727374

"If it bleeds it leads". -famous newsroom adage. This is true for all news and media, always. Humans are drawn toward stories that arouse fear and negativity.

It largely depends on the market you're in and timing. In SF Bay, things are extremely skewed toward renting right now because so many people are sitting on houses benefiting from ZIRP and locked-in prop taxes. It's probably 50-70% cheaper to rent on per month, even considering building equity. We own and are looking to rent a bigger place because it's much, much cheaper. I'm surprised more people aren't talking about how incredibly skewed things have gotten. The problem is there's a lot of "high-end" rental stock in our area, but it tends to be outdated.

NYTimes has an excellent calculator for informing your own rent-vs-buy decision based on factors like rent, purchase price, marginal tax rate, interest rate, etc.

edit: adding link - https://www.nytimes.com/interactive/2024/upshot/buy-rent-cal...

Smart watches for kids are great compared to alternatives. My kids have Apple Watches and aren't asking me for phones/tablets (and the related evils like IG, TT, etc). Love to see this from Google (and Apple).

Scams / Identity theft cost people 1 Trillion dollars a year (https://www.theverge.com/2024/5/14/24156212/google-android-a...). This is a huge problem for the elderly. I don't see telecoms preventing this adequately. So huge kudos to Google on this. Not sure what value Google would get from “Google will use your phone calls to train their AI”. They already have a huge corpus of YouTube speech data.

Or if you want to eat it right away, make a carpaccio - https://youtube.com/shorts/WZ_orsakLG0

Sorry for the staff who’ve lost their jobs. I did a phone screen with one of their recruiters a couple years back. They told me they couldn’t afford my (slightly above market rate) salary, but went on and on about how the company enjoyed a very “comfortable” work life balance. WLB is great, but I’ve never had a recruiter spend so much time on that and so little on the “exciting” product/role. That was a red flag to me at the time.

Dataloader (https://github.com/graphql/dataloader) eliminates many n+1 issues. Basically, entity requests are queued up during evaluation of the query and then you can batch them up for fulfillment. It’s a great addition to other asynchronous contexts as well.

WRT random requests, there are libraries for query cost estimation you can use as a gate.

+1 to all of these, especially #1.

A lot of companies will have a career doc that lists expectations for each role. Take a look at the manager version and make sure it is actually something you want to do and then start doing some of those things. For example - lead a small group on a project, mentor someone who needs help, implement a change that will improve team health, reduce chaos on a project, align stakeholders, etc.

Tell your manager you’d like to explore that career path and frequently ask them what they need help with. If you don’t have a good relationship with your current manager, you may consider finding a new one who will help you grow. Look for a manager that is ambitious and looking to develop someone they can delegate to and you can ride their wave.

Everywhere I’ve worked in the last 20 years has needed strong managers, including startups and FAANGS. If you’re slightly patient and show some signal of readiness an opportunity will arise.

Yeah a lot of places are doing this. My old company calls it "ad hoc" office attendance.

One thing I've always appreciated about Amazon is its willingness to experiment with everything. Offering reduced hours for some would hit so much resistance at most large corps and never get off the ground.

My definition is "Finds and creates the most valuable work for other engineers to do."

I really started to see the value once I moved from Principal to Manager. All of a sudden I was busy with boring manager stuff and needed an IC who could get deep into the details of our initiatives and make sure everyone's efforts were lining up. You don't need the title of Principal to do this role, of course, but ICs who see the big picture and understand what's important for the business are invaluable.

The other points in the article are on the money as well, especially around modeling and amplifying a good dev culture for the organization.

Nomad is the BetaMax of container orchestration. Better quality than K8s in a lot of ways (supposedly), but there's no ecosystem around it so it's a moot point. Want a continuous delivery system that plugs into Nomad a la ArgoCD or Spinnaker? Or a serverless platform that can run on top of it. Or access to people that know how to use it. Be prepared to write your own Nomad bindings and train your team on it.

