HNHacker News
TopNewBestAskShowJobs

jakewins

3,664 karma · joined March 9, 2009

I build hardware for the energy transition at https://www.fluxrite.com

I previously built virtual power plants at Tibber, databases at Neo4j and fleet management Equipmentshare.

jake@davis-hansson.com https://tech.davis-hansson.com http://github.com/jakewins

submissionscomments
jakewins··on [dead]
Yeah this just seemed like people making rude (“designed by monkeys”, really HN?) unspecific statements about a standard while the examples were about a library they didn’t specify..
jakewins··on Vector: A high-performance observability data pipeline
My experience with using Datadog at (some) scale was that they focused on making it really, really easy to integrate their agent with your apps, and then once they had a large base of users with high switching costs they started rapidly raising prices.

In other words: My claim isn’t that they were better at ingestion but at onboarding and at creating switching costs.

Since that was how their leadership acted last time I used their code, I expect the same leadership to act the same way again with this other piece of code they own.

jakewins··on Building a streaming SQL engine with Arrow and DataFusion
Maybe I’ve not used DuckDB enough but: How would you set it up to consume, say, a Kafka stream, apply some transformations or windowed aggregates and continuously output the result?

In my experience DuckDB is a batch tool, vs this thing which is a continuous stream tool

jakewins··on JetBrains' Temporary Solution for macOS 14.4 Issue with Java
It makes a lot of sense for them to try though - the obvious reaction here is to switch to VSCode, devs using JetBrains can’t wait for Oracle to fix this. I imagine JB are churning a lot of users because of this
jakewins··on Building a streaming SQL engine with Arrow and DataFusion
This is a different use case - DuckDB is large queries over your entire dataset; Arroyo is continuous queries executing over a live data stream.

So if you have a stream of sensor data from a bunch of IoT devices, you might use Arroyo to run “live” queries over the flowing data, and you might use DuckDB to do analytical queries over archive data

Although they are both hammers to whack nails with I guess, so a lot of use cases likely work well to solve with both systems

jakewins··on Vector: A high-performance observability data pipeline
Yeah the trick is if they can lock you into a stack that sits everywhere in your apps, it’s very expensive to switch vendors, letting them extract high rents. This is what happened with the Datadog agents.

In that context, OTEL is an existential threat, because it makes them a commodity. Then it becomes relatively clear why they wouldn’t put OTEL support in the Vector roadmap.

jakewins··on Vector: A high-performance observability data pipeline
To deliver a cross vendor narrow spec, you mean?

Given my experience with Datadogs pricing lock-in-and-switch, yeah 100% I’d rather run agents that allow me to pick the collection backend than another tool from Datadog.

jakewins··on Vector: A high-performance observability data pipeline
But this is a vendor specific forwarder? Isn’t this just Datadogs attempt at slowing down community focus on OTEL?
jakewins··on Vector: A high-performance observability data pipeline
What aspects were you missing from OTEL? We swapped the agent side out from New Relic to OTEL in NodeJS, .NET and Python - I’ve not found any major missing feature?

Or are you thinking more on the UI/analysis/collection side?

jakewins··on How Figma's databases team lived to tell the scale
Separate logical databases, within the same RDBMs, so sharing CPU, disks, RAM etc
jakewins··on How Figma's databases team lived to tell the scale
The problem - conceptually - is made much simpler this way; we make use of this at work.

However you will still have shared resource problems - some rogue query destroys IOPS in one tenant now ends up bringing down all tenants etc. There are in theory databases that solve this as well, but my experience has been that at that point what you buy into is a bad version of resource sharing - ie what an operating system does - and you’re better off using OS mechanisms

In other words: yes, but you still have noisy neighbours, and may be better off running lots of small fully separated DBMSes than a big logically separated one

jakewins··on Show HN: Creating custom coloring pages from photos. Great for parents/teachers
I think this is just a thin wrapper around GPT-4 or similar? Parents I know have replaced colouring books by just asking GPT-4 , and the colouring pages it generates look exactly like the example outputs here.
jakewins··on MicroZig: Unified abstraction layer and HAL for Zig on several microcontrollers
Agreed! I’ve similarly had good experiences with Rust on the rp2040, using embassy which builds on the rust HALs.
jakewins··on How Netflix really uses Java
Not sure what the poster above was thinking of, but it seems kinda the same as every other language that’s adopted it - powerful, but footguns abound. I ran some async C# in a debugger - in Rider - the other month, and the debugger just goes off the deep end.

Does C# have the same issue Python does with accidentally calling blocking code? In async Python - which I mostly quite like actually - it’s terrifying to bring in a library because if you’re unlucky it does some blocking network call deep inside that stalls your whole app..

jakewins··on How Netflix really uses Java
This was part of some “testing in production” type scheme?

Are there any talks or blog posts on how y’all wired this up, lessons learned etc?

jakewins··on [dead]
The thread that proposed this change may be a better source of context: https://lobste.rs/s/ttmyvf

David mentions as an aside “this place has other mechanisms for dealing with [spam]”. A lot of the complexities around “karma” - far from all of course - relate to spam. But lobsters has a separate giant hammer for that nail, so the karma feature complexities for lobsters becomes quite different. Not necessarily harder or easier, but certainly not the same compared to a site like, say, SO, where karma is central to who gets to take what actions.

jakewins··on Institutions try to preserve the problem to which they are the solution
One of my first tasks at Equipmentshare was automating invoice generation, and we did a lot of that basically pair programming with one of the back office specialists that did that work manually - it was super, super fun, we made really good friends, and now, ten years or something later, she’s a manager overseeing whatever systems replaced what we built.

But it was driven by both sides being made clear from the beginning: nobody is losing any jobs here, the goal is to 10x the number of accounts we could do per back office person; their new jobs will be overseeing the software and dealing with edge cases.

