Mesh WIFI networking is AWESOME... bullshit

I was on Ars Technica today reading about heavyweight mesh WIFI networking gear that promises better WIFI in your home.

The article makes some excellent points against these products, but it hides the failings deep.  I guess writers don't want to knock their advertisers.  Here I point out Ars' hidden arguments against these mesh networking products.

1. The Ars Technica article says, "like I discovered when I tested Eero for the Wirecutter, multi-hop here works, but it’s best avoided if you don’t absolutely need it". Stated more simply: avoid Mesh networking unless you absolutely need it or don't care much about performance.

2. The article says "some client devices actually perform worse when overwhelmed with too much signal."  Actually, all client devices perform worse when overwhelmed with too much signal.   Unsurprisingly, noise in the radio spectrum is a serious problem when dealing with radio communications.  The more powerful transmitters you have hanging around, the more noise devices will experience.  And that means more retries and slower performance.

3. The article says that that all these mesh routers are "still a lot more expensive than a standalone router, and most people's homes can be served fine from a single, good standalone router"

Furthermore, it is clear that Ars failed to led its readers through many other limitations of mesh networking:
  • Better placement of their atrociously-placed "base unit" would likely permanently solve all of their coverage issues using a single traditional router.
  • Adding nodes is going to lower overall network reliability.
  • Adding nodes is going to increase inter-node traffic, thereby increasing radio noise and lowering network performance.
  • Adding nodes is going to increase network installation, administration, and maintenance efforts. 
  • Adding nodes is going to increase cost.
I've installed WIFI access points in a number of 1, 2, 3 and 4  level homes that were up to 4,000 square feet in size, and with over 30 WIFI clients.  A single, properly placed mid-tier WIFI access point easily covered each property with great performance and reliability.  I have never had the need to install more than one access point on a residential property.

WIFI problems are almost always due to:
  • poor placement of the base station
  • poor (outdated or 3rd rate) equipment
  • poor channel selection
Mesh networking has its place:  Maybe you like to play with WIFI.  Maybe you live in a very large steel and concrete house and proper placement of gear is impossible.  Or maybe your house has numerous wings and outbuildings, but you can't afford proper wiring and would like to pursue a passable, less reliable solution.  Then, just maybe, a mesh network might be an reasonable option.

But the for the vast majority of home users, a mesh networking product is a complete waste of money and time.  Save your money and save the headaches.  Mesh networking products are bullshit.


On the failure of democracy

Democracy cannot succeed if people and politicians set up government policies along the idea of majority rule.  Instead, you have to go back to the guts of it all: happiness for the people.

With a concept like Brexit, where roughly half the population of the UK want to leave the EU, and the other half do not, compromise is in order.  Brexit shouldn't mean Brexit, and to run with that conclusion is to blindly follow a nascent and unrefined buzzword.   Is Brexit going to hurt the UK?  Then just maybe you need to REFINE the theme.

"Oh, but a majority voted for Brexit".  Bullshit!  They voted for all different things, and for all different reasons.

Brexit SHOULD mean fucking fixing the problem so that WAY MORE than 50% of the population is better off, and the few others should at least not be harmed.

That's what democracy is SUPPOSED to be about: compromise, make the imperfect world a little better, and repeat forever.

The US has the same problem.  The gun lobby can't even begin to compromise on obvious ideas like "risky unstable nuts shouldn't be given guns".  Politicians, who are real humans, actually support this idiotic position despite the realities of a complex world and a well-oiled legal system.  Instead of these politicians being fucking idiots, why don't they work to fix the fucking problem in a way that is achievable and makes reasonable sense?  Oh no, that's too hard?  You're unwilling to compromise?  You're a politician, you idiot!  It is what you signed up for!


Tracking Spending - and a Strategy

I don't like spending good money on stupid crap.  And so I have a new strategy that I've developed to help me keep it in control:

Basically, I keep track of my monthly recurring expenses across a ton of categories, such as:
  • Cable TV
  • Internet
  • Insurance
  • Electricity
  • Netflix
  • Mortgage
  • etc. 
If prices go up, I make a note of it.  If prices go down, I make a note of it too.  My goals?
  • Financially respond to every price increase
  • Reduce at least one recurring cost every month
So earlier this month, I knocked down my cell phone costs from $70 to $40 a month (win!).  I also eliminated my iTunes Music subscription, which I didn't use much, from $15 a month to $0.  Another win!

Earlier this year, I changed from my old employers healthcare plan (a "COBRA" plan) to a healthcare market place plan, saving me $260 a month (!).  And some people complain about Obamacare.

I also updated my auto/property insurance, which increased by $14 a month, but which gives me about 5x the coverage.

So with these changes, I'm saving $3500 a year.  But there is still a lot more to do!


My Ugly, Old MacBook

I have an ugly MacBook.  I like it.

I bought my current MacBook for a mere $40, cleaned it up, and then upgraded its RAM and stuck in a 1 TB SSD drive.

But it's still ugly.

The prior owner sold it to me as broken.  He was a heavy smoker of something, and the machine was a disgusting disgrace.  It had scrapes and stickers on the back of the display and on the case.  Everything was covered in a thick grime. And the keyboard smelled bad.

The only things to repair was the magsafe connector and the battery, and while doing that I cleaned out a lot of the disgusting. The tough sticker adhesive, burn marks, and scrapes are still there, but I put a cheesy case on it to cover up most of that. And it has a dead (always red) pixel in the middle of the display.

But it's a good computer. It is reliable, and it's reasonably fast. The whole working setup cost me about $300, but that includes the 1 TB SSD which was more than half of that cost.

Why dump it? A new, non-upgradable low-end MacBook Pro, with 1/4 of the storage, would cost me more than four times more money. I plan to keep this going with this beauty for at least 2 more years.


New Car versus Old Car

As I said before, I'm a car guy.  I love cars.  They are awesome fun, and many are beautiful works of art and engineering.

My car is beautiful.  My car is reliable, and it costs very little to own it.  I'm going to keep it that way.

Many years ago I bought a new car.  I put 168,244 miles on it, and then I sold it to a friend.  Then I bought a used car.  I kept it for four years.  Then I bought another used car.  I still have that one, 17 years later.  And I have no plans to replace it.

