107 karma · joined December 15, 2020
Note that I’m not commenting on how high performers are doing in terms of getting new jobs! It could be bad, it could be good, I don’t know.
But layoffs with less worker protections have always been the trade-off for higher compensation for top performers. Just because the understood risk occurred doesn’t mean that it was a bad risk to take, or even that it was much of a risk at all (especially if top performers have emergency funds that tide them over to their next job)!
All this to say that your rebuttal to the high performer argument isn’t super convincing. I’m sure it could be made better, but as it stands it’s not offering much.
There are certainly financial firms that people are less interested in though, namely banks.
From a business perspective it takes you out of the business of managing your own infrastructure. Even at big tech now we have to decide between managing our own infrastructure and “outsourcing” it to a managed service. Most advice I’ve gotten encourages the latter. It can cost more but if cost isn’t prohibitive it frees you up to deliver features and value for customers because you no longer have to develop (as much) expertise in ops.
To me the intellectual curiosity is inherent, and I don’t think I can, should, or want to tone that down. The best way to protect your time and remain recognized is to work less hours at the same efficiency/capacity, not to reduce the efficiency/capacity. It also helps to have management that is good and has trust in you. This way you can work your 10x style but deliver at a 3x timeline (or whatever x < 10 but greater than expectation) without the pressure to deliver at 10x all the time.
According to levels.fyi, 90th percentile in the Greater Seattle area [0] is $344-363k USD while 90th percentile in the Greater Amsterdam area is $162-170k USD [1]. Taking the mid point of both ranges, you make 2.13 times more in Seattle before tax. After tax, you get $244,933 in Seattle [2] and $118,078.95 in Amsterdam (with requisite conversions from/to USD when using the linked tool) [3], bringing the difference to 2.07x more take-home in the US.
The next step is to compare cost of living. I don't know which source is most reliable, and not all sources include international cities, but a Reddit comment suggested Numbeo so that's what I went with [4]. It reports that local purchasing power in Amsterdam is 18.1% lower than Seattle, while categorized prices in Amsterdam are anywhere from 18.8% (restaurants) to 39.3% (groceries) lower in Amsterdam. So it seems you would need to have to be paid quite a bit more than 50% of your US total compensation to achieve the same standards of living.
I'll spare the details because I didn't look too deeply, but I also did look into the cost of major purchases that people value - houses (3 beds) and cars (specifically the Polestar 2, which is an EV more on the luxury side). These things are definitely not half as expensive in Amsterdam as in Seattle.
So, I can only conclude that US workers do end up with higher purchasing powers despite less worker protection. While European workers do get a lot of benefits, many I probably don't even know about as I don't live there, Big N-tier companies (besides Amazon) usually have generous benefits like employer-paid health insurance that lessen the gap in benefits somewhat.
[0] - https://www.levels.fyi/t/software-engineer/locations/greater...
[1] - https://www.levels.fyi/t/software-engineer/locations/greater...
[2] - https://smartasset.com/taxes/washington-tax-calculator#L3MFa...
[3] - https://thetax.nl/?year=2022&startFrom=Year&salary=900&allow...
[4] - https://www.numbeo.com/cost-of-living/compare_cities.jsp?cou...
There might be perfectly legitimate reasons to be employed. Maybe they were aiding instruction of earlier students as they got more experience, akin to a tutor or teaching assistant in colleges. Maybe they took a temporary job because they were unable to find a job right out of bootcamp, and the reasons for not being able to get a job don’t necessarily have to be negative (e.g. they had familial commitments).
Getting a job from a bootcamp is already hard and it wouldn’t quite sit right with me to reject someone outright for something that could be explainable. Perhaps if you are drowning in applicants then you are forced to be harsher out of practicality, but even then surely there’s a better filter then someone reporting what is ostensibly an experience applicable to job they’re pursuing.
