atxgeek 


just one more geek in a sea of austin techies

May 18, 2026

Is CNet pandering (a little) to Apple?

This past week a CNet article headline caught my eye:

We Tested 33 New Phones to See Which Charge Fastest and Crown 2 Winners

Source:
https://www.cnet.com/tech/mobile/cnet-lab-exclusive-fastest-wired-wireless-charging-phone/



I can't say this subject (fastest charging times) is likely to overly influence anyone's phone buying decision but the article is interesting to me in that I'm curious to learn the kind of charging rates many of the most recent phone models are now achieving.  I was dismayed, however, to discover that CNet's ranking of the "2 winners" was clearly backwards rendering the remainder of the article's phone rankings suspect.

The issue is that the method of comparison used embraces a fallacy of false equivalence.  The article compares percentage charged but ignores battery capacities which effectively robs the percentage charged values of meaning.  By the author's logic, attaining a 50% charge on a 4,000 mAh battery is equivalent to attaining a 50% charge on a 5,000 mAh battery.  Put another way: a phone charging a 4,000 mAh battery from 0% to 50% in 30 minutes is judged to have "charged as fast as" a phone charging a 5,000 mAh  battery from 0% to 50% in 30 minutes.  Obviously in this scenario the 5,000 battery phone would have charged 25% more capacity (2500 mAh versus 2000 mAh) in the same 30 minutes and would therefore be considered to have delivered a 25% faster rate of charge.

We could assume that CNet's test intentionally disregarded factoring in battery sizes in favor of taking the view that "100% charged" is considered the same benchmark across phone models based on different phones' design and battery choices.  That might be stated something like, "Phone A was designed to have a 4,000 mAh battery, Phone B was designed to have a 5,000 battery, and we are testing to see which can be recharged the fastest as a percentage of each model's own battery capacity."

Judging a phone's charging rate against its own capacity is, however, a terrible basis when purporting to compare the charging rates of dissimilar phone models.  That's like judging who "ran the most distance" by measuring how far each runner ended up from the finish line with different runners having different finish lines.  If you want to equally compare runners' performances you should be measuring how far each runner ended up from a common starting line.  It's a little hard to imagine a "measure each phone only against its own battery capacity" being a thoughtfully-chosen approach for a serious "best charging rate across 33 phones" competition leading to "CNet Lab Awards".

For Your Consideration (Example 1): 
If we were to add to our comparison group a 34th phone with a paltry 1,000 mAh battery (let's call it "baby phone") and only half the other phones' collective average charging speed the baby phone would still be almost certain of outperforming the rest of the pack despite having comparatively terrible charging performance.  By CNet's measurements baby phone would be considered to outperform both the iPhone 17 Pro and the Samsung S26 Ultra if baby phone achieved only 30% of the S26's charging rate.  That's because baby phone's small 1,000 mAh battery would be 100% charged after 30 minutes while the S26's 5,000 mAh battery would only be 76% charged and the iPhone 17 Pro's 4,252 mAh battery only 74% charged.

Example 2:  CNet's approach would be like judging who can fill up a liquid fuel tank the fastest and comparing a professional F1 pit crew to someone filling one of those long-stemmed butane candle lighters.  An F1 crew with special equipment can achieve a refill rate of 12 liters per second (until F1 banned in-race refueling back in 1984, that is) but the butane lighter fills nearly instantly with just a single squirt.  Based on how full by percentage each respective tank ends up after half a second, the lighter with its teeny-tiny fuel reservoir wins every time despite not even being in the same universe as the F1 crew in terms of actual fuel refill rate.  

To really compare apples-to-apples in this example the "rate of charge (fuel refill)" must be based on comparing the amount of energy (fluid) transferred.   It's almost meaningless to compare how full two tanks are as a percentage of tank capacities while also completely disregarding the different sizes of the tanks being compared.


The article's two top-rated phones are both excellent phones with excellent charging rates.  I thought every phone on the list of 33 phones compared was a good-or-great phone model and was a suitable choice for inclusion.  The problem is that CNet gives top honors to Apple's iPhone 17 Pro at 64.5% average overall charge (wired % over 30 minutes + wireless % over 30 minutes / 2) with Samsung's Galaxy 26 Ultra a respectable second at 57.5% average overall charge.  In reality, though, Samsung outscored Apple with a 5% better actual charge rate per CNet's own numbers -- a 12% swing in favor of the S26 and a reversal of the #1 and #2 spots.  



