HNHacker News
TopNewBestAskShowJobs

Olreich

711 karma · joined July 9, 2012

submissionscomments
Olreich··on To make programming easier, first be conscious of what makes it hard (2014)
Wix, Squarespace, etc.
Olreich··on Reconsidering the way I explain programming (2020)
I find the easiest way to get past the abstraction of recursion is to describe what a function actually is, rather than the handwavey definitions we usually steal from math. A function is a stack and a list of instructions. When you call a function, you add things to the stack, and when you return from the function, you throw those things away.

Stack of papers, flight of stairs, layers of onions, circle of life, etc. they all just harken back to what it actually is without having to define it directly.

Olreich··on Handwriting is better than typing when learning a new language, study finds
I wonder if this has to do with how fast typing is. Is there similarly diminished learning in very slow typers? Does a shorthand expert fail to learn as much?
Olreich··on GitHub Copilot regurgitates valid secrets
I think if your code requires reconstruction of nonlinear pieces to read it, you’ve written bad code. Fundamentally, a program is a list of instructions for a computer to run. The more linear that list is, the more efficiently the computer can run it. Linear code is also much simpler to read and understand for us humans as you pointed out. Iterate on the code until you find the linear path through it, otherwise you’re going to be in a world of pain if you need to understand it in the future.
Olreich··on GitHub Copilot regurgitates valid secrets
I’d be very surprised if management at the least didn’t have their heads in the sand about the potential failure mode. There are often “must deliver” dates at large companies because someone made a promise about a deadline and now heads will roll for missing it whether anyone actually cares or not. So long as middle management thinks the C suite is watching them, they are desperate to meet quota.

Hilariously, this results in stuff like Copilot getting released to great big legal problems. Only then does the C suite actually notice the project and get upset that it is a legal nightmare for them.

I think the real secret to winning in big tech is that your job is just to keep your head down and keep the money rolling in without causing headaches for higher ups. Increase sales, make customers happy enough to keep paying, maybe release a cool product. But more importantly, don’t cause a major outage, burn the PR team, or get caught up in a legal kerfuffle.

Olreich··on Highview Developing 50MW/500MWh Liquid Air Energy Storage Facility in Chile
People are working on molten salt batteries: https://ambri.com/technology/ is the only commercial one I know of though.
Olreich··on Write a time-series database engine from scratch
It seems they are relying on the partitions to act as sync frames for this implementation.
Olreich··on Secret Recordings Reveal How Exxon Lobbyists Manipulate Politicians and Public
Eh, I think it’s worth the old college try of “anything new please” without further corroboration. Everything we’ve tried at large scales has sucked for the majority of the population.
Olreich··on We Infiltrated a Counterfeit Check Ring Now What?
Idk, I pay my water bill using a check. I think it’s just restaurants that use “check” to mean bill.
Olreich··on Husband of Amazon employee sentenced to prison for insider trading Amazon stock
It’s not very effective…

https://en.wikipedia.org/wiki/2020_congressional_insider_tra...

Olreich··on Ask HN: Post Burnout Ideas
If you’re talking about the file system share, there are better tools for people with the skills to get it set up. The tech seems good, but the set up is hard. The limitation of local network is rough for getting people excited, you should set up a tiny relay to let people connect.

The documentation is a little unfocused, but the pitch is good. If we assume the pitch is the primary purpose, simplifying the user experience should be the top priority. There’s a bunch of tools that have tried to solve the “throw file at Mom” problem, but I haven’t seen one work without putting my files in someone else’s server.

I’m not sure how you differentiate this from Plex, WD myCloud, OneDrive, or Google Drive. That’s an important factor considering how easier some are and how powerful others are.

You built a desktop UI in the browser to share files? That’s a bit much for anyone to take a serious look at. It’s the same problem as documentation and this post: needs more focus.

If you’re enjoying making it, and it does useful things for you, have fun! But I don’t see it getting much feedback or traction due mostly to lack of focus, but solving some of the technical limitations would make it much more broadly applicable and would help as well.

