HNHacker News
TopNewBestAskShowJobs

mikepurvis

13,612 karma · joined October 27, 2009

Mechatronics engineering grad at University of Waterloo. I've led teams and shipped products— hardware and software. Areas of interest include DevOps and DX, tools, pipelines, and packaging. Experience in Nix and Debian primarily, increasingly GHA, TypeScript, and Docker Buildx Bake. Clearpath Robotics / OTTO Motors alum, now at Simbe Robotics.

All opinions expressed are entirely my own.

https://www.linkedin.com/in/mike-purvis-6166a42/

https://meet.hn/city/ca-Kitchener

submissionscomments
mikepurvis··on The Waymo effect: how AI is quietly making research less collaborative
True, though I wonder if 2026's "Claude make me a minecraft plugin" is 2006's "View Source" button.

Web development took a long time to gain respect in part due to the accessibility of it. When everyone had a nephew who "makes websites" and anyone could get started with a $5/mo Dreamhost account, it was harder to see why a consultant was worth $200/hr to do the job properly.

It'll be interesting to see how this plays out, particularly accounting for bots continuing to get smarter and smarter. Is the "craft" of software engineering dying before our eyes? Or will this be like webdev, where there's an explosion of access and ideas for a time and then it settles back down into being a thing that normal people mostly leave to the nerds.

mikepurvis··on The Waymo effect: how AI is quietly making research less collaborative
I remember fifteen years ago I would tell people leaving my house "oh yeah, ignore your GPS that tells you to go that way, just down the road here and left turn; that'll get you on the highway much faster."

Nowadays obviously Google Maps gets it right, but it was interesting that period where you had to actively preempt everyone's instinct to just rely on the little table hanging off their dashboard.

mikepurvis··on The Waymo effect: how AI is quietly making research less collaborative
I can see how the situation is annoying there; I've had a similar frustration when doing manual code review for someone's obviously AI-slop PR only to have their bot auto-respond to my comments. There's an asymmetry of effort that feels unfair. Like, don't ask for manual review of something that contains lots of things that your own manual review should have caught first, and if I provide human feedback, I expect human responses.

That said, I wouldn't necessarily take it at face value that the person you were working with trusted the bot more than you— at the end of the day, the bot was doing all the implementation anyway, so all the interactions between her and the bot were probably framed in language like "what if we" or "I think it would be better if", so her saying "my programmer colleague thinks X" would naturally generate that "yeah your colleague is right, let's go that direction, makes sense" that she passed back to you, probably thinking it was a complement.

Maybe the real frustration is her not understanding the value-add that still exists there when someone with real expertise and more context can step in and provide that nudge in the right direction.

mikepurvis··on What do Visa and Mastercard do? An intro to card networks
Transporting and handling cash is absolutely not free, and unlike joe sixpack taking out $20k in a stack of benjamins that one time to buy a motorcycle with cash, it's much worse when

a. a lot of it is coins, so it's heavy and bulky, and

b. it becomes a regular thing that everyone on the street observes you doing over and over

mikepurvis··on Stockfish 19
The Square One Chess channel on youtube turns this kind of exploration/commentary into quite digestible content, though the thumbnails are, like everything else these days, a little clickbaity.

It's been fun viewing for me and my chess-interested 9-year-old.

mikepurvis··on iPhone Duo
Agree. The comparison to a passport is ludicrous since a passport can be curled around your leg when it's in a pocket where this obviously cannot.
mikepurvis··on iPhone Duo
I switched off my 13 Mini this year to go to a Pixel 10 for the better Pebble support. I still miss it, it really was the last perfectly-sized phone.
mikepurvis··on What do Visa and Mastercard do? An intro to card networks
I mean, cash is also expensive in its own way. It has to be handled and insured, coins may need to be rolled before being taken to the bank, a register is a mechanical device to power and maintain, petty theft by employees is a concern.

It's not really that hard to see why a business charging $5-20 per transaction from people who all have phones and credit cards anyway might choose not to accept cash.

mikepurvis··on Lotus Notes and the dangers of starting from scratch
> For most users who only use a single platform

Also depends if your "platform" extends beyond just the OS. If platform includes IDE/editor/terminal, then a lot of users are juggling multiple keyboard schemes from different eras (ctrl-C has entered the chat).

mikepurvis··on Nitter and XCancel resume service after legal advice
For a while after the acquisition you could syndicate your external blog's RSS feed into your Facebook profile. Didn't take long for that to get the axe.
mikepurvis··on The largest electric aircraft just flew [video]
Seems like an extremely practical way to handle that requirement without hauling around a ton of extra battery capacity that will rarely get used.
mikepurvis··on “It works better in the app”
It's definitely a case where I see the value of the Asia-style super apps, where you just attach transactions like this to LINE or Grab or whatever, and everyone knows what to do with it.

The only thing remotely comparable in Toronto might be if Presto (payment app for multiple transit agencies in GTHA) took on parking payment, but the Presto network now supports just tapping a credit card so I don't know how many people would be getting cards or using the app for that either.

mikepurvis··on GUIs should be fully keyboard-driven
Using tab to cycle between interactive elements in a window or page. As soon as you tab and it goes somewhere unexpected, you're cooked.
mikepurvis··on “It works better in the app”
That seems reasonable. The kiosk is still there to pay with cards or cash directly, but the app/website lets you skip that, avoid printing a slip, maybe buy a pass or something.

The risk is mainly that down the road someone looks at the usage numbers and goes hmm, well the kiosk is only getting used 5% of the time, let's just discontinue supporting it.

mikepurvis··on “It works better in the app”
Valid, though I expect this being down the block from the festival theatre, it probably was being actively patrolled. I was a bit shaken by a prominent sign threatening a $400 fine for illegal parking, though having just checked the website for the City of Stratford, it looks like the standard municipal parking ticket is $40 if paid promptly, so the $400 fine was probably specific to parking in the theatre's dropoff zone or fire access or something.

All that say, I think you're probably right... next time instead of using my phone to install and sign up for a parking app, I'll just use it to check what the fine is and wipe hands on pants.

mikepurvis··on “It works better in the app”
Yes, just a plain tap-to-pay on the meter itself (or a nearby kiosk) is the obvious solution. I think doing it on the meter is awkward because then they all need to be online, so you're either dealing with individual M2M cards or some kind of local mesh network, not to mention powering them all. The kiosk is another thing to have to install and maintain and power. And doing a plain tap would have no way of inputting how long you wanted to pay for.

All that to say, I can fully understand how we ended up with "here's a sticker with a number on it, pay on your phone"; it's just, the pay on your phone option should be a one step apple/google pay webpage where you at most have to enter your license plate, and it certainly should not be collecting credit card details within the app.

mikepurvis··on “It works better in the app”
I guess, but no one carries coins, and the machine doesn't accept them. Me not having coins changed nothing about this situation.
mikepurvis··on “It works better in the app”
Ehh maybe. I mean, it was a three hour show and I don't carry coins any more. So that would have been ATM -> go to a store and break a twenty -> put in a few bucks -> remember to go out at intermission and put in more money.

Paying by phone is a step up from that old world; I just want it to be one tap, not a whole sequence that tries to upsell me on a monthly pass or establishing an in-app "parking wallet" or whatever.

mikepurvis··on “It works better in the app”
There are a lot of businesses that have a semi-captive hold on a market. Like a recent example was I drove my daughter and I to a stage play, and to pay for street parking I had to install an app, make an account, and add a credