albeec13

More like, "Why is Intuit so awful?"

My personal story on this:

I was a Mint early-adopter, and some time before Intuit purchased Mint I was also a TurboTax online user. I liked using Mint mostly just to track my net worth / assets and did so for several years.

Fast forward a year or two, and Intuit had purchased Mint. A few months later I came to realize that Intuit was now creating TurboTax accounts for all existing Mint users without bothering to cross-reference existing account email addresses between the two databases. When I signed up for Mint originally, it was using your full email address as a user name. So now, I had a Mint account which was my email address, an old TurboTax account which had that same email address on file, but a different user name, and a new TurboTax account which WAS my email address, tied to the existing Mint account.

So what happened? I tried to log into Turbo Tax using the account I already had used for taxes in the past to import my Mint information, but I was told that email address is already in use by another TurboTax account - the one Intuit created for me instead of merging my existing accounts...

Several phone calls later, I was assured the only path forward was to start using the new account to do my taxes if I want Mint integration. I decline and asked why they couldn't just get rid of the erroneous new account they created for me and let me combine the ones I already had. I was not about to lose years of tax returns on my old account to start a new account without that information. They said it was impossible.

Ultimately, I decided I'd delete the old Intuit/TurboTax account and start fresh, to take advantage of the integration features. They told me they were required by law to keep my old TurboTax account data, but that I could change the password to something random/long and change the email address to something non-existent to prevent myself or anyone else from being able to log in anymore.

At this point, I was very annoyed, and chose a suitably-annoyed fake email address directing hatred toward Intuit, which I won't repeat here, but rest assured it included a reference to Intuit and non-standard sex acts. I also changed the password to a 64-character random string, which I did not write down. Fast forward a couple weeks...

Lo and behold, I tried logging into Mint a few weeks after changing my Intuit password and wouldn't you know, that account was inaccessible now, the exact opposite of what I was told they were going to do... Better yet, I couldn't reset the password because I had changed the account email address on file to the aforementioned vulgarity. Now you see where this is going... I called customer service, and the rep assured me she could help if I could recall the email address I had used... So, I apologized in advance, and proceeded to tell her the address was "Eff"IntuitInThe"A"@"eff"you.com (something in this ballpark, and edited for language).

Needless to say, we had a good laugh. I haven't really used Mint ever since, and have been considering deleting my account (if I can even do that without destroying my TurboTax account...)

Frankly I'm shocked at how anemic the radar chip demonstration and marketing materials are compared to what they demoed about the chip a few years ago.

Google had video showing how they could differentiate individual finger motion and motions like rolling 2 fingers together to adjust volume up and down.

What they're shipping with Pixel 4 seems to have exactly one gesture: swiping a few inches above the screen to skip songs back and forth or silence alarms and calls.

You know what else had this feature? The 2013 Moto X. And it didn't use a radar sensor, just a group of 4 light sensors, one at each corner of the face of the phone.

Granted, if what they demoed a while back is accurate, the radar chip is much more capable than a few light sensors, but what they're shipping and marketing is not that. Either they were not able to make the proof of concept work in a shipping device, or the final product is not as capable as it seemed.

Yup, agree 100%. We bought a 2014 Mazda 6, and while it has a touch screen, I find myself only using the control knob 99% of the time. It's vastly superior to any touchscreen interface I've ever used, and also far better than some of the mouse-like joysticks in some cars (think later 201X Lexus vehicles and such).

I don't disagree with your comments about tradespeople's hours, but I'm not sure I agree with the reasons. A lot of those long hours are due to taking on more work/jobs to make more money. That is probably somewhat due to lower income per job as a tradesman, and somewhat due to the desire to work more "overtime" to make more money, depending on the going rates in the area of work.

Moreover, it's been my experience that people in non-trade work (particularly non-exempt STEM workers) often work just as much overtime and don't even get paid for the pleasure. Granted, their base salaries are generally higher, but a tradesman working similar hours can approach or surpass that. (I say this is I sit at 2:40am waiting for code to compile...)

Non-sequitor: I misread that first word in your comment as "Transgenders" and was flabbergasted halfway into that first paragraph until I went back and re-read it :D

That's a good point. I suppose as long as I'm careful with updates, I should be able to leave it alone until there's a long-lost security bug found.