I went down the Bloom Filter rabbit hole for a project a while back and then wondered if I could actually fit the entire set space in memory. The set was IDs up to 2^32, so basically a giant bit array. I believe the IPV4 universe of this article is actually the same size as in my project. I coded up my project using Java bit arrays and got things working decently well, using a big heap. Then I found out there are a bunch of compressed bit array libraries such as EWAH and Roaring Bitmap. When I substituted Roaring for the stock Java Bitset implementation, I saw space and computation improve by many orders of magnitude. Roaring uses a bunch of tricks to achieve this, but it mostly comes down to encoding large runs of 1s or 0s. Obviously, not as tiny as bloom filters, but still pretty small for most modern machines if you have a sparse set. https://www.roaringbitmap.org/

As a customer, I see a lot of value from Apollo's SAAS, which helps in "productionizing" graphQL. GraphQL has more sophisticated instrumentation and access control needs than simpler APIs like REST/RPC and there are few if any open source projects that can do this. It was a pretty easy buy vs build decision for my team. Granted open source alternatives may pop up at some point.

Every site that needs a CDN probably already has one, because it's a quick win. For every site that doesn't need one (small user base, small asset size, etc) it's likely not worth the added complexity.

I agree it could do better explaining what Jewelbots are, but I wouldn't assume there's a dedicated "marketing team" for this product yet.

"They’re wired behind the neck but they’re every bit a competitor to Apple’s AirPods"

Umm, no. The killer feature of AirPods is the form factor and not having an annoying cable connecting the buds. As other posters have noted, Amazon is full of cabled headphones. Language translation is certainly cool, but for most people that's not an everyday need.

How ridiculously good is Amazon at getting free press? They have already likely decided on where to put the new HQ (1), and will get featured in these overwhelmingly positive articles for months to come. I'm also reminded of that 60 minutes segment on drone delivery that just happened to air right before cyber Monday a couple years back. Still waiting on that one, although I do actually think it will happen eventually.

*1 https://www.brookings.edu/blog/the-avenue/2017/09/08/amazons...

The sensationalism and missing facts/context of this story is truly precious. I'd love to hear the specs on Ms Lile's Manteca house. I'm guessing it's significantly bigger than the median house size in SF or neighboring towns, where she could surely afford to live in a 2BR bungalow or condo.

Annoying. I was happy paying $100/yr for Dropbox to store my photos along with a few extra files. The photo features weren't great, but good enough for basic browsing through galleries. I bought into the idea that if I was paying for the storage it would be more durable and permanent than some fly-by-night photo startup or google/amazon feature. I realize the photos aren't going anywhere, but my use case is. Canceling my subscription to Dropbox... now.

Yes, this and also parallelizing disk I/O. For example you could fit a 5TB table on a single machine, but if you have an operation that requires doing a full scan (e.g. uniqueness count over arbitrary dates), that will take a very long time on one disk. Yes you could partition into multiple disks, but Hadoop offers a nice generalized solution.

Yikes. Gambling is just a dumb idea in general

Or, a thrilling and potentially expensive form of entertainment. I don't feel it's dumb when I lose $200 at the blackjack table when I'm in Vegas for a bachelor party.

If people like this, also check out his course at Udacity - https://www.udacity.com/course/design-of-computer-programs--...

Sales plummet 22% in SF, yet there's a bidding war on every property? This implies that demand still outpaces supply so I don't see how you could predict a crash, yet. In reality, prices only dropped 1.8% y/y. And that was for March, which is traditionally a low-inventory month. I personally think the correction is still a year out.

CA property tax increases are limited to 2% a year and only get re-assessed after a change of ownership or new construction.

https://en.wikipedia.org/wiki/California_Proposition_13_(197...

Agree based on my experience at Amazon. 95% of the devs I met were extremely professional. Very little joking around, beer drinking, ping-pong, etc compared to other places I've worked. Granted 90% were white/asian males, so not a lot of diversity. Diversity issues - yes, brogrammers - not IMO. Brogrammer seems to be a slur for white males in their 20s.

Really? This article sounds very over simplified, but I haven't worked at google so I wouldn't know. I'm assuming if you want to change some much depended on library, there's a way to up the version number so you don't hose all your downstream users. That's the way it worked at Amazon at least. Also, I wonder why the people in the story think Google's codebase is larger than that of other tech giants, not that it really matters.

Or sick people eat less spicy food?

Also, in addition to the other responses, excel is great for quickly modeling problems/systems with lots of inputs and outputs and calculations in between e.g. discounted cash flow.

Half sarcastic, maybe. But, genuinely curious about how they are managing the risk around scoping/estimation. Perhaps using overly conservative estimates or else assuming that on average their estimates will be close. Or maybe those Google/MIT engineers are magical.