Just don’t even think about buying a Fitbit Versa

Or at least, buy it from Costco, and try to remove / replace the band before leaving the parking lot, so you don’t have to make an extra trip to return it… because you WILL return it.

Google “fitbit versa band problems”.

Seriously, worst wristwatch band in history. Tiny little bend/breakable pins and a bizarrely difficult ‘push down to release band’ pin that can’t be reached because the band is in the way.

(Oh yeah, still no calendar… “so-Called-Smart” watch with no calendar. Still does not allow you to change the volume on some devices like Apple AirPods when listening to music.)

HemmorrAndroids even more painful than IOS

Daughter called from a borrowed phone because her Samsung Galaxy S9 bricked itself during a routine update. So she gets to drive an hour to the nearest place that can re-flash her Android ROM.

Yes, apparently that shit still happens in Android-land.

Now don’t get me wrong, I despise IOS since IOS 9. And Apple’s jack-booted-thugs killing the ability to install older versions of IOS that actually worked. Especially because Apple software quality has gone right into the shitter.

But I guess Apple can get away with it as long as Google/Android continues to suck so bad.

Making Heroku CI (almost) as fast as running Rspec locally

Once Heroku support showed us how to specify the Postgres version for in-dyno Postgres in our app.json we re-ran our Rspec rspecs and were pleasantly surprised how fast they ran. (Specifying Postgres 9.6 seems to be required for Rspec for a Rails 4.2.11 app; using the default Postgres 10 causes errors related to the change of the increment_by field in P 10)

 Mac OS XHeroku CI "M" dynoHeroku CI "M" dyno
IN-DYNO postgres
Heroku CI "L" dynoHeroku CI "L" dyno
IN-DYNO postgres
Dyno cost/hour$0.34$0.34$0.68$0.68
Run Time5:5359:0013:27 !!!31:438:03 !!!
Heroku cost$ 0.34$ 0.08$0.36$0.09

The “trick” to specifying the Postgres version for in-dyno Postgres is to use an environment variable. If this is documented anywhere we could not find it, so perhaps this will be helpful to someone else:

In order to change the Postgres version for the in-dyno add-on you’ll need to create a new environment variable in your app.json like this:
  “environments”: {
    “test”: {
      “addons”: [“heroku-postgresql:in-dyno”],
      “env”: {
          “value”: “9.6”



Heroku Deploys via Github are not faster than using git push

While we did not expect to see any difference, we did confirm that our 10 minute deploys to a Heroku staging server typically take the same amount of time whether we use the old git push vs Github integration.

We did not test whether a deploy with significant new data (a ton of new images added to the repo for example) was materially faster from Github.

Heroku CI is remarkably slow, but half as slow if you use the largest dyno type

UPDATE: using in-dyno postgres puts Heroku CI on par with running Rspec locally!

Experimenting with using Heroku CI for a Rails app, we were a little surprised just how slow Rspec specs run.

As you can see however the runtime cost is about the same since the largest dyno performance-l costs twice as much as the default dyno used by Heroku CI.

Unfortunately, we cannot test using in-dyno Postgres since Heroku does not seem to offer a way to specify the Postgres version for in-dyno instances.

 Mac OS XHeroku CI "M" dynoHeroku CI "L" dyno
Dyno cost/hour$0.34$0.68
Run Time5:5359:0031:43
Heroku cost$ 0.34$0.36

Next Step: If we can get parallel tests to run, it will be interesting to see if the overall runtime (and cost) remains constant so we can shorten the test to say 2 minutes by running 16 dynos.

If so that may be GREAT news for using Heroku CI vs Travis CI or Circle CI, whose fixed monthly cost is HUGE for even a few concurrent instances.

4 Most Remarkable Shortcomings of Apple Watch Series 4 (e.g., why I returned it)

That’s a lie. It’s NOT an AirPlay source:
See https://support.apple.com/en-us/HT208728

Ooops. Even on version 5 of Watch OS,
Apple forgot to ‘mirror’ your iPhone’s

default alert setting to your Watch, or provide
any other way to add an alert to an appointment.

For someone who uses Macs and iPads and iPhones, there’s a lot to like about the Series 4 Apple Watch (GPS + Cellular, watch OS 5.1.2).

But I returned mine to the Apple Store after only 3 days because:

  • The Deal Killer: Cannot create appointments with “alerts” – if you use the Watch to create a Calendar appointment, there is no way for the Watch to add an alert, such as an alert 30 minutes beforehand. So your appointments created with a Watch behave differently than the ones you created with your iPhone. (If you create an appointment via an iPhone, or an iPad, or a Mac you can set a default alert, but not with the Watch.) This was confirmed today by Apple Support who escalated the issue to engineering.
  • Calendar displays only current month. Of the many bone-headed things Apple has done in the post-Jobs era, a Calendar app that will not display months other than the current month deserves special mention. If you display the month-at-a-glance calendar, and turn the scroll knob, or swipe, the Watch displays… the same (current) month.
  • No 5 GHz Wifi support  – had to dumb-down my home wifi network to 2.4GHz for the Watch to connect correctly to the iPhone, even though it said it was connected it wasn’t connected to the iPhone via Wifi. Apple documentation is silent about the need to dumb-down your Wifi to 2.4GHz… One symptom is ridiculously slow music syncing when adding playlists to your watch if your iPhone connected via 5GHz and your watch connects via 2.4GHz… but the iPhone offers NO way to “see” what Wifi speed your iPhone is connected, so good luck debugging this.
  • No Airplay support – despite offering an “Airplay” option right on the Control Center when selecting audio output, the Watch is NOT on Apple’s list of Airplay audio sources. Confirmed with Apple Support on lengthy call: if you have a nice stereo connected to an Apple TV, no, you cannot stream audio to the Apple TV’s sound system. So the “Airplay” option on the Watch is a lie… iPhone and iPad and Mac Airplay options DO connect to my Apple TV (audio and/or video) the Apple Watch will not show an AppleTV among the audio destination options. (I thought perhaps it was because mine was a 3rd Gen AppleTV but Apple Support said no version of AppleTV can play music from a Watch.)
  • BONUS Shortcoming: No Phone or Messages quick-launch option for the new Infographic faces. Oddly, the fancy new faces do not support the two most important options (which they call complications)… no phone and no messages.