This 5% difference in actual charge rate feels like a whole lot of nitpicking.  In the real world that's pretty much a wash with both phones being really great with respect to actual charging rates.  I would happily support declaring the top five or six phones (most of which are iPhones) as basically being "equally the best" at fast charging.  My issue is not with any of the excellent phones cited in the article but with the article's terribly flawed judging methodology.  Which leads me to...


Small Conspiracy Theory of the Week

(This "conspiracy" section is mostly just for fun.  I don't wear tin-foil hats and I do happen to really like Apple products but CNet's overtly-flawed battery-charging-rate methodology is just a little too odd to ignore...)

The very premise of the article is to judge which phone will "charge the fastest" and the only way to place varying models with varying battery sizes on an even scale is to measure "how much energy is added/stored in a given amount of time".  The article instead measures, by percentage, how much each phone's "tank" is filled after 30 minutes while completely ignoring the fact that these are different-sized tanks.  The article goes further by declaring (more than once) that iPhones are more efficient than other phones and so use power more slowly.  This seems aimed at suggesting that, somehow, battery discharge rates should be considered when measuring how fast a phone's battery is able to be charged.  

For what should be such an easy, straightforward approach (amount of energy over time) that oversight and the resulting unequal method of scoring seems far too obvious to not question whether there might be a little Apple favoritism at play.  Why, though?  Do Apple products account for a notable portion of ad revenue across CNet / Ziff Davis sites?  Is it a little telling or just happenstance that, near the end of typing up my observations, I browsed to CNet.com's home page to see what advertisements might be presented there and was welcomed with a huge T-Mobile iPhone 17 banner?



Perhaps I should have been suspicious from the start upon noting that that the top of the article -- which covers many top-tier makers and even mentions five different phone manufacturers in the article's subtitle -- only provides a "Follow" button option for one phone manufacturer.  Hmm.



April 19, 2026

NOAA Sand Crawler... now with USB-C!

This NOAA weather radio is giving strong sand crawler vibes, right?

It can't just be me...



November 9, 2025


...Now you can enjoy the song in your head for the rest of the day whether you wanted to or not.

You're welcome!

May 25, 2024

Substance Over Style

Warning: mini-rant ahead

For years I've maintained a quiet argument with American grammar-style rule-followers.  This argument is different from my other style argument concerning Americanized dates (it should be the international standard YEAR-MONTH-DAY, not MONTH-DAY-YEAR) but that's fodder for a different post.

Despite that fact that I am American-born and American-educated my argument concerns the American practice of including as part of a quoted value whatever punctuation immediately follows that quote.  For example, in the sentence:

  Stephen said the name of his lawn business is "Yard Masters." 

The business name itself almost certainly does *not* include a period and so was not actually part of the quote from Stephen.  The practice of enveloping non-quoted punctuation into preceding quoted text may seem rather unimportant.  In this example most people would assume the business name is "Yard Masters" with no period.  Even if someone mistakenly believed a period was part of the business name it's hard to imagine any serious negative consequences of doing so.

People who deal with the necessary preciseness of computer code, however, should be at least somewhat irritated at this practice.  Instead of quoting a business name, what if the sentence were:  

  Stephen said the code to disarm the explosive is "n0b!0w." 

The inclusion of the non-quoted period as part of the quote now has very serious consequences.

When pressed with this dilemma the normal response from grammar experts is to restructure the text so that the quote is no longer at the end of the sentence, such as:  Stephen said the code "n0b!0w" will disarm the explosive.  This is an effective, though somewhat clunky, workaround.  Unfortunately it is not a workaround you can depend on -- you will still get cases of people including quotes at the end of sentences (or immediately preceding a comma) so the issue persists.  I was recently reminded of this in some Google technical help instructions:

Following American grammar convention results in incorrect information in Google's instructions.
https://support.google.com/maps/answer/1725632?hl=en#zippy=%2Chow-do-i-opt-my-access-point-out-of-google-location-services

In the case of Google's instructions shown above, users who wish to maintain privacy by opting out of Google Location services must append the text "_nomap" (with no period) to the end of their Wi-Fi access point(s) SSID names.  Unfortunately the writer of the instructions above decided to follow American grammar convention and included sentence punctuation within the quoted values.  The cited values end up being "_nomap." and "12345_nomap." (with periods included) instead of what the values should be: "_nomap" and "12345_nomap".  This is a critically-important technical distinction as anyone wrongly including a period will not opt-out as intended.  Worse, though, is that in this case there is no independent way to verify that a Wi-Fi access point has been successfully opted out so anyone who includes the period as quoted in Google's instructions will mistakenly believe they are opted-out when they are not.