I’m not sure this would’ve been possible to do in such a way if the company wasn’t rapidly growing though.

Makes me wonder: what are things that are easier like this in orgs that aren’t in growth phases?

jakewins··on In Defense of Simple Architectures (2022)
> Python and TypeScript by definition don't have a concept of const generics the way it applies to C++ templates or Rust.

Can you help guide me where I'm misunderstanding the type system aspect of const generics that's missing in Python? What I meant was that in Python I can say something like this:

    import typing as t

    class FooEater[NUM_FOOS: int]:
        def eat_foo(self, num_foos: NUM_FOOS):
            print("nomnom")
    
    one_foo = FooEater[t.Literal[1]]()
    one_foo.eat_foo(2)  # type error, NUM_FOOs for one_foo only allows literal '1'

And a PEP695-compatible typechecker will error at build time, saying I can't call `eat_foo` with `2`, because it's constrained to only allow the value `1`.

I admit to not being a type system specialist, but at least as I understand Rust's "const generics" the type system feature is: "Things can be generic over constant values". Isn't that the same as what Python supports?

jakewins··on In Defense of Simple Architectures (2022)
Sorry - can you put the snark aside and answer the question then, it seems given my lack of understanding here as a mere Go simpleton it should be trivial for you.

The question I responded to said: "Better static analysis, for one."

I asked: "What is something you can express in the C# type system that MyPy cannot statically check?"

Can you provide some examples where C#'s type system can express something that modern well-typed Python cannot type-check?

jakewins··on In Defense of Simple Architectures (2022)
We have services in Python and C# at $dayjob, and my experience is that Python has mostly left C# behind here - what is something you can express in the C# type system that MyPy cannot statically check?

Conversely, I find myself constrained by C#'s typesystem not having features now widely popular from Rust and Typescript and available in modern Python: const generics / literals, protocols/structural interfaces and matching over algebraic datatypes, to name a few examples.

jakewins··on In Defense of Simple Architectures (2022)
The author is ex Microsoft, Google and was a senior staff eng at Twitter; i don’t know them but my experience with their blogging over the last decade is that they seem generally very well informed. To me it seems unlikely “the penny hasn’t dropped”, like you say.

On the actual criticism you’re raising: In what concrete way would moving to micro services help with reducing data integrity bugs? In my experience I’d expect the exact opposite - micro services systems generally being harder to test and generally having more possible interleavings and partial failure paths.

jakewins··on Diseconomies of scale in fraud, spam, support, and moderation
I’m being unnecessarily confrontational because I’m having a bad day, sorry, that’s not helpful.

I would say this: I agree with you that massive faceless monopolies with no accountability are terrible, and often a direct consequence of economies of scale.

It is also, simultaneously, true that the labour saving effects - human hours spent per calorie of food or pair of socks - is a number of orders of magnitude such that modern society (ie the not starving/not freezing part of it) could not exist without it.

These two things are simultaneously true, which is what makes the problem hard; we want one effect of economies of scale and not the other.

jakewins··on Diseconomies of scale in fraud, spam, support, and moderation
Sorry I’m having a crap day; you’re totally right there’s a ton of gradients between “one mega factory makes all t-shirts” and “we all make clothes ourselves with hand made cotton thread.”

I think my argument was mostly like - don’t throw out the baby with the bath water; economies of scale are a real thing that brings us real benefits, and that should go in the scale too.

jakewins··on Diseconomies of scale in fraud, spam, support, and moderation
Tailored vs ready-made is a rounding error in the economies of scale involved in making clothes - both will use factory made fabric, dyed with factory dyes, woven from factory thread, made from factory farm cotton, grown from factory fertilizer, made from massive scale fracked natural gas, produced with millions of tonnes of factory steel, mined from factory mines..

The economies of scale that put food on your table and mine and clothe our kids - for all the horrors - will not be replaced by artisans doing small batch t-shirts from yard-grown cotton using hand made tools.

jakewins··on Diseconomies of scale in fraud, spam, support, and moderation
> In the sense of benefit to the overall economy, to the society in which that economy sits, or to humanity in general, virtually all economies of scale are actually harmful

You had me until here.

Economies of scale in commodities is the thing that makes it so we don’t have to do sustenance farming and everyone gets to have sewn clothes on :)

There’s truth in both sides there - we want nail production to not be done by hand, but also real humans with power to change things to answer the phone when we call support.

jakewins··on Basic proxy implementation using io_uring
What are the open security issues?
jakewins··on Ledger: Stripe's system for tracking and validating money movement
This processes three orders of magnitude more transactions per day than ethereum.

What would blockchain possibly be helpful with here?

jakewins··on Problem solving across 100,633 lines of code – Gemini 1.5 Pro Demo [video]
Na we’ll all be moved to perpetual on-call, every day an endless fire drill as hundreds of services are launched on top of the crumbling crash looping burning landscape of millions of services launched last quarter, a Mad Max world of endless adrenaline and New Relic AI-enhanced alerts.
jakewins··on Wi-Fi jamming to knock out cameras suspected in nine Minnesota burglaries
My understanding is this is rapidly being replaced by NFC phone tapping, at least that's what Hyundai is swapping to.

Annoying though, I've really enjoyed not having to dig through pockets to open the car :(

jakewins··on Wi-Fi jamming to knock out cameras suspected in nine Minnesota burglaries
What model vehicle stalls if it can't find the key? That seems like a severe safety problem - you mean it would brake if the battery dies on the key dongle?

At least in the two Hyundai's I've owned with this type of system - an Ioniq hybrid and a Kona EV - it drives just fine once started even if you chuck the key out the window.

← PreviousPage 6 of 19Next →