HNHacker News
TopNewBestAskShowJobs

vp8989

754 karma · joined January 4, 2019

submissionscomments
vp8989··on Go 1.17 Release Notes
Serving web requests with ephemeral workers introduces some new problems though. Generally efficiency in this space comes from pooling, reusing connections, multiplexing etc ...

Ie. amortize expensive work over many requests.

vp8989··on Go 1.17 Release Notes
Because GC will pause threads while requests are being served, this leads to latency spikes at the higher percentiles.

It introduces a whole distinct type of complexity into operating the server.

The latency of a request to your server is not just caused by the code that is ran to serve that request (which is how people intuitively would think about it), it can be caused by GC pauses that are happening because of the memory pressure caused by previous requests.

vp8989··on Go 1.17 Release Notes
"It feels like a missed opportunity"

I agree with that. I think looking back on it, Go arrived on to the scene at a perfect time where lots of people were rearchitecting so there was lots of natural appetite in the industry for a new language. Go vacuumed up a lot of this opportunity but ultimately hasn't really evolved the practice of software engineering all that much.

vp8989··on Go 1.17 Release Notes
"The Go team originally wanted to replace C++, especially inside Google"

Im glad you said that and I was not just imagining that myself. I remember following the Go language closely in its earlier days and it was often spoken about as a "systems language" but it actually seems to have ended up settling as a language to write servers for people who are sick of OOP but still like their imperative C style code.

vp8989··on Go 1.17 Release Notes
GC is quite problematic for web servers too. Lots of people "solve" that by giving AWS more money.
vp8989··on Google employees who work from home could lose money
"The European vision of living in dense, walkable neighborhoods, with easy access to quality amenities without the need to drive makes sense, to me at least, as the answer to this problem."

What vision? People in Europe drive cars. The cars are smaller and things are less sprawled out than in the US, but people still drive. Europe is not just central London or Barcelona.

vp8989··on Why don't tech companies pay their engineers to stay?
"in my experience most developers do their best work after about 6-8 months of starting a job at a new company, and that comes to an end after about 24-36 months of working at the same company. After that the vast majority of people just stagnate, get complacent, the job stops being interesting to them and they move on."

Thanks for sharing this, I think you are probably "right" even though what you said is not an easy/popular thing to say.

What is your observed sample size on this? Could you elaborate on how you see people stagnate, and why you think this might be?

vp8989··on In praise of defensive football
The evolution from this seems to be to overload just one side of the pitch (or effectively just a quarter of it, instead of the entire opposition half) while leaving a wide attacker on the other side for a quick switch of play.

The defending team is then forced into either committing more defenders into the overloaded side (and leaving 0 or 1 defenders to cover the lone winger) or staying symmetrical and thus likely not impeding the buildup enough.

This is basically how Italy played in the Euros, overloading the left and leaving Chiesa open on the right thus creating valuable opportunities for him to take on a defender 1:1 with lots of space behind them.

This is probably the best you can do if you want to have high possession without constantly running into brick walls.

vp8989··on Where has all the productivity gone?
This empiricism theater that so many engage in these days is just another form of CYA.
vp8989··on Message Deduplication with RabbitMQ Streams
"Exactly-once semantics is virtually impossible to achieve".

Impossible to achieve for a broker, yes. Combination of at least once delivery everywhere upstream and deduping in the consumer gives that consumer exactly once processing.

Lots of people don't grok that and they can be duped into buying certain products with claims that they offer "exactly once". Kafka does or used to do this, SQS (FIFO) also gives this false impression although they are a little more subtle about it. As soon as 1 vendor does it, they all have to do it.

"Did you hear Rabbit doesn't dedupe messages?!"

vp8989··on Migrating Facebook to MySQL 8.0
Are cross-server transactions scalable? Would anyone architect a system that did them en masse unless they absolutely 100% had to?

I am just speculating here but I doubt that YouTube migrated to Spanner so that they could do cross-server transactions. I would think consolidating the Ops burden into the GCP org and also serving as a trophy "customer" were higher on the list of reasons. But again, just speculating...

vp8989··on Migrating Facebook to MySQL 8.0
YouTube built Vitess, which lets you scale out MySQL writers. If I can scale out both my writers and my readers then what could be "vastly more scalable" than that?
vp8989··on Show HN: Wiser – minimal hypervisor boots Linux VM, written in C
Are you implying that Rust takes less time to "learn" (whatever that means) than C? Have you seen Rust?
vp8989··on We don’t use Kubernetes
The blog post itself explains it's existence:

"the following question comes up a lot:

“So… do you use Kubernetes?”

This has been asked by current and potential customers, by developers interested in our platform, and by candidates interviewing for roles at Ably. We have even had interesting candidates walk away from job offers citing the fact that we don’t use Kubernetes as the reason!"

Why do you think this is an "edgy blog post"? This is a piece of Marketing, both recruitment and regular. The implied subtext of the post is: "we don't use Kubernetes but its not because we don't know what we are doing".

Separately, what does having studied distributed systems or not have to do with Kubernetes?

vp8989··on The Idempotency-Key HTTP header field
They are 2 distinct concepts for sure, but if you also need to track each request independently then it would seem more robust to generate a separate Id for that from within the scope of your system. I don't think you would want to rely on a client to differentiate "physical" requests for you.

Maybe "Logical-Request-Identifier" and "Physical-Request-Identifier"?

vp8989··on The Idempotency-Key HTTP header field
It's more appropriate IMO to name it something like "Request-Identifier". All the client should indicate is that if you get 2 or more POSTs with that same Id, you are free to no-op on any after the 1st successfully handled one.