Some day my old car may suffer a catastrophic failure, or get wrecked.  It may get to the point where repair is an economic impossibility.  Then I'll replace it.  Until then, it's my car.  And I'm keeping it.

Yes, I do have to have my car maintained, and that costs real money.  Hoses, belts, brakes, tires, alternators, water pumps, and so-on can cost $1000 a year.  But the fact is that $1000 a year is a hell of a lot less money than the cost of a $30,000+ car.

Furthermore, I'd like the majority of my $1000 of maintenance costs to go to the local mechanic, versus spending $30,000 that mostly goes to an overseas conglomerate and its sales team.  Even so-called "American" cars are 80%+ built elsewhere.  And that doesn't even consider the car loan interest, which amounts to many thousands more used exclusively to line the pockets of billion dollar banks.

So why do people buy new cars?  It isn't reliability - a good properly maintained car is just as reliable as a new car, and is much cheaper to own.

I guess people buy new cars because they want something fancier to drive around.  Maybe they want new features, or want a new hip style.  There's nothing wrong with that.  I don't need it.  And I'd rather have my money go to something good.

The Smartphone Industry Isn't Green.

Google and Apple.  They're doing us wrong.  They don't build things to last.

Let me take that back.  They -do- build things to last.  But they withdraw security patching from their products way too early, making older devices a literal danger to use.

I love my iPhone 4.  But it is stuck at iOS 7, and who knows if it has dangerous vulnerabilities.  It is unclear if Apple is watching over iOS7, and it is a safe bet that Apple will never release a security patch for their older devices even if dangerous vulnerabilities are known to exist.

Don't get me wrong: I think Apple is right to leave the iPhone 4 at iOS 7.  After all, the iPhone 4 was designed and developed long ago.  Technology has moved forward, and no one in their right mind can expect an old device to run the latest and greatest operating system.

But I'm not talking about that.  I'm talking about security patches for over a million useful, active devices that are "stuck" at iOS7.  The security of a million+ people are at risk.  Certainly Apple won't say that they want them to all go in the recycle bin (or, just as likely, the garbage bin).  But certainly nobody wants to use an insecure device.

I don't mind if Apple charges a nominal fee for such support.  $5 a device per year?  No problem.   That way Apple could cover its costs of a team to work and address issues.  And clearly there should be an end-point - for instance, the vintage iPhone/3G/3GS models are pretty obsolete.  Maybe there should be some kind of OS software switch that just turns old smartphones into a dumbphone, and disables all the stuff that has a higher potential for a security failure.

Now Google's Android platform is just as bad, if not worse.  Many Android phones currently in retail have major security flaws, and it seems like no one - not the manufacturers, not the retailers, nor Google seem to give a crap.

The auto industry is actually better at fixing dangerous cars after their release.  Smartphone makers should take it upon themselves to at least rise to the level of the auto industry.


Lance J., the Anti-Sales Guy

I now see all these friends and family who get entrapped in a salesperson's bullshit.  I'm tired of it.

I'm fine with people selling products.  I'm fine with people over-paying for a product because that product is great, convenient, or well-supported.

Unfortunately, the majority of corporations ask their salespeople to sell product using lies, distortion, and manipulation.  It's bullshit, it's wrong, and I'm gonna stop it.

I hear it all the time:
  • Auto Repair Industry:  Your brakes are down to 2 mm, so driving could be dangerous.  We can't be responsible. (Using Fear)
  • Mobile Telephone Industry: You should get a bigger mobile data plan because you use more than simple email. (Using Fake Advice)
  • Cable TV Industry: If you sign up for Cable TV, Internet, and Telephone, your monthly bill will actually be a little lower than if you just signed up for two of them. (Using a Fake Bargain)
  • Electronics Retailer: The $30 gold plated HDMI cable will give you a better picture then the $5 HDMI cable, (Touting Fake Quality)
All these sales pitches are nothing but lies.  These sales pitches use fear and manipulation to encourage people to hand over money.  Instead of being honest and providing good service and advising their customers, these corporations attempt to entrap people and sell them useless garbage.
"Oh, we don't KNOW what people need and want!" Bullshit.

These salesmen never EVER provide quality service or advice.  They never say:
  • "You have 2mm of good brake pad left.  They are 100% safe.  You should get your pads replaced after your brakes start to squeal, as that's the indicator of needing your pads replaced".
  • "You likely only need a 2 GB data plan; the unlimited plan may be a total waste". 
  • "This Triple-play package will lock you into a multi-year contract that will likely cost you".
  • "Get the $5 HDMI cable.  It's digital.  The picture will be identical to what you'll see with the $30 cable".
Have you heard lies like these before?  Or something like them?   Please share.

Our New Lighting Strategy Saves Us 70% on Electricity Costs

I'm responsible for managing the energy use at my condominium complex. 

My mission is to always minimize our energy costs to in order to help keep our association fees low.   By strategically using LED lighting, we have been able to reduce our lighting costs by 70%.


We have a bunch of common area lighting that, for safety reasons, needs to be on 24 hours a day.

However, our lighting doesn't have to be super-bright.  It needs to be strong enough so that people can safely use the common areas, and safely enter and exit the building.

Over the past decade we have used 14 watt CFLs in all of our fixtures.   With a 25,000 hour service life, each CFL bulb lasted about 3 years (there are about 8750 hours in a year).    Some CFLs died sooner, and some lasted much longer, but given that the bulbs are lit 24 hours a day, they don't see the stress of being switched on and off.

But even with CFLs, we were still spending about $3000 a year on electric power (or $250 a month).  So I aimed to cut that electricity cost.

The Strategy

In my own unit I was using a 6 watt LED for a desk lamp, and I thought it was pretty darn bright.  And the 14 watt bulbs we were using in condominium's common areas were very bright.  So that got me thinking: could we replace the 14 watt CFL bulbs with lower brightness bulbs?

That idea helped me build a strategy:  Figure out what the right bulb is, instead of just slapping any cheap or available bulb into a fixture. 
  1. Determine the lowest brightness you need for each area.
  2. Find the lowest wattage brand-name bulb that meets or exceeds that brightness.
  3. Watch the electricity costs go down.
So we did some research to determine what brightness bulb would be sufficient. We bought a few sample bulbs from various manufacturers and tried some out.  We started with the lowest watt bulbs we could find.

