The tricky requirement that ends up existing and torpedoing attempts to clean up feature flags is a requirement for long-term holdback.
e.g. "Test was successful so it's rolling out to all users, minus a 0.5% holdback population for the next 2 years"
This then forces the team to maintain the two paths for the long-term, ensuring the team might get re-orged / re-prioritize their projects sometime a year later making the cleanup really hard to eventually enforce.
The exact same point applies to Twitter's 3rdparty API getting killed off, and both moves are still a mystery.
It'd be easy to say "as a condition of getting this API key, you agree to display ad elements as they are served in the feed, and on click, open their associated URL in the system browser". All the ad-targeting is done server-side anyways, and attribution via unique links is easy.
Yes, it would be interesting, because you'd expect materials to cost roughly the same wherever you are in the country (modulo shipping costs), and cost of labor to account for much of the city to city difference.
A former landlord of mine, whose house I was renting in Palo Alto at the time (2021), shared that they were planning to kick off a major renovation that would total around 800K$ all-in.
That's an absolutely stunning figure to renovate a 3bdrm home, considering I've also heard anecdotes from outside California, of completely stripping down a similar-sized home to the studs, redoing all plumbing / electricity / walls / flooring / high-end-everything in the kitchen... for under 250K$.
So, where's the extra half-million dollars going? The delta in renovation costs alone between these anecdotes represents 10 years of the average California constructor worker's salary [per the BLS](https://www.bls.gov/oes/current/oes472061.htm).
Any hash calculations using a "read from stdin or a pipe" strategy, in my experience, is fraught with issues caused by an extra newline at the end of the input possibly being there today, and not in later checks, or vice-versa.
When people claim they wrote a prediction at some later date, they always have to document the EXACT command used to avoid this, e.g. `echo "smart prediction" | md5sum`
I'm shocked and disappointed at how expensive retrofitting insulation into a minimally insulated 1930s Bay Area house actually is. I estimate gas heating costs at around 2500$ / year (nov-march), while insulating floors, walls and roof would cost a shocking 60K$ for an average size 3bdrm.
If the insulation cuts heating cost in half, that's a 1250$ savings per year, meaning a 48 year break even time assuming zero discount rate. It absolutely doesn't make sense to do the work unless there's a massive tax subsidy or contractor costs come down (which is bad for the environment)
I thought I’d like the OpenComms but the Bluetooth stuff seems flaky. Not sure if this is everyone or just me, but if you pair them to 2 devices (e.g. a phone and a laptop), then shut down one of those, the Aftershokz will beep every few seconds thereafter to let you know you lost the connection, even while you’re using the other audio source.
It’s crazy: shut the laptop and walk away to listen to a podcast on your phone and they just beep beep beep.
Some product categories seem to have the middle of the market (quality wise) hollow out and disappear, so all you've got left is the low-end crap and high-end pricy versions, with no middle-range left.
One example: try to buy a nice metal or leather band for an Apple Watch. There's the low-end ones, which are 4$ on alibaba re-sold on Amazon for 12$. There's the X00$ Apple-made first party ones, and that's it. Nobody's making a really nice ~45$ leather band without the Apple name (that's obviously not just the cheap alibaba ones with a better marketing site).
The Palo Alto school meals are by no means healthy. At least at my kindergartner’s school last year, nothing’s really prepared onsite, it’s mostly microwave-in-a-bag fast food (factory made burritos, pizza, 2-ingredient sandwiches). Often this would come with a side of fruit (canned and sweetened) and crackers.
My kid would always bring a lunch from home but often return with it uneaten, because when you pit healthy home cooked food against microwaved pizza and crackers, for a six year old, it’s no contest.
I’m still supportive of the program - if there are starving kids in our community, of course having free options is great, I just wish they’d managed to have a cook onsite so it wouldn’t be so factory-made and artificial.
> Also, you can call the place in advance and just come to pick it up, no wait time
I suspect both the orderer and order-taker are both a bit happier to have orders come in online instead of taking calls. It enables a bit more asynchrony and batching, and removes the potential for miscommunication.
Big advantage of moving food ordering online is that it's gotten rid of trying to read off an order through the phone to someone who barely speaks English, standing next to the phone in a loud kitchen with dishes clinking and people yelling in the background.
How do you do present window sharing to a Meet call?
Using OBS gives you a very flexible way to stack a window capture with video sources, composite them, etc, yes, but your output is then routed through the virtual camera driver and sent to Meet as a regular webcam feed.
So because Meet thinks it's just a webcam feed, your coworkers will just see your shared window as a tiny video tile downsampled on the sending end down to 720p, and the only way they have of seeing it large is to pin your video tile manually.
