reporting this issue. What I find interesting about this failure was that it took 7 days before my account access was restored and many users are still affected, 9 days later.

 

Mobile mapping isn’t good enough

Today I confronted a limitation of Google Maps. I’m in Buena Park and I need to be in downtown LA around 8AM. Last night, I consulted Google Maps on my iPhone for planning purposes. It answered 31 minutes. I also checked in a browser on my laptop, in case a feature existed that would let me choose a departure or arrival time that accounted for Monday morning’s typical traffic. There is no such feature.

Because of LA’s infamous traffic, I gave us 2 hours to assure we arrived on time. I also consulted Google Maps at 6:00AM when we departed. It offered up 38 min with traffic accounted for. I checked again after arrival and Google Maps predicts 55 minutes. The actual time: 77 minutes.

Perhaps the Google Maps time estimates are typically reasonable in typical cases and LA traffic is an edge case that is poorly handled (as I learned several times in the past week). However, I think this is a limitation of how Google calculates traffic times. It only accounts for traffic as it is at this moment. It doesn’t account for the typical traffic patterns that predictably recur on given days of the week.

This is where other competitors such as Waze and hopefully the new iOS maps app can offer a significant advantage by mining crowd sourced data to make much better predictions. This is almost certainly something we’ll see in a future version of Google Maps.