The lowest wattage we could find - 3 watt bulbs - looked great in areas with multiple fixtures, but were too dim in areas served by only one fixture.  The next step up - 5 watt bulbs - were definitely good in all areas.

After testing out these bulbs in a couple zones for over a year, and without any failures, we ended up buying a bunch of bulbs and changed out our CFLs.  We replaced 70% of our 14 watt CFLs with 5 watt LEDs, the remaining 30% were replaced with 3 watt LEDs.  On average, that's 4.4 watt per bulb, or about a 70% reduction in power costs.

So far, after a year of full service, no LEDs have failed, so I do expect we'll get something close to the 50,000 hour lifespan - that's about 6 years.  That's great from a maintenance point of view.

As a bonus, our lighting is now of a consistent quality and color.  It no longer looks like we just slap up any bulb regardless of brightness or color into a fixture.   We've only heard positive comments from residents. 

What did we do with all those used, but working CFLs?  They were still working, so we didn't want to throw them away.  So instead, we offered them "for free" to the members of the condo association for use in their own units.

So what does that all mean at the end of the day?    It means that our monthly electric bill has gone down a lot.  We were running at about $3000 a year.  Now we spend less than $1000 a year.  That's $2000 saved every year, or $20,000 over 10 years.

  • Electricity costs reduced by 70%
  • Bulb replacement costs reduced by 50%

  • None so far


Leaving Sprint for Huge Cost Savings

So my friend is under a Sprint contract, and it's killing him.  He pays $145 a month for service for two lines, and he has 8 months left on his contract.  That is $1745 a year!

He came to me because he wanted to upgrade his phones - he has two iPhone 5c's with 8 GB of storage - not enough - and so he was thinking of signing up with Sprint for another two years to get "new phones".  Stop!  Stop the contracts!  $1745 a year is a LOT OF MONEY.

So we have THREE problems:
  • He is under contract for 8 more months, at $145 a month.
  • He has two phones that are too limiting for his day-to-day use
  • He has two phones that are locked to the Sprint network
We have FOUR goals:
  • Reduce money being sucked out the door
  • Get phones that are not so limiting
  • Get phones that are not locked exclusively to Sprint
  • Get out of any "contracts"
Stage 1: New phones!

So our first step is to get new phones - phones that will work with Sprint or any other carrier.

Well, not NEW phones, but good and affordable USED phones.  We are ordering a couple unlocked 32 GB iPhone 5Cs.  They're pretty inexpensive, at $140.

Then we hope to sell the 8 GB iPhone 5Cs for about $100.  That's an $80 expense in total, if it all goes our way.

Stage 2: Validating the new phones

Stay tuned!  We haven't received the "better" phones yet. 

Buzzing Mac is actually my Mighty Mouse

My Mac started to make this crazy buzzing sound, and it was bothering me for several days.  At first I thought it was the hard disk drive failing.  Then I thought it might have been the internal fan getting worn out.

Then I realized it was my Apple Mighty Mouse.  It was buzzing.  Making a racket.

Imagining how a mouse would buzz, I first thought it might be an internal electronic component about to fail.  But then I learned that the Mighty Mouse has a tiny a little speaker which makes a "click click click" sound as you rotate its tiny trackball.

That little internal speaker is now clicking all the time, dozens of times per second.  And the left/right trackball movement seems non-functional.  Ug, its REALLY ANNOYING.  The mouse is about 6 years old, so it could be an internal failure related to age, or it might have gotten a little, um, wet.

I'll open it up soon to see if there is anything I can do.  If not, maybe I'll just cut out the speaker circuit.  Alternatively, I might just buy a used Mighty Mouse.

I'll update this as I explore this thing.

Update 1: The Speaker in the Apple Mighty Mouse

Well, I've confirmed it.  I bought a used USB mighty mouse and fooled around with it.  I confirmed that the USB Mighty Mouse does have a speaker in it.  Here's a way to check it out!
  • Unplug mouse.  Roll trackball around and listen for the little clicks.  They aren't there.
  • Plug in mouse. Roll trackball around and listen to the little clicks.  Click click click!
Update 2: Take-apart of a USB Mighty Mouse

I don't really feel like tearing apart my broken Mighty Mouse, but something draws me to it.    I dig in, levering the outer ring off to get inside.  It's crazy dirty inside, with lots of long hairs and other garbage that you'd never think could make it into these very tight spaces.

There she is, a little speaker, making all that noise.   By disconnecting the scroll ball ribbon cable, the clicks stop.

The white round speaker of the Apple Mighty Mouse
I take apart the scroll ball mechanism.  It's unholy - scum and dirt all over.  I clean it out and remove just the cable.  It looks worn down and gross.  I clean it the best I can.  I plug in the cable - which has four sensors on-board - and it clicks the speaker.  ug!

The scroll ball mechanism seems to use four rotating magnets to sense scrolling, and evidently a hall effect sensor has failed.  So at this point, I re-assemble the mighty mouse with the track ball circuit disconnected.  Now it functions like an old-school mouse without a scroll mechanism.

The very dirty track ball cable, with four embedded sensors

Update 3: Into the Parts Bin

I bought a working used USB Mighty Mouse on eBay.  It was under $10 and it works great.   My well-worn, semi-functional Mighty Mouse goes into the parts bin, for future cannibalization, exploration, or eventual recycling.


My MetroPCS Experience and Review

I decided to try out MetroPCS.  I have been paying AT&T about $70 a month since the advent of the iPhone 3G.   Competitors offer less expensive plans.   Now is the time for me to look for alternatives.

My approach for selecting a new carrier is to (1) Figure out what level of service I need, (2) Figure out what quality and features I will be happy with, and (3) locating the low cost provider given that.

Any plan I sign up for should closely consider my usage.  I went to my AT&T account on-line and looked at the use of my AT&T plan over the past 18 months, and was able to determine my usage pattern.  With this information, I can find the most affordable plan for me.

Max of 300 minutes per month
Max of 50 texts per month (thanks to iMessage!)
Max of 5.5 GB per month (outlier)
Max of 2.5 GB per month (typical)
 table 1: my usage pattern

What I want out of a mobile phone service

Everyone has different interests when it comes to service quality and features. 