We can argue that Google's writer should have restructured the wording to avoid having quoted values  appear at the end of sentences but the real problem is that American convention is to purposefully change such quoted text from an explicit value to an ambiguous value.

What I (usually) Do
In practice I typically end up using a hybrid of including/not including punctuation within quoted values.  If the piece of writing is conversational I'll lean toward American convention to help hold the rule-followers at bay.  If there is a quote with a critical value, or if the writing is more technical in nature, then I only include within quotes whatever is exactly the value intended.   Trailing punctuation  that's not part of the quote gets left outside of the closing quotation mark (as it should be).  For technical items where every detail matters it's better to be exact and run the risk of someone thinking I don't know my grammar rules rather than follow a style convention that (like the Google example above) results in wrong information.

<end of mini-rant>


February 6, 2021

PayPal, Bitcoin and Taxes (oh, my!) #CryptoGeek

Over the past year, PayPal's stock value has surged considerably along with many others in the online payments processing realm.  Indeed, prior to 2020 PayPal had already solidly established itself as the fourth-largest payments processor behind Visa, Mastercard and Discover.  PayPal is now the largest of all pure-digital payments processors with more currently-active PayPal accounts than the entire population of the United States.  
 
If this weren't already enough, in 2020 PayPal decided to go "all-in" on Cryptocurrencies like Bitcoin, Ethereum, and Litecoin.  You can now easily buy and sell Cryptocurrencies (a.k.a., "crypto") with a standard (free) PayPal account.  This move by PayPal brings crypto accessibility to the masses more than anything has before.  However, the masses are largely not up to speed with how crypto is different from typical currency and, perhaps more importantly, how spending crypto can have a major impact on taxes for US citizens.

The short version: 

  • Treat crypto like stocks, not money.
  • Every crypto transaction has end-of-year tax implications.
  • It is super-easy to get into crypto with a (free) PayPal account.  
  • You can quickly make or lose lots of money with crypto.

The long version: 
Read on...

December 29, 2020

MPOW BH079A driver for Win10 #BluetoothGeek


If you need to add Bluetooth to a Windows 10 PC it's harder to find a cheaper or easier solution than this well-reviewed $7 USB dongle from MPOW.  There are, however, questions about the safety of trusting included hardware driver CDs or even downloading drivers direct from lesser-known manufacturers.  Even the most honest of vendors is not guaranteed to have clean, virus-free driver installation software.  
 
If you purchase a device that is not automatically recognized by Windows then you face the quandary of trusting unknown third-party driver installation software or changing to a different hardware solution.

Such was the case with my own MPOW BH079A USB adapter purchase.  The adapter boasts a large number of great reviews and is compatible with many versions of Windows back to XP. Windows 10 did not, however, automatically detect and set up the device and I have a healthy distrust of driver CDs from lesser-known manufacturers.  To be fair, no MPOW reviewers cited any cases of malware but I still prefer to avoid software from lesser-known sources when possible.

Fortunately the adapter is basically just MPOW's casing wrapped around the trusty Broadcom BCM20702 Bluetooth single-chip solution which has been around for a decade.  Any official Broadcom BCM20702 Windows 10 driver should be all that is needed (e.g., the driver doesn't have to be from MPOW).

Although my MPOW USB dongle is being used on a home-built PC, I went with a Hewlett Packard (HP)-supplied Broadcom BCM20702 Win10 driver (circa 2017):



Install the driver then insert the USB dongle and it should be activated automatically -- no need to turn it on in Windows settings.  If you already had the USB adapter plugged in when installing the driver you may need to remove and then reinsert the adapter.

Cheers!

December 8, 2020

Save, then save, then save again #ShoppingGeek


We're well into the holiday shopping season as I write this but it's never too late for tips on saving money.  This year I've been doubling, tripling and even quadrupling my online shopping savings opportunities.  How?  With some new virtual takes on a few old payment standbys.

When everything is clicking, my best online shopping savings scenario is:
  1. Fill up my online cart with items
  2. Get a discount applied automatically
  3. Get part of the remaining balance paid for via saved credits (bonus gift cards)
  4. Get part of the remaining balance paid for via discounted virtual gift card
  5. Earn cash back on whatever balance is left that I actually have to pay
  6. Earn credit towards future bonus gift cards (see #3)
That is, quite literally, savings on top of savings on top of savings (on top of savings on top of savings).  I *do* have to jump through a couple of hoops to get things set up, but I'm frugal enough that I don't mind doing so when I consider some of the larger payments and purchases we make.  Read on to see how I stack up these savings...