Nice write-up. I have a very similar setup, only I didn't delve into the netflow montioring/traffic shaping because it seemed a bit overkill for my needs.

Have you given any thought about what you might do with respect to Shorewall, given the this news?

I've been running Arch as my main home server for going on 11 years now. I love it. In that time span, the biggest hurdles have been the sysvinit/systemd transition and a handful of issues with mongodb feature deprecations that required manual intervention.

My original impetus for using it was to "re-learn Linux" after a hiatus out of college, and because it's not as bloated as something like a full Ubuntu install, but doesn't require the full compiling of packages like Gentoo, it seemed like a good choice. Unless I want a GUI right out of the box, I don't use anything else.

As you noted, the wiki is fantastic as well.

High likelihood it was this: https://wiki.archlinux.org/index.php/Router

The Arch Wiki is pretty awesome, generally, and I used this pretty heavily as a reference. I will say, this was not 100% perfect and I had to use some other outside sources, but most of the info is here.

I did consider writing my own step-by-step post (if for no other reason than for me to not have to remember it), but haven't gotten around to it yet.

Welp. I recently completed an Arch-linux and PCEngines-based DIY router build and stumbled upon Shorewall as an alternative to straight-up IP tables...

I'm not looking forward to changing this setup.

Yep, I've been mulling some DIY work on replacing the headlights and adding foglamps... Despite appearing bright on my garage door, the actual street visibility of the standard headlights is pretty awful on this car.

I'm not sure about 2015 Mazda 3, but I know 2016 and later cars can be upgraded to support Android Auto for a fee at the dealer (I want to say ~$500, which may or may not be worth it).

Those aren't necessarily mandatory, but it certainly seems a lot of the safety tech is not available unless you also get things like that.

For reference, my wife and I bought a 2014 Mazda 6 Touring, which seems to be one of the few cars with a well-balanced mid-tier that doesn't cost an arm and a leg and still has good safety features (and did not come with heated/cooled seats).

With the exception of safety features (think blind spot monitoring, backup cameras, lane keep assist, parking sensors) and mobile connectivity features like bluetooth, frankly I'm getting tired of all the crap they're throwing into vehicles these days for many of the reasons you mention.

It seems every manufacturer or parent brand has their own infotainment system, NAV, etc. and you pay the price for it when you buy the car. And god forbid you want the upgraded safety features or leather seats, but don't want any of the other bells and whistles, because you can't have it without tacking on $4000 for the "technology" package.

My ideal car these days is one with the safety bells and whistles, limited luxury features (automatic and heated/cooled seats, climate control), but without any complicated infotainment beyond bluetooth/Android auto/Apple Car Play connectivity. There's no reason to have redundant technology built into cars by companies that aren't as good at it when we have perfectly capable, and generally more up-to-date mobile devices with us at all times.

Is anyone making reasonably well-appointed vehicles without throwing everything and the kitchen sink in there?

The title here, the title in the actual article, and the content of the article aren't telling the same story.

Here: "Toyotas and Chevys Are Holding Up Better Than Most Luxury Brands"

Article: "Toyotas and Chevys Are More Reliable Than BMW and Mercedes, J.D. Power Finds"

Article content (paraphrasing): "Lexus, Porsche top the list" "Most German manufacturers lag behind US"

So which is it? Bloomberg doesn't appear to be trying too hard here.

Good to know it's not just me. I've reluctantly been pulled into using the feed (as you say it has some nice features, and it's right there on my phone when I swipe right, so hard to avoid...), but this recent change has really put me over the top in terms of annoyance with Google's tracking behavior, especially in light of recent changes to Chrome.

I'm hoping it's just a bug/oversight, but I wouldn't put money on it...

Awesome project! I browsed through your headlessPi repo and saw how you configure hostapd and whatnot, but I'm having trouble seeing where you actually get the uap0 interface created. Have you tried this on a Pi Zero W yet?

That's as fantastic looking application! If I can find some time, maybe I'll consider helping out.

Thanks! Hope it works out for you. If so, I'll update the article to indicate it applies to RPi 3 as well (I don't have one to test with myself.)

Appreciate it!