In terms of quality, my provider must be able to provide me with decent service.  Considering that Sprint doesn't cover my neighborhood very well, I am not currently looking at Sprint-based providers.  This might change over time as Sprint builds out its network in my region.

Another notion of quality is data performance - basically ping times and bits-per-second.  Sadly, no provider will honestly tell me with a what kind of performance I'll likely see in my neighborhood, so I'll have to try prospective providers in order to make a decision.

My notion of "features" needs an explanation. Modern phones like the iPhone support features like Visual Voicemail and WiFi calling.  I'd much prefer a vendor that supports advanced features.

Finally, there is cost.   I want to to pay as little as possible very month.  I don't want surprise charges.  I don't mind if I have to pay more if my usage radically increases, but I want that to be my explicit choice. 

I built a little table that describes the very basics that are important to me.  This table isn't all that intellectually sophisticated, but it does boil it all down to the essentials:

Service should provide adequate reliability where I typically go.  I'm not so interested in service quality in rural Wyoming, and am more interested in service where I spend 99.99% of my time (near home & work)
Service should provide modern smartphone features I want
Service should be of the lowest available cost
table 2: desired service features

Services In The Running

So I scoured the Internet and priced out as many providers as I could find.  It took me several hours to do this.  In the end I boiled it all down to this chart:

Total Wireless
AT&T (current)
Monthly Cost
Adequate talk/text
Adequate Data?
5 GB
3 GB + music
Advanced Features
Annual Savings
Tower Provider

One first thing to note is that although all the plans are different, they will all meet my voice, texting, and data needs.   Yes, some provide more than others, but that's immaterial to me.  Providing me with 3 GB versus 100 GB is moot if I only use 2.5 GB.

Secondly, it is important to note that Total Wireless doesn't support modern features like Visual Voicemail.    Why?  I'm not sure, but I speculate that they've agreed with Verizon that they would not deliver advanced features.  But right there, it means that Total is worth a little less to me.

Finally, it is clear that choosing an alternate carrier WILL save me between $360 to $440 a year.  That's good money.

But there is a risk.  My AT&T plan is grandfathered in, and if I leave them I can never get it back.  So I want to try before I buy.  And so I will try MetroPCS and keep AT&T.  If MetroPCS services my needs, I will then terminate my AT&T account by having my AT&T number ported to MetroPCS.

So without further delay, here is my story of my MetroPCS "Trial".

Signing Up for MetroPCS

I went into the MetroPCS store to get service for my iPhone 6.  I went in for the 40$ plan, which provides 3 GB of data.

The store was empty, then crowded, then empty.  Very wavy.  Maybe people come in to pay their bills. There was one irate customer, but some people are nuts and she seemed a little crazy-stressed.  The poor sales guy definitely handled her professionally and everything was smoothed over.

I went to the counter and met Kevin, a sales guy.  Nice guy, good salesman, professional demeanor.   After some sales-pitching by Kevin, I convinced him I would only go for the $40 3GB plan.  He tried to get me to go for other things, but I told him that this was an experiment.  He didn’t try to hard-sell me… I think he realized I wasn’t going to be convinced (he was right).

The plan itself was $40/month total for 3 GB.  All the "executive compensation fees" that the big carriers tack on are not part of the MetroPCS strategy.  So my total bill was $40.  That's awesome, and it's the way it should be.

MetroPCS charges for a sim card, which is basically $10 + sales tax.  OK, take my money.  But in contrast, AT&T had the habit of charging me $30 (or is it $40?) for "activation", which was totally bullshitty.   Another win for MetroPCS.

Kevin told me of an option an extra $2 a month you get caller ID with name, but I didn’t want that so I declined.  Phone calls are so 20th century.

My plan runs for 30 days, which seems to imply that it “ticks back” through the months (as months are, on average, longer than 30 days.)  But instead it seems to be monthly, with payment due on the same day of the month every month.  So maybe a plan actually runs for 30.44 days (or 365.25÷12)

MetroPCS garbage moment #1: I asked Kevin about how I can pay my monthly bill.  Kevin told me that I can pay online (free).  If I pay at the in-store kiosk, they charge me $2.  If I pay at the store register, they charge me $3.  I think this totally blows, because it penalizes people who can’t readily get online (basically, the elderly, the poor, the inept, the illiterate).  This is shameful and MetroPCS should change their policy to not penalize those who are down on their luck.  People WANT to pay on-line because it is a convenience.   Those who cannot pay online can least afford the extra $2 or $3.

Of course, I'll pay online via credit card and avoid this "fee for the poor and elderly".

Kevin and I made a test call in the store using my new MetroPCS number, and it was all good.  I left.

The Service: First thoughts and beyond

The first thing I tried when I got to my car was a data speed test.  About 10.9 Mbit down and 1.45 Mbit up.  Little slow on the upside, but I’m OK with that if it's typical.

Once I got home, I found that I was getting one to two “bar” of signal at home, which is what I get with T-Mobile providers.  But I made several phone calls, and the quality seemed great regardless of the bar count.  In any case, I enabled WiFi calling, which also seems to work great.

Since I kept my AT&T number, I decided that I'd forward my AT&T calls to my MetroPCS number.  Sadly, due to a weird AT&T issue, AT&T wouldn't let me forward calls to anywhere.   I decided not to pursue this issue with AT&T.  And so therefore I have two phone numbers.  Good thing I have a spare phone.

Next I set up my MetroPCS voicemail.  This seemed just fine, and it worked just like any other modern visual voicemail service.

MetroPCS garbage moment #2: Next, I tried to move my MetroPCS SIM card to another device.  It turns out that MetroPCS links the SIM to the phone, and so you just can't move SIM cards around in MetroPCS world.  You have to contact MetroPCS to make such a device change.  You can do it in the store for $15, or (smarter), you can do it over the phone for free.

The MetroPCS Web Site

Once I had service, I tried to create a new account on the MetroPCS web site.  Lame old-school password rules don’t like any special characters.  That's weird, because security guys absolutely LOVE long, complicated passwords.  Doesn't MetroPCS respect security guys?

As I set up an account, the web site told me that it would send me an SMS to verify me.  That makes sense.  Except I couldn’t finish setting up account because I never received the SMS.  I tried many times, no SMS.  Humph.  Maybe I’ll get those SMS tonight?  Hope so.  But I’ll mute my phone in case they all come in at 2 AM.