Olreich··on Testing in the Twenties
Testing is important, but unit tests are the least important of them.

Powerful type systems and compilers that catch your mistakes before you can run anything is the best kind of testing. It’s fully automated and will inform you about most things early on. Practically, this kills dynamic languages for projects expected to become larger than proof of concept (pydantic, typescript, etc. mitigate, but must be in and enforced from the start).

After that manual testing is the most important. Specifically because it can be done by hand with isolated inputs and outputs. This won’t catch the refactoring breaking something disparate, but will ensure you can reproduce production issues with the code locally where you can expect it.

After that, integration tests are the most important. Give the whole component/system an input, and verify the output. This should be as fast as possible but should mock as little as possible. This should be able to be done from a local machine with a debugger running and still be fast. If an integration test takes half an hour, it’s never going to get run except on a Thursday at 6pm when you’re trying to get out a change before the Sprint review. Then it will fail and make life hell, because running it is so slow that running with a debugger is going to be hell.

After that, logging is the most important form of testing. Every input or start of a workflow should get an ID that follows it through the system with logging to watch it go. Any problem identified should include that Id so that you can grab logs from all the disparate systems. Specifically, that ID should be attached to the inputs or each component so you can replay it in a controlled environment later.

Then you finally get to unit tests, because once your software works as expected, they let you lock down the implementation and ensure that any changes will get flagged by the tests. If you lock down the unit tests before the software works, then you wind up having to go back and redo the unit tests for every issue found, as the implementation wasn’t correct so the unit tests can’t test it. This is the big lie of unit testing. They don’t prove correctness, they prove implementation details.

After that you want to get to end-to-end that workflow the whole user experience with a product or system. But if you don’t get there, that’s really okay.

Most problems are caught by the compiler, manual tests, and integration tests during development. Any problems that do come up can be identified quickly and repaired by the highly inspectable infrastructure that provides you IDs to dig out logs for any input, and while you’re fixing the issue, your unit tests will let you know when the implementation has changed in a way that requires more thought (though the compiler probably did first and the unit test complaints are just noise).

Olreich··on Drunk Post: Things I've Learned as a Sr Engineer
Obscure tech and proprietary tools do translate… if you can translate them.

Programming languages are all the same, so learn 3 or 4 new ones and discover that you can probably write in any language for an interview (then do some in relatively unfamiliar languages for kicks and giggles to practice).

Tech is all the same. Take data in, poop data out. That’s the whole job. The formats and protocols change, but once you starting thinking about your stacks as data-in, data-out, they all start looking the same.

Make video games or hardware drivers from scratch, those are the hardest things to make. Video games from complexity overload, and hardware drivers from interface complexity.

Olreich··on How Lucky Is Too Lucky in Minecraft?
I can already do this for most popular speed games, and it’s not a big problem. The TAS tools will help the community much more than they will hurt integrity.
Olreich··on The mortifying ordeal of pairing all day
Pair programming only seems good when there is a large disparity in experience or seniority. The senior engineer is there to teach the junior engineer and is sacrificing a great deal of time and energy to do so. That trade off may be worth it. Pairing people with the same skill set? Mostly that’s just a waste.
Olreich··on The compiler will optimize that away
The point stands anyway. Why are all my dev tools super slow? If my time is important, shouldn’t those be fast and heavily optimized? And don’t try to tell me that 4 billion cycles a second across 16 cores operating on in-memory contiguous values isn’t enough to update a watch window more than 1 time per second.
Olreich··on How many engineers does it take to make subscripting work?
If you don’t know the column will live forever, avoid adding it. Adding columns is easy, migrating and deleting columns is super annoying, even with auto-migration and an ORM, because the risk of data loss is significant.
Olreich··on How many engineers does it take to make subscripting work?
As an industry, software development needs to get better at cleaning up TODOs. This is especially bad when business needs take prime spot always. Sometimes, technical health needs to take priority and drive down these issues. If there’s confidence in the organization that we can take on some tech debt and pay it down a little bit later, then you can move faster (just like leverage in finances). It’s when the debt you take on is so great or so well swept under the rug that it turns into a monster and ruins your day. If the tech debt items are allowed to live forever, the org gets into the position of quickly maxing out their credit and never being able to add more debt which slows delivery.