Idempotency is an emergent property of a system so it doesn't make sense that a client would dictate the key that's used to provide that behavior. So if the Idempotency-Key is quite likely not to be the actual "Idempotency-Key" in any sufficiently complex system, then don't name it that.

vp8989··on Don't Feed the Thought Leaders
This is absolutely spot on, great interpretation.
vp8989··on A Project of One’s Own
It's curious that you are defending a system that evidently provides poor social mobility, because the person who is trying to de-emphasize it's significance is rich.
vp8989··on Work Hard and Don't Burn Bridges
"Being scared of burning bridges reduces personal agency."

Totally agree. "Don't burn bridges" is also such common, banal advice that it's strange the OP is writing it as if it's a contrarian opinion.

For every story like the OPs there are probably hundreds or thousands of people who tacitly perpetuated bad work conditions instead of "burning bridges".

vp8989··on Auditing Io_uring
This reminds me of things like SOX, HIPAA where for any set of compliance/audit requirements there is a layer of people in the middle who always suggest to add all kinds of un-necessary things "just in case" it might be required. These suggestions are based on their interpretation or pre-empting of the compliance process rather than anything that is actually explicitly required.

Why can't the auditors explicitly state what needs to be auditable about io_uring? Instead of guessing and debating.

vp8989··on Stack Overflow sold to Prosus for $1.8B
Hah, bingo. I suspect all the people who left Basecamp will soon realize this when they finish their paid sabbaticals and move on to other companies.
vp8989··on Using Kubernetes to rethink your system architecture and ease technical debt
For the purpose of my joke, it doesn't really matter if it's incompetence or intentional under-engineering.

It was a satirical observation (based on my experience) that "the business" tends to scoff at unsexy foundational rework of existing tech. The trendiness of Kubernetes is used effectively to "trojan horse" that kind of work in.

I am jokingly suggesting that at this stage maybe that's the real value of Kubernetes.

vp8989··on Using Kubernetes to rethink your system architecture and ease technical debt
If you are a Director of DevOps, what is easier to say to the CTO:

"I need X million USD in engineering spend this year because we need to be on Kubernetes".

"I need X million USD in engineering spend this year to fix half a decade of awful technical execution and decision making (that all happened under your watch, BTW)".

vp8989··on My thoughts about the Principal role
I have personally not found the staff/principal level very enjoyable or satisfying. You end up being a hub for the larger team, helping everyone do their jobs. I'm not just referring to the coders.

It's very draining and I resent that the industry classifies this type of work as technical IC work. A small subset of these higher "IC" roles do get to focus on technical IC work, but for most people a staff/principal role is basically going to be that of a co-manager with no authority.

I've been considering down-leveling or getting into some software niche where technical expertise is actually valued and existentially necessary to the business.

vp8989··on “About one-third of Basecamp employees accepted buyouts today”
The details do matter, though. By omitting details you create a hypothetical position that is less reasonable and easier for you to criticize. In this particular situation, this type of thing is particularly harmful because it paints the supposed originators of this made up position as something they are not. You could arguably consider it as libel by omission.

I don't even mean to single you out because I've seen this type of thing countless times on here and especially on Twitter over the last week or so.

"What's more interesting is the awkward situation they have created, which I'd argue is a false dichotomy - where does "societal politics" end and "politics that directly relate to our business or products" start?"

That can be tricky, which is why the policy is not zero tolerance and they already anticipate and are OK with people sometimes getting it wrong.

vp8989··on “About one-third of Basecamp employees accepted buyouts today”
"I'm not sure it's right to say that any discussion about anything political is banned when a company policy that affects an employee can be political in nature."

Why do you think that's the policy? Did you read the blog posts?

According to DHH's post, political discussion that is related to the business is allowed. Any other political discussion is encouraged but must take place in a dedicated non-work channel. The policy is also not zero-tolerance.

vp8989··on I hope work from home continues
"sounds to me like raising a baby using a six-axis robotic arm"

No it's like raising a baby using advice you read from books that are peer reviewed and revised over time vs. just listening to what your parents say.

vp8989··on I hope work from home continues
Engineering culture should be disseminated via documents rather than oral lore.

If it's disseminated via oral lore it's expensive to change and is often misinterpreted across a large enough org.

A silent killer of engineering orgs is bad practices that stick around because that's the way it's always been done. No one knows why because it wasn't written down. The only way to fix that is for someone high enough to provide blessing, but maybe none of those things are individually disruptive enough to get the CTO involved. So everyone eats the costs in perpetuity.

Much easier to just write everything down.

Increased difficulty for juniors to ramp up when you are remote is a symptom of a culture with lots of overly complicated processes that are not well documented.

vp8989··on U.S. rent has increased 175% faster than household income over past 20 years
Rent to buy is a good strategy because you can justify to yourself living in a crappier place than you would want to own outright. Over time that allows you to purchase a house at a lower total outlay, including the rent you paid. During that saving period you also had the additional hard to quantify benefit of being able to move quickly and cheaply.

If you are doing rent to buy but living in frothy areas where rents are always rising and apartments rent for 4k then obviously it's not going to be a winning strategy for you.

vp8989··on Ask HN: Is Google still a cool place to work?
I don't think these problems are exclusive to Google or Google-sized companies.

Once devs have worked on a particular system for long enough, they become a nuisance because they know all the (good) reasons not to do something. Not necessarily because they are lazy / don't want to grow.

← PreviousPage 3 of 7Next →