No worries, 45 minutes later all the SMS come in and I am able to log into my account.  I hope SMS’s aren’t always deferred.  That’d be dumb.

Once I'm on their web site, its clear that its pretty basic and old-school (2010-vintage).  It seems like a great marketing opportunity to up-sell is being missed, but I do appreciate that the site fairly “lean” without too much junk.  That said, their web site could be leaner with fewer pointless resources.

I find where I can pay my monthly bill, and it looks like they’ll auto-deduct from my card.  Nice.  Will set this up later if I decide to stick it out.


I happened to do a bit of regional travel this month (250 mile radius), and the MetroPCS service seemed fine. There were some dead spots when I traveled in one rural part, but that seemed to be at a rare point, and would impact me for fewer than 10 minutes out of a year.

Data and Performance

I did speed tests during the month and performance was quite variable - from a crazy fast 99 Mbit/s down to 5 Mbit/s.  Ping times were normal, and speed averaged 15 Mbit/second.  It was perfectly great for watching HD video over Netflix.  Upload speeds were worse, but I don't really upload much data (which, by the way, saves data).

After 29 days, I used about 1 GB of data.  In researching my data use, it was clear that music app data was not being counted against me.  I verified that Pandora and Apple Music data was "free" and not counted towards my monthly limit.

Since I was at the end of the month, I decided to exhaust my remaining "high speed data allotment"  to see the service's behavior.  I used Netflix to do finish off all my data.  About 70 minutes of HD video later, I got a text from MetroPCS that I had 10% "high speed" data left.  About 20 minutes after that, I got a text that I used all of my 3 GB.  The service immediately slowed down to about 132 kbit/second (maybe choked at 10% of 1024^2 bit/second?).  Of course, it remained on LTE.  The choke is at the IP layer, no doubt.

I wanted to see how my phone functioned while in "data used up" (choked) mode.  I asked Siri to give me directions to a friend’s house.  Given the slow data, my phone took about 20 seconds to start directions.  Asking Siri to give me directions to my local Target took about 25 seconds.  Asking Siri to get me directions to a friend's house 3000 miles away took about 3.5 minutes.

Streaming music was playable even after running out of high-speed data.  Pandora buffered for a long while but was able to play  tunes.  I’m not sure if music data is subject to the automatic choke, but speculate that it is.

I fetched my email, and that worked quickly.

So therefore, choked data speeds are still useful.

There is a way to add more data, but I didn’t look into that.

Ping (ms)
Download Mbit/S
Upload Mbit/S
MetroPCS (high speed data)
MetroPCS (choked mode)
table: typical data performance I experienced

After a month, I decided to leave AT&T and stay with MetroPCS!

Being happy, I tried to later port my AT&T number over to MetroPCS.  I had to call 611 to do so.  But there was a ($12.50?) fee to change my numbers.  I decided it would be both simpler and cheaper if I simply dropped my MetroPCS plan and port it at sign-up time.

Weirdly, after my call to 611,  MetroPCS started to SMS me my “secret code” that I set up at account creation.  Anyone who stole my phone would then have my code!  And clearly my secret code can be converted to clear text by MetroPCS.   I don’t think this is smart, but maybe there is logic to it?

Dropping MetroPCS

My 30 days were up and I wanted to cancel my MetroPCS service.  I found I had to cancel via a phone call, as canceling is not available on the web site.  I called them up via 611.

But canceling isn't via 611 isn't as easy as it should be.  MetroPCS's automated Voice Response System (IVR) isn't always helpful.  No way to cancel it via IVR, so I had to get to an operator.  What a pain!  I had to say “operator” 3 times in a row in order to get out of an IVR loop and talk to an operator.  It said it really wanted to help me with the IVR, but clearly it could not help me.  CRAZY.  Ya know, if your IVR can’t handle all business issues, then you need a clean way to skip out of the IVR and you need to spend more money on operators.

An operator came on (Daniel).  I asked to cancel.   He was good spirited and had professional demeanor.  But he pushed me to another operator (no doubt they want to try to keep me.)  Preston answered, and I told him I was canceling because I was moving to Canada.  So he didn’t try to hard-sell me to stay.  Yay!

Cancelled service.  Phone line dropped immediately and my MetroPCS service was terminated.


Commitment to MetroPCS

So yesterday I went back to MetroPCS to leave AT&T for good.  I brought along three critical things: my phone, my AT&T Account Number, and my AT&T Account PIN.  The Account Number and PIN are required to move my AT&T phone number to MetroPCS.

I went up to a sales guy, Michael.  Again, the MetroPCS sales people have been fabulous with me. We briefly talked and he proceeded to give me an account and ported my phone number.  The new service using my old phone number took no more than 5 minutes, and my entire time at the store took less than 15 minutes.  Fast, efficient and accurate.

Since I was porting my number over, MetroPCS didn't charge me $10 for the SIM like they did when I wasn't porting.  So my total outlay for a month of service was $40.  Yay! 

A little while later I got an email from AT&T, telling me that my phone number was no longer associated with my AT&T account.   I'm not sure if AT&T owes me money for my terminated plan, but I suspect that I'll see a check in the mail soon. It's kind of sad - I was an AT&T customer for more than 15 years.

Then again, whatever its name, I've given AT&T more than $10,000.  So maybe it was time to break up with them.  If I keep this MetroPCS service for 15 years, I'll save about $5,400.

My Overall Experience with MetroPCS

  • Services my needs
  • Saves me about $360 a year
  • No contract, so no commitments.
  • Great people in store and on phone
Cons.  (none of the cons are show-stoppers):
  • SIM card mobility restrictions.  You need to contact MetroPCS to move a SIM between devices.
  • Dumb extra charges that can cost the poor and elderly. Distasteful.
  • Weak 611 IVR system (at least for what I was trying to do)
  • Lame web site password limitations
  • Dated web site, that looks and works like 2005.

So I've gone through a month and it has been a glowing success, with no issues.  I do notice that MetroPCS and the iPhone count data in inconsistent units.  The iPhone incorrectly counts a Megabyte as 1024*1024 bytes, while MetroPCS counts a Megabyte correctly, as 1000*1000 bytes.  The network industry has never used the obtuse "kilo=1024" convention.  It is surprising that Apple gets this wrong.