The package I was using had multiple artifacts per version, including the raw zipped source code and wheels for each ABI they built for. It’s certainly convenient that the most applicable one for your current system is pulled automatically by pip. But in this case that’s not the behavior I wanted, and I could not successfully get pex or pip to download a “less optimal” artifact.
You can blame C++, but if Python is already papering over deficiencies it’s not unreasonable to expect improvements that aren’t fundamentally impossible. It should definitely be possible for there to be one easily understood argument to specify which artifact pip should download. In my (potentially flawed) experience, figuring out which ABI value to pass was already not straightforward. Even then, the arguments for selecting an ABI or even asking for the raw source instead of a wheel weren’t seeming to have an effect.
I admit to being unaware of "versions inside versions" where a version may have multiple published ABIs that are not compatible across systems, but a nice packaging system would still make it easy for me to use platform X to build for platform Y.
One day the release pipeline broke because a dependency's authors published a new version containing a wheel for a newer ABI, though the older ABI was still supported. My build pipeline pulled the one for the later ABI because it was compatible, but the release pipeline environment, which could not be upgraded, did not have the newer version of gcc required. It was a nasty introduction to the fact that one Python package version can have multiple published wheels targeting different C++ compilers!
But the worst part was not that I learned something new! It was that I wrangled with both pex and pip for a few hours trying to figure out how to download a wheel for an older ABI as both tools resolve the latest compatible by default. The options to do so are ostensibly there but the effect was not. I definitely could have made some mistakes but in my mind it shouldn't be _hard_ to get it right.
I ended up downloading the wheel for the internal package in the build and copying it directly into the artifact. The release pipeline resolved public dependencies and installed the internal package from the local file.
This is awesome to hear! I did check out your website but wasn’t sure how closely it tracked with the GitHub method.
Honestly, I had a thought to build something like this after my GitHub interview but wasn’t sure how much traction it would get. It’s cool that you’re working on it and that my random thought was validated; best of luck!
Edit: to be clear, the interview was live with an interviewer. So it wasn’t a take home in the scheduling sense either.
I agree that take homes can be simplified with your suggestions above, and that certainly makes a better experience for the candidate. The hybrid format is also great - future interviews become an extension of your previous work, so it’s more comfortable than having to context switch for a new challenging problem each round.
I didn’t find this in your linked database, but I also enjoyed GitHub’s take home. I only recall spending 45-90 minutes on it, and the setup process was seamless. A recent blog post describes their approach here: https://github.blog/2022-03-31-how-github-does-take-home-tec...
1. Some candidates only need to brush up to be in interview shape.
2. To mitigate the need to spend time to prepare, simply use initial assessments to get back in shape. That is, apply to and interview at companies you don’t want to work at first to get more practice problems. If you pass, you get the added benefit of more leverage when negotiating your compensation!
3. As another commenter mentioned, as Leetcode style interviews are shorter, there should be a threshold where $prep_time + $leetcode_interview_time <= $take_home_interview_time. Of course, the threshold adjusts up and down based on your initial level of preparedness.
I recently interviewed with Ramp and enjoyed their model of practical problems, like a take home, but condensed into just one hour with less of the end-to-end expectations of a true take home.
[0] - https://www.theglobeandmail.com/real-estate/guinness-family-...
Regardless, I’ve not been soured by the slow intro of CarPlay. Android Automotive is sufficient for what I need the infotainment for (maps and music), and frankly I’m not surprised a Google built OS is having trouble integrating an Apple product. That they have plans to offer CarPlay at all via Android Automotive was a surprise to me.
Source: own a Polestar 2.
In addition, given an Expected Family Contribution (EFC) of $91k a year for assets == 1 million (plus other variables of course), there seems to be a decent amount of room to reduce the total assets <1 million and still pay the full price of $80k a year. Then even by strict definition, not only millionaires and billionaires are going to be paying full price.
[1] - https://docs.docker.com/engine/security/trust/#client-enforc...