How to allow two Vagrant apps to talk to each other

An issue we ran into recently is: if the same development machine is running two Vagrant instances, how can an app running onfoofetch data from a url onbar.

The two-part ‘trick’ if foo wants to fetch data from bar, is:

1) each app’s Vagrant file needs a line:

config.vm.network :private_network, ip: PVT_NETWORK

where PVT_NETWORK is a local IP, is different for each Vagrant file, and probably needs to be in the same subnet. For example PVT_NETWORK might be (foo) and (bar)

2) foo accesses bar via the PVT_NETWORK IP address not the “real” IP you would use with a web browser.

In our Rails example, we have each app running on a different port, so foo is on localhost:3000 and bar is on localhost:3001, so foo would access a url on bar via

JibJab irresponsibly, ineptly, and possibly illegally blasts ecards

I despise irresponsible mass-emailers like JibJab.

A truly unfortunate real estate team in California was horrified yesterday to find that JibJab (jibjab.com) pissed off legions of their customers as a direct result of JibJab’s inept and irresponsible practices (which are also, in my non-legal-expert opinion, illegal under CAN-SPAM).

The agents’ intent was simply to blast out a holiday card to their email list via JibJab. It became, in their words in the apology they then had to email to their entire email list, a “debacle” and “nightmare”.

By irresponsible (and possibly illegal), I refer to the fact that when JibJab sent an email on behalf of the agents, they did not provide any opt-out link. In my opinion, that seems to be a fine-worthy CAN-SPAM violation because these were not transactional emails.

By inept, I refer to the fact JibJab blasted many duplicate emails to the same recipients, which in the words of the agents made “tons of people angry at us.” (Even if we allow the possibility – although I have no reason to believe this is the case- that the agents somehow submitted duplicate requests to send the same ecard to their list, JibJab is still in my opinion woefully incompetent for allowing the same card to be sent to the same recipient from the same account multiple times.)

As an example of JibJab’s ineptness, here is a copy of our inbox for one day:

Naturally, JibJab does not answer their support phone (the greeting tells you to email them), and they (thus far) have ignored both email we sent alerting them of the dual issues of repeat sending and no opt-out mechanism. (I tried contacting JibJab within minutes of the first duplicate, so they could have saved countless pissed off recipients had they offered real-time support option.)

Ironically, JibJab’s most recent tweet boasts their product “keeps on giving” but I think we can all agree that blasting 29 duplicates isn’t something to boast about.

I can only wonder how many hard-earned customers and prospects JibJab customers have lost due to JibJab’s ineptness and irresponsibility.

For healthcare plans, Bronze is less expensive except in a narrow range of care usage

Short version: the $200/month additional premium for a Silver plan (vs Bronze) almost never gets recouped through the lower copays. (Remember, although you hit your lower Silver deductible sooner, you still pay 20% of all costs until you hit the out of pocket maximum, which on the Silver does not happen until your expenses hit something like $12,500. But on Bronze, once you hit $4000 in expenses your coverage in 100%. The extra premium for Silver, combined with paying out of pocket until you hit $12,500, is the one-two Silver sucker punch.)

Long version: For those of you deciding which ‘color’ version of healthcare plan to choose, there is a very high possibility that the Bronze is less expensive. This guy does a detailed analysis which, despite being a few years old, definitely applies in my case. http://ideas.4brad.com/surprising-math-obamacare-levels-go-…

I did my own spreadsheet for our employer BlueCross plans and found that by the time you factor in premium costs, Bronze was less expensive for anything other than the case where covered medical expenses fell into the narrow range of 4000-6000 for the year (and even then, in that narrow range where Silver was less expensive, Bronze only cost a couple hundred dollars more than Silver).

Emotionally, the “intuition” that it sucks to pay 100% of expenses until we hit the Bronze deductible ($4000 in our case) is simply wrong. The numbers were unambiguous, the Bronze wins.

So here’s a good question (echoed by the guy whose analysis I linked above): why in the hell don’t employers or insurers or the various HCA websites provide a chart or calculator to “compare” the real TOTAL cost of each plan under different care usage scenarios?

Here’s the results of my spreadsheet for BCBS Bronze, Silver, Gold.

The top chart is the usual case where only ONE subscriber incurs the medical expense on the X axis (and the other subscriber is generally healthy), the bottom chart is if BOTH subscribers use the same level of medical expense.

Caveat: If your “member cost” (the insurance allowed amount) for prescriptions is significantly higher than your Silver co-pay, that difference in effect “adds” to your premium cost for Bronze. For example suppose you are on Silver or Gold right now and every month you take XYZ, and on your insurance claim each month you see the Provider (retail) cost for XYZ is $600, the insurance allowed cost $320, member pays $20. Your copay benefit in this case is saving you $300/month vs the normal insurance-allowed rate for that drug.