Regardless, I was way under using my monthly data allotment.

Another Update!

I went with my friend Aaron to MetroPCS so that he could swap his SIM into a new phone.  Again, the sales guy was amazingly good.  He said that if he switched it at the counter there would be a $15 charge, but if the customer did it via the telephone, it'd be free.  Aaron did that successfully.  I've updated the above contrary information. 


Extending Battery Life

So you have a smartphone or laptop that you've owned for a year or two and the battery seems to lose charge very quickly.  However, some batteries in a friend's identical device seems to have lasted for years longer!  So what's the deal?

The deal is that modern devices use lithium-ion batteries, and it is very easy to mistakenly misuse a lithium-ion battery. 

What?  A misuse a battery?  How does THAT happen?

The common killer of lithium-ion batteries is heat.  Heat just sucks the life out of a lithium-ion battery, and they never, ever recover.  My goal for battery care is to always avoid unreasonable heat.  So here's the short list of the rules I follow for long battery life:
  • Never leave a device in a car.  Cars interiors can get very hot.  If I leave my device in the car even once for 15 minutes, where the temperature can get over 120 °F, I have forever stolen some life out of my battery's soul.
  • Never leave a device in the hot sun.  This is really a variation of the above statement.
  • Never use cheap, generic chargers.  A cheap charger can send poor quality power to a device, and then the device needs to smooth out and reduce that voltage to something it can safely use.  All that voltage smoothing results in substantial heat - which, as we now know, kills batteries. 
  • Never use "fast chargers".  Fast chargers inject more energy into a device over less time, resulting in more heat and more stress on the battery.  It's best to use a normal charger, and avoid speed-charging.
  • If a device feels hot, restart it.   Bad software can spin out of control, resulting in a lot of internal heat production.  A restart terminates all software and gives it a chance to behave normally.  And make sure the device's software is up-to-date.  Up-to-date software is less likely spin out of control and heat up a device.
  • Never let a battery completely discharge.  Once a lithium battery is allowed to sit for months without a charge, it might not be recharge-able.  I never store a device with very low charge, and I never let a device sit for several months without charging it up.
Other myths to debunk:

I think the following silly ideas come from the age of the NiCad batteries of the 1970s.
  • Do not purposefully "drain" or "calibrate" a battery.  Doing so simply weakens the battery.
  • Do not worry about "overcharging" a battery.  All modern devices have a managed charging system.  These systems are necessary to protect the battery and to prevent event leaks, fires, and explosions. Attempting to trick the charging system will only reduce battery life.
Buying a replacement battery

It can be hard to buy a replacement battery.  First-party manufacturers (like Samsung, Apple, Dell, etc) all charge quite a lot for a battery.  But the battery they sell you has a real warranty and will certainly be of higher quality.

Third party suppliers generally sell batteries without a warranty.  Some might claim they offer a warranty, but most don't provide any mechanism to make a warranty claim.   That's no warranty.  Furthermore,  some of these 3rd party batteries can be ill-fitting, low-powered, and can have very short useful lives.  But they are very inexpensive, often one quarter of the price of a manufacturer's battery.

I almost always buy 3rd party batteries on Amazon.  Here are the 3rd party battery purchases that I have been happy with.  Of course, double-check your model number before buying anything!


Powerful Utilities for Mac Performance and Optimization

A lot of people as me about the best utilities to install to improve Mac performance and reliability.   This is a great question, and I'm going to answer that here.

I use just about the oldest supported Mac, and it works great.  But it has taken some work to make it great.  If you have a slow Mac, it is likely easily fixable!

Here is my step-wise procedure for trying to address a slow Mac:

Common Mac Performance Issues

There are a few common things that greatly slow down your Mac:
  1. Out of Date software
  2. Lack of Disk Storage
  3. Sick Hard Disk Drive
  4. Slow Hard Drive
  5. Lack of Memory
  6. Bad "repair" software
 My Speed-up-that-Mac procedure
  1. Restart.  There's nothing like a restart to clear out memory.
  2. Empty Trash, which frees up some disk space.
  3. Software Update.  OS updates and core applications should be kept up to date.
  4. App Update.  Applications and  Plug-ins should be kept up to date.
  5. Disk Utility First Aid/Repair.
  6. Networking: Restart all networking equipment (wifi router, modem)
  7. Restart again (restart at least once per week) 
From here, check to see if performance is improved.

  1. Perform network speed/reliability test (speedtest.net)
  2. Using Activity Monitor, see what (if any) apps are consuming significant CPU or Memory.  Update them, or remove them if reasonable.
  3. Inspect free disk space.  More than 20% free is good, otherwise bad.
  4. Inspect RAM.  4 GB or more is good, otherwise bad.
  1. If less than 4 GB of RAM, update to 4 GB or 8 GB.
  2. If a traditional hard drive is in use, upgrade to an SSD.
  3. If storage is reported to not be healthy, upgrade SSD.
  4. Upgrade to the most recent version of MacOS.
  5. If network is bad, consider upgrading WIFI router to a newer model (anything newer than 2014 should be fine). 

Well that's the short list I use, generally from the top down, to address any Mac performance issue.  There are some weirder things that can happen that aren't addressed by the above procedure, but I find that to be very rare.

Note that I never install or use any 3rd party utility software - as the vast majority of it does more damage than good.


Update on MacOS Sierra

At this time we have installed Sierra on about 50% of the Macs we maintain, and so far it is going well.  Performance and reliability seems on-par with El Capitan:
  • No kernel panics
  • No significant bugs/failures or "first party" app crashes
  • No app compatibility issues
  • No performance issues
  • No data migration issues (we use Migration Assistant)
 Here is our hardware minimum:

4 GB of RAM.  We have Sierra running well on machines with 4 GB of RAM.   Although the majority of our Macs have 8 GB, 4 GB is totally acceptable.  We haven't had a machine with less than 4 GB of memory for several years.

As we've said earlier, if you're upgrading from less than 4 GB of RAM and need to buy new RAM, we suggest buying 8 GB.

Amount Today Advice Rough Price
Less than 4 GB Go to 8 GB $40
4GB + Do nothing $0

Non-Traditional Drive.  As of earlier this year, we no longer have any traditional hard drives in our Macs.  All of them have been upgraded to either a hybrid drive ("SSHD"), or a straight-up SSD.