Is there a trick here where you got OBS to output the composite feed to a chromeless window then you share that as a regular application window presentation?
> It’s frustrating, because you have sunk cost in entering the employment agreement
It's even more precarious for US immigrants where many types of immigrant visa (H1B, L1, TN) are tied to continuing to hold a specific job.
Don't like that job anymore? Get laid off, or reclassified into a different position? Oops, you're in violation of your visa, you have a week to leave the country.
There's a huge power imbalance between the class of workers that have US permanent residency or citizenship, and those on work visas.
I'm not sure how regulation can hope to really solve this when the problem isn't that an employer is forcing anything on you, but usually more one where your coworkers are independently making decisions about out-of-hours availability that create an unspoken culture, and usual and customary expectations that differ from the official written ones.
You're perfectly free to completely ignore work after 5pm! Nothing bad will happen! A company will swear up and down all day long that they support flex time and you're not expected to ever answer any queries outside of your own working hours!
But if a number of your coworkers decide of their own volition they're going to be in touch at all hours of the day and that becomes the default soft expectation, then you might look worse in comparison. Maybe not in the short term, maybe people will make a strong effort to un-bias themselves during performance reviews and not pay that any mind, but over time, you're sharing the bonus and raises pool with workers who valued their time less than you and perhaps ended up getting higher visibility for it.
Should regulation then mandate that your coworkers disconnect? Maybe. We do it for truckers, pilots, and flight attendants in safety-critical jobs: they're not allowed to work more than a certain set of hours regardless of whether they'd gladly do so.
I think of this more like: Just while you're not offering your work for sale in a country, then we all just agree as a society to look the other way as it's pirated. But you can end this get out of jail free card at any time by providing an official alternative.
Not doing this creates memory holes for artistic content. TV shows from decades ago with unclear rights ownership will be gone, disappeared.
How do you propose someone actually review VPN services? All of them work, have many countries available as egress points, and are fast enough.
The only actual differentiation is in attributes that are impossible to measure from the outside: Do they log accesses? Make logs available to governments? Sell it to data brokers for advertising purposes?
That's pretty much the only difference between good and bad providers, and there's no way to tell, so in-depth reviews are an impossible task.
Isn't that basically a "no true Scotsman" fallacy, though?
I can completely believe both sides of this: the original Fishkin article, and this rebuttal which claims that "actually, marketers are competent and knowledgeable".
One way to look at this, mindful of [Sturgeon's Law](https://en.wikipedia.org/wiki/Sturgeon%27s_law) stating that ninety percent of everything is crap, is that good marketing agencies will rise above, but if I own a store selling scented candles and Google "online advertising" and hire one of the random agencies listed, maybe, just maybe, I'll get one of the 90% of bad ones that will try to hoodwink me into believing unsupported incremental conversion numbers.
If you're in tech, things are changing so rapidly I don't think it's realistic to walk away for 15 years and then be able to come back and get hired as a practicing, competent engineer, unless you've been keeping up on personal projects in the meantime.
What's incomprehensible to me about the Tether saga is: the idea of a stablecoin pegged to USD is so incredibly useful as a bridge between financial systems... why go through all the trouble and risk of doing it fraudulently when you could... just not be crooked?
The basic idea seems sound: people want to transfer around cryptocurrencies (or speculate on them, fine), but actually selling them for USD is somewhat hard and peppered with regulatory issues. So swoop in with a partial solution: a stablecoin worth exactly 1$ USD that you can transfer quickly around using the blockchain, but doesn't have any price volatility so is a safe store of value until you can move out and redeem for USD through a slower, regulated offramp (of which there are fewer).
That seems like the hard part? They solved all the hard problems, and it's so useful that it should be possible to get rich just by charging small transaction fees, success just requires doing all that plus not being crooked and lying about the 1:1 backing.
As a child in the 80s/early 90s I played little league baseball. Our annual fundraising drive would get capped off by a random draw amongst all the players on the team, performed by the coach at home, and he'd tell us at the next game who won. Prize was 50$, a not-insignificant sum for 8 year olds in 1989!
Sadly, child-me didn't realize that it was a tad fishy that the coach's son won the draw two years running. What fantastic luck!
For some reason, the current crop of streaming devices all fail the use case of: 2 people want to watch a movie on headphones, without waking the kids. The current solution to this is airpods, where each person gets either the left or right ear. If there's one thing a premium device that costs twice what everything else does could do to differentiate itself, it would be supporting bluetooth pairing to multiple headphones at once.