Are you referring to the issues I had getting it working prior to my workaround script?

I would love to know as well. I mentioned it somewhere, but during boot, there are a few cryptic Broadcom driver errors which a Google search turned up little to know info on. It's hard to say, but it seems like something in the system has to stabilize before everything works.

Thanks for the info! I recall reading about that before, but forgot about it.

With respect to your question, I didn't post a log from iw dev after a reboot, but ap0 does appear after reboot.

Before proofreading my post, I originally included output from iw dev at the end to show everything working, but realized output from ip addr was more useful, since it shows actual network addresses and up/down connectivity. I may just re-add that for clarity. Thanks!

Unfortunately, the antenna on the RPi is not great, due to it being a PCB-etched pattern antenna.

For client, the range seems really good, thought not nearly as good as something like a cell phone, probably by about half.

For AP mode, the uplink speeds from the device connected to the RPi seems to be almost completely uninhibited, while downlink is pretty bad. This is where the poor antenna comes into play, since the RPi can't broadcast its SSID very strongly, and really only works well within LOS of the device. From another room, download rates drop pretty low, to sub-5Mb.

It may not work very well for transmitting video or something, but it's certainly good enough for low-bandwidth applications.

So I've tested it at home, where I have 2 different APs on 2 different channels (11 and 6, if that's of any consequence). I purposely rebooted the APs the RPi was using for internet access and it seamlessly switched over to the other AP without killing my laptop's connection to its own AP. I then rebooted the AP it was connected to after the first one had come back up, and it switched back again.

While this seems to behave sanely, you may run into issues where your device will drop session with the RPi AP due to "no internet connection". Assuming you're running Android, your device may ask if you want to remain connected to that AP, or it might just disconnect due to lack of internet. That's entirely device-dependent, but there's nothing inherently making this not work based on my instructions. It just might take some extra work on the device-side to make sure it doesn't try to drop the connection.

Thanks! Hopefully it works for you. I've got it going on 2 RPi Zero W boards so far, and it seems pretty stable. I actually wrote most of that post last night while passing all my laptop traffic through the RPi with no issues for over 3 hours.

Thank you. To be totally honest, I don't have a plan fully fleshed out, but the gist of it is this. I'd like to use a few RPi Zero W's as part of a home automation setup, primarily for either monitoring temperature/motion or for remote switching of lights/powered devices using relays.

One way to allow those devices to communicate with each other is via a central server (which I already have available), but it would also be nice to allow devices to communicate directly to each other in an ad-hoc manner. For example, Amazon dash buttons can be used to run arbitrary scripts from a router using dnsmasq when they connect to their configured access point. With the Raspberry Pi running an AP and dnsmasq, the dash button can directly connect without having to go through the server first. However, having the RPi also connected to the WiFi allows for it to access the internet (say to upload logs of activity, or register it's services to a central IoT server on the LAN or WAN) without requiring me to have to script in tasks to bring the WiFi interface up and down to switch between AP/client modes. This would also mean it would always be available remotely.

One of the downsides of the RPi Zero W is that it doesn't have a wired ethernet interface, so having both a WiFi AP and Client mode running at the same time allows it to have LAN/WAN access, while still being able to directly connect with other devices. The added benefit is that since both interfaces are wireless, it can go almost anywhere within range of WiFi, without cable restrictions.

Finally, having the ability to route traffic through the RPi would mean they could potentially be set up in a daisy-chain manner to extend the range of available WiFi access.

I've had a hard time finding resources on getting a Raspberry Pi Zero W to operate simultaneously as and AP and WiFi client. I pooled together whatever resources I found that failed me, but got part of the way there, and figured out the rest on my own. I figured it would be a good thing to share, since there appear to be a lot of people asking how to do it, and not a lot of solutions out there.

Great write-up. I find this system a lot more friendly than previous attempts I've made at game programming, usually involving derived classes of a an Entity class that posses inherited function calls for things like Draw(), Move(), etc.

By using data-only entities, it becomes a lot simpler to create Draw and Move classes which take an entity as an object and use the data therein to perform the action described by their class name. This makes it a bit simpler at the "main" code loop, because you don't need to pass handles, etc. down to the entities in order to draw using the game engine.