Hybrids are much less expensive per GB, so they're a good option for those with a lot of storage needs on a budget.  SSHDs are not as speedy as a good, low-cost "full" SSD, but they're less than half the price.  And an SSHD is a tremendous boost over a traditional quality hard drive... and only about $25 more expensive.

Here's a matrix summing up my thoughts on storage:

Storage Type Rough Price Storage Performance Reliability Battery Use
Traditional HD $55 1 TB Too Slow!!! Very Good Higher
SSHD $80 1 TB Very Good Very Good Medium
SSD (low-cost) $240 1 TB Excellent Excellent Lower

Age.   We have no concerns about running Sierra on any Sierra-capable Mac.   Even our oldest eligible Macs - models over 5 years old - run Sierra well.

Bugs and Stuff: Broken Keyboard Driver with Bootable Sierra Installer

The only bug we've seen is when performing a clean install from USB boot media.  The keyboard driver of the bootable installer seems to be broken on particular Macs.  A keyboard is required to use Terminal, or to perform some Disk Utility actions.

Our work-around is to use an older bootable OS X installer (from El Capitan, Yosemite, or whatever) to prepare the machine, and then boot with the Sierra Installer to do the actual OS installation - no keyboard required.


Faster Internet and the Bullshit of Cable Modem Marketing

People assume that a blazing fast 340 Mbit/second cable modem will give them better performance than an 80 Mbit/second modem.  "4 times faster!  8 times faster!"

All of that talk is simple marketing garbage, a hold-over sales pitch from the days of dialup modems.

The fact is that the Internet doesn't work like that.  You will likely get the same exact performance with a low-cost standard cable modem versus an expensive fast modem.

The reason?  Internet bottlenecks.  The Internet is a vast, shared network, and there are plenty of bottlenecks on every network path. With your 340 Mbit/second modem, you'll know that your modem is no longer a bottleneck.  But there are more than a dozen network devices between your computer and Google and so your cable modem is likely never the slowest device on the network.

What other devices are slowing things down?  First, there is your home WIFI.  Then there is bandwidth that your ISP provides to you.  Then there is the network equipment at your ISP, which is shared among perhaps hundreds of thousands of users.  Then there is all of the shared networking equipment between your ISP and Google's ISP, which is used by millions of users.  And then there are all of the network devices at Google - which is used by hundreds of millions of users.

Save your money.  Get a normal, reasonably priced modem, like this TP-Link DOCSIS-3 Modem, or this Netgear DOCSIS-3 Modem. Your performance experience will likely be identical to a "blazing fast high performance" modem, and you'll save a ton of money.


Apple's Self-Driving Car

Today it was reported that Apple's Self-Driving Car project is going through some significant challenges - with some layoffs, reorganization, and perhaps a change in focus.

But let's be realistic here: Self-driving cars are very hard, on the order of magnitude of the Manhattan project... or maybe harder.  An incredible amount of new science, AI, and associated engineering needs to be solved.  It isn't a "simple" hardware/software engineering exercise like creating the first iPhone, or implementing the Android OS from scratch, or implementing Twitter.

It took about 40 years for the helicopter to go from concept to production, as a lot of science and engineering had to happen before a helicopter could be a reasonably safe way to travel.

The idea that a self-driving car project wouldn't go through significant struggles and major changes in direction is fantasy, as that even happens with simple projects, such as the creation of another new fighter aircraft.

Kudos to Apple, Tesla, SpaceX, and others for taking on tremendously challenging projects.

MacBook: burnt out magsafe connector repair

My MacBook (A1342 model from 2009-2010) stopped working, thanks to a cheap knockoff magsafe charger. The magsafe connector looked burnt with heat-distorted melted plastic and some black carbon surfaces.  Plugging in the charger resulted in no lights and no action.  My MacBook was dead.

First I tried to clean up the connector with some Q-Tips, tooth picks and solvent, but that did little, as seen in this "after" picture (below).   The heat generated by the aftermarket charger permanently deformed and distorted the MacBook's magsafe connector.  Clearly the damage was significant.

Burnt Magsafe Connector - Replacement Required!
Looking at iFixIt, I determined that it would not be difficult to replace only the magsafe connector with the right tools.  I hoped and prayed that the problem was isolated to the connector assembly and not the logic board.

I bought a replacement magsafe connector assembly (available from Amazon) for under $20 and started to replace it by following the instructions on iFixIt, using a proper tool kit like this one.

When I pulled out the old Magsafe, and boy did it look nasty.

My removed Melted Magsafe Connector, thanks to a crappy 3rd party charger.
While I was in there, I also cleaned out a lot of dust that accumulated inside of the MacBook.  I cleaned up the following locations:
  • Fan, heat sink, and vents
  • Logic board (front and back)
  • Case
  • Battery
  • DVD Drive
To remove the dust, I first quickly vacuumed up the big blobs.  Then I used a dry, unused paintbrush and lightly brushed dust off of all the surfaces.  This was followed by more vacuuming.

Also, while I was in there, I carefully cleaned out any scum that developed around the port openings, using some Q-tips and some solvent.

After re-assembly, my MacBook fired right back up and I was back in business.  Success!  I'm very happy, and will certainly never use a cheap aftermarket charger ever again.

My new Magsafe connector allows charging once again!

Now that this MacBook is again in operation, I have since updated its memory and the hard drive, as seen in my other articles.


The Trick to Properly Fixing the Rubber Feet of a MacBook Pro

A black rubber foot of my trusty old MacBook Pro finally broke off, and I didn't like how it made the laptop wobbly.  A fix was needed, and quick!

First I bought some replacement feet, the kind that just stick on.   These feet are nice because there is no need open up the machine to fix a broken foot.  They're very inexpensive, and the ones I bought have adhesive tape already applied - just peal and stick.  I bought feet like the ones found here.

I have read many complaints about how the adhesive doesn't stick, but I think that's because people don't know the two tricks required to get the new feet to stick properly.

New Feet to replace a lost Foot

Important Trick 1: Make sure to remove all remnants of the old foot - including the pieces in the hole.   My factory-installed foot was attached both mechanically and with adhesive, but when the old foot sheared off, some pieces of the old foot remained in the hole.  Those plastic bits would have prevented the new foot from seating properly.  I used tweezers to carefully break up and extract the remnants of the old foot.  Of course, never dig too deep - there are sensitive electronics under there!