In the scrum world, prioritizing a feature story over a debt story should require a strong reason with significant debate. Don’t let them live longer than a couple of sprints.

Olreich··on Ask HN: Learning Distributed Systems as a Junior Engineer
Good advice for the system design level of abstraction in the other comments. But do try to remember, no matter what, everything is just a computer running some code and shoving some data around to various other computers or devices. We lose sight of that a lot and consider the systems we abstract as somehow different from a computer. Avoid it as much as you can.
Olreich··on Tetris-OS: An operating system that only plays Tetris
It's no more misleading than saying it runs on bare metal. If I flash the correct motherboard ROM chip with the necessary bits to play Tetris in the UEFI (or completely replace it), would I be running closer or farther from "bare metal" than this is? Am I still on bare metal if I choose to use GRUB for loading up the game? Do I become an OS if I use GRUB instead of embedding that logic within my binary?

We call things BIOS and Boot-loader and Operating System by virtue of history, not necessity. It could even be confusing to a newcomer looking into development closer to the hardware and being told that all these programs are fundamentally different, instead of just being instructions and data getting shoved through a bunch of tubes in the processor and PCB.

Olreich··on Ask HN: Does anyone else find the AWS Lambda developer experience frustrating?
While that’s generally true, the inputs and outputs and execution environment that you can count on is well-documented on AWS Lambda. That means that recreation of execution environment for the function in question is pretty reasonable locally (why is this not a thing then Amazon?!?). For the rest of AWS, you can either get credentials in place for access or mock those out somehow, but that’s not officially supported either.

I feel like the service is leaving a bunch of easy work on the table that would let people have a local development process that mirrors the remote development reality.

Olreich··on Why Rust strings seem hard
The problem isn’t the forcing you to handle it, it’s the horrific API to handling it. It feels like all the OS representations were haphazardly bolted onto the core string API which is already kinda janky.

Rust just needs to sit down and make a good API for the whole set of string-ish types they have.

Olreich··on Ask HN: I built it, nobody came, now what?
> As in you can't do basic maths AFTER the data's been processed.

That's not true at all. It's just a different query language to yours and a different methodology to utilizing it. All the big BI and SIEM tools have post-processing capabilities able to do statistics at the Excel level, typically by creating new columns in the dataset. PowerBI has an excel-like function system for calculated columns. Tableau has a cross between Basic and Excel functions for calculated columns. Splunk and ELK are the standouts for letting you directly modify columns in their query languages, but the same sort of functions exist. In all of these tools, you can extend the query language as well through a plugin system, so for situations where you need a full programming language, there is an escape hatch (though a hard one to use).

> further processing in Excel. This is the part I want to compete with.

If you want to compete with Excel, you're going to have to get on its level. Specifically, the JavaScript must go. The reason big companies use Excel is because they have a couple of people at the company in the reporting jobs that can handle some programming, but most of the data entry and management has to be done by people who wouldn't know where to begin if you sat them down in a programming environment (IT gets called when people accidentally end up in the visual basic environment in Excel).

The majority of users do not want to live in a programming environment, so by having that be the primary interface, you're cutting the already small potential user base by another order of magnitude with the JavaScript environment. The advantage you get for an Excel-like query language is familiarity to the potential user base, less threatening environment, and additional control over the interface with the user (programming languages are generally poor interfaces, and JavaScript is no exception).

Olreich··on Ask HN: I built it, nobody came, now what?
It gets worse than the comments I’m reading are saying. They’re implying that you could get there with marketing. But the thing is, you aren’t competing with Excel. You’re competing with PowerBI, Tableau, Splunk, ELK, and Jupyter, but without any of the enterprise or scaling selling points. So this is a cool demo of a way to define charts once you have the data mostly there, but I don’t think it will scale to very large data volumes. With no way to hook into some already-scaled query engine, you won’t have much use by people that need to analyze production data.