September 25, 2020

eero 6: Will it work with original eero? #NetworkGeek


Will eero 6 work with
first-generation eeros?

The new eero 6 (link) and eero 6 Pro (link) lines of mesh routers and access points are now available bringing improved connectivity, faster wireless speeds and integrated Zigbee smart hubs.  

There are plenty of articles and reviews covering the improved features but I've yet to see anyone answering one burning question:  

What about existing eero owners?  Specifically, can existing (first-generation) eero mesh networks incorporate new eero 6 devices?  Or do existing owners have to replace their old eero units in order to upgrade to eero 6?

Read on for the answer...

December 8, 2019

6-month review update: UltraLoq U-Bolt Pro

Last May I shared info on my then-latest purchase of a crowdfunded item, the "UltraLoq U-Bolt Pro Combo".  This was a new, fancier smart lock from a company that was already shipping smart lock solutions with good reviews.  The new "U-Bolt Pro Combo" was notable in that it included a Bluetooth-to-WiFi bridge adapter for a smart deadbolt with access options including fingerprint scan, code entry, wireless entry (cell-phone proximity detection), and traditional key entry.

Was it any good, though?  YES!

Read on to get my impressions after 6 months of heavy use...

November 12, 2019

SQL Server snippets #SQLGeek

This post will be an ongoing, growing collection of SQL Server code snippets that I happen to find useful. There is no particular theme, order or categorization -- just a random collection of helpful items that I hope others will find useful, too.

My apologies if I eventually include anything particularly unique or complex without citing an original author -- please comment with links to original posts so I can verify and add citations accordingly.

Read on for the snippets...

May 16, 2019

Save 64% on a COOL smart deadbolt

I love to window-shop crowd-funded products and, occasionally, purchase.  So far I've had great experiences with the products I've gambled on.  For instance, I pre-ordered a full eero mesh WiFi system 14 months before its actual release and the system has been every bit as good as I'd hoped.  I threw some dollars at a Coolest rolling cooler and was one of the lucky bunch that actually received a unit right away (the blender worked, too!)

Smart Lock
Today I laid my credit card info down on a recently-released smart door lock -- a deadbolt for the front door.  Sure, smart locks aren't a "new" gadget category but this product and pre-order deal is the best combo I've seen:  a deadbolt lock supporting Bluetooth access, code-entry access, physical key access, fingerprint access, a WiFi bridge and a companion smartphone app for an early-bird price of $159 shipped ($449 MSRP).

The U-Bolt Pro Combo lock has typical smart-lock features like auto-unlock when your phone gets within Bluetooth range and it offers the ability to remotely lock/unlock with a smartphone app for Andriod and iOS.  It's with the less-typical capabilities, though, that things start to get interesting:  IFTTT integration, support for both Alexa and Google Assistant, and an exterior-accessible power port (micro-USB) to power up the lock if the primary batteries go dead and you don't have the physical key handy.  Got a big family?  Lots of close friends?  This lock lets you set up access for up to 60 users including one access code, one smartphone registration, and two fingerprints for each user.
 
Even More Interesting: Feature Scheduling
You can create schedules to determine when certain features are enabled.  For instance, you could allow a maid service to unlock via entry code or fingerprint only during certain hours of the day.

Using IFTTT, you can cobble together a wide range of scheduling options.  Imagine automatically allowing a trusted friend the ability to unlock via code if the door hasn't been unlocked in more than 3 days (i.e., when you go on vacation but forget to alter access settings for your friend). 
 

Caveats
Note that, at this point, I don't have a U-Bolt to test myself so I'm only throwing out ideas with the expectation that the listed features can be trusted to work as advertised.  There is also no guarantee that any crowd-funded product will ever actually be delivered.  I tend to lean heavily towards the, "I'll believe it when I see it" camp but, in this case, the company has already produced a couple of more basic deadbolt products as well as 500+ units of this "pro" version of the smart lock.


I'm expecting my own unit to ship sometime over the next two months (!) and will update with my initial impressions when it arrives.  If this is something you're into and you decide to buy during the early-bird discount period, consider using the ATXGeek referral link below so I'll earn some credit towards whatever my next gadget will be.  Early-bird pricing is only available through May 31st.


(Thanks!)