Photo 1: Be sure to fully remove the old foot.

Important Trick 2: I used some solvent to soften and release the old clear adhesive tape.  I didn't even know that this adhesive tape was on the laptop body until I managed to loosen it up with some nail polish remover.  The tape is in the shape of a small transparent disk, and will prevent the new adhesive from sticking well unless the disk and all the old transparent adhesive is completely removed.  Of course, again, never let anything get into the hole.  That could do some damage.

Clean the recess very well

Whoa!  A little transparent piece of old adhesive tape appears!  Tweezers help me pull it away

Remove the old tape, and finish cleaning out the old adhesive from the recess

Once the recess was completely clean and dry, I properly aligned the new foot and stuck it in place.

Now my MacBook Pro is steady on its feet and ready for more years of computing bliss.  


Not the End of an Era: the continuing life of a 2009 MacBook Pro

The 2009 MacBook Pro's days are numbered.  As you likely know by now, the 2009 MacBook Pro will not be supported by Mac OS Sierra.  I have one of these beasts, and it is a wonderful machine despite 7+ years of heavy use.

But that doesn't mean my 2009 MacBook Pro is dead or will become immediately useless.  Quite to the contrary, I expect to use my MBP securely for several more years.

Mac OS X El Capitan will continued to be supported by Apple with security patches for some amount of time.  How long?  My research shows that Apple supports a deprecated operating system for at least two years after the release of its subsequent operating system.

So if Sierra is released on October 1st, 2016, El Capitan will be supported by Apple until at least October 1st, 2018.

Then what?  Well, by rights, most people should no longer use El Capitan as it becomes security-obsolete.  But some more technically-minded people might have some options:
  • Hack Sierra onto the machine.  This may extend the security of the machine until Sierra is obsolete, perhaps around October 1st, 2019.  No one can promise that Sierra will continue to operate properly on a 2009 MBP, but it could.
  • Manually patch El Capitan.  This may be reasonably possible depending on the security vulnerabilities revealed after obsolescence.  For example, there is an NTP security patch that can be manually built and applied to the obsolete Lion.
  • Install another operating system onto the machine, such as Linux or Windows.
  • Use the machine in an off-line way.  For example, the machine could potentially be used securely if it never shares data with the outside world.
So see?  It isn't all that bad.  After all, a notebook computer with a 9+ year active lifespan has done well.  Even if El Capitan gets all obsolete in the 2018 - 2019 time frame, my  machine will continue to be able to provide good service for years to come.


Maxing out your White 2009 2010 MacBook for Performance and Longevity

Even in the later half of 2016, the 2009 & 2010 white MacBook can be a fantastic laptop if upgraded with modern parts.   Here's the deal.

I am responsible for maintaining a small fleet of these MacBooks, and have found that it is very easy to upgrade these MacBooks to perform in a manner competitive with a new, $1000+ laptop.

2009-2010 "A1342" white MacBook

The upgrades I typically perform on these MacBooks are Memory, Storage, and Battery.  Depending on your needs, this can typically cost from $25 to $125 if you buy and install the parts yourself.   This is a great way to get another 3 or more years out of your MacBook.

Let me be clear:  I am talking about the white plastic unibody polycarbonate MacBook from 2009 and 2010, as exactly pictured above.  Check out "About this Mac" under the Apple menu of your Mac to verify that you have a "Late 2009" or "Mid-2010" MacBook.

Here are the details of the upgrade parts I usually use:

Memory Upgrade: Typically about $35 ($25 - $35 depending on your budget)
  • 8 GB RAM - $35 (Recommended)
  • 4 GB RAM - $25 (Cheaper option, for those on a budget)
The MacBook was typically delivered with 2 GB of memory.  Today, 2 GB of memory is below the minimum amount of memory recommended for the smooth operation of modern Mac software.  I recommend upgrading the MacBook's memory to 8 GB of memory.  If your budget is tight, I recommend upgrading to 4 GB of memory, at a cost of about $25.

The MacBook only likes specific memory chips, namely PC3-8500 DDR3 SO-DIMMs.  That sounds complicated, so I'll make it easy:  I have had excellent results with RAM from Amazon, such as these: 8 GB Memory Kit, 4 GB Memory Kit.

SSD Solid State Drive Upgrade: Typically $70 ($45 - $125, depending on your needs and budget)
  • 250 GB SSD - $70 (Recommend for most)
  • 500 GB SSD - $120 (Recommend for those low on space)
The MacBook was typically delivered with a 250 GB hard disk drive.  Today, laptop hard disk drives are near-obsolete.  Instead, modern laptops are delivered with much faster "solid state disk" (SSD).  I advise upgrading the MacBook's hard disk drive with a 250 GB SSD, at a cost of about $65.

If you are using more than 200 GB of hard drive space now, you may want to upgrade to a 500 GB SSD.

I have had good experiences with most (but not all) SSDs.  I have had issues with the very cheapest SSD, so I recommend staying away from the "bottom of the bargain bin".  I have been very happy with the SanDisk II SSDs, which has good price, performance, and reliability.

New fast 240 GB SSD (top) versus old, slow 250 GB Hard Drive (bottom)

Battery Upgrade: Typically $35 (only if your Mac's current battery bothers you)

Many MacBooks have a 5+ year old battery.  If your battery is OK, or if you always use your MacBook plugged in, then I wouldn't replace the battery.  If your battery is no longer holding a reasonable charge AND you use the MacBook while disconnected from the wall, then I would replace the battery.

Make sure you get the right battery, such as this one for the White MacBook 2009-2010

Tools and stuff

When upgrading the memory and hard drive, one needs some tools to do the work.  In particular I  use the following tools, which can be found for under $2 each:
  • Phillips "00" (small) head screwdriver - for removing the back of the computer and the hard drive
  • Torx "T6" head screwdriver - for removing the hard drive mounting pins
  • Tri-wing screw driver - for removing the battery.  This is often included with the purchase of a new battery.

EasyN WiFi Camera Firmware Upgrade

I wanted to upgrade the firmware of my EasyN WiFi camera.  I just like to keep my firmware up to date.  I am sure it is a security sieve no...