If this was very performant compared to Excel, then that would be a use case that could get you a niche while trying to build out the scale and interoperability required for successful BI tools.

Olreich··on They Write the Right Stuff (2021)
This is still sadly applicable to software dev as well, not just ML and statistics.
Olreich··on New 5G protocol vulnerabilities allow location tracking
AT&T is the public-facing company with a ton of cash. They want to advertise the next speed bump in cellular. They hire a contractor to build it nationwide. That contractor hires 4 contractors for regions of the US and pockets some profit. Each of those contractors hires another 20 contractors to handle the rollout and pocket some profit. The 20 contractors talk to a variety of hardware integrators about deploying the new technology and start getting permits.

The hardware integrators have a 90% chance of being from the world of industrial systems where closed networks and point to point runs means they get to pretend security isn’t their problem. They apply the same principles to their hardware installs.

AT&T wants to be able to administrate and manage their new hardware without keeping 1000 hardware integrators on payroll, so they have their contractors’ contractors develop a manual that can be used by AT&T Network Services, but this is out of the initial CAPEX quota, so it’s done on the cheap and no one does any real work. They just grab all the default passwords and architecture estimates from the integrators and put it into a set of hundred page books for different parts of the network.

The government, customers, and shareholders choose to not hold AT&T accountable for any failures in their network. They instead accept it as “the way things must be”, and AT&T is happy to buy into that story and do the exact same thing on every capital project and maintenance.

This is how you get telcos failing at security every time for decades. Their incentives as a customer-facing company running the worlds largest infrastructure projects encourage them to farm out all the infrastructure work to several layers deep of contractors with no real responsibility enforced by anyone.

Olreich··on Reinstating Net Neutrality in the US
You don’t criminalize behavior because it will happen, but because it could. Net neutrality prevents companies from doing harmful things, like installing data caps and providing preferential treatment to certain data. Those things don’t have to happen, and maybe aren’t happening because the companies think they’d get bad PR. But they could happen, and that’s why you regulate to prevent them.

Perhaps the big kerfuffle was overblown, or perhaps it instead slowed down the plans that companies had (since it was now a provable PR nightmare post-kerfuffle). We don’t have enough data to decide that. We do know that companies are currently allowed to act in anti-consumer fashion.

Olreich··on The GameStop Mess Exposes the Naked Short Selling Scam – The American Prospect
The market doesn’t reflect reality unless everyone in the market agrees to reality. The hedge funds choose their reality and try to convince everyone of it. They’ve been doing that for decades. The only change here is that a bunch of people on social media started doing it too, and it’s working.
Olreich··on When Everyone’s a Genius: A Few Thoughts on Speculation
Desperate times call for desperate measures. Most of that growing group of people see their lives getting materially worse over time. It makes sense that they’d be desperate and incurious enough to hitch their wagon to a strongman figure who says the things they desperately want to hear. The governments of the western world have done a pretty bad job on the whole with making the lives of the people better. Reversing that trend would do a lot to help solve the problem (few people want to overthrow governments when they are safe and content).
Olreich··on A Simple Way to Reduce Cognitive Bias
Rational thinking doesn’t preclude data gathering. There’s lots of data points to inform that position: who will I be playing against? What is my relative skill compared to them? Is skill a major factor in my poker play? Can I expect to lose money? If I expect to lose money, would I enjoy that anyway because the game itself is fun? Would I have a particular threshold over which my losses would ruin my fun? Can I come up with a sure fire way to prevent my losses breaching the fun threshold?

After data gathering, it’s no longer garbage in and garbage out. You’ll note the questions assumed a randomness to the results, captured the worst case scenario, and minimized it to something I can live with.

You can do this with everything. That’s the beauty of rational thinking. It’s less about knowing the absolute correct solution and more about being conscious of the process of decision, rather than letting intuitive cues be the only deciding factors.

← PreviousPage 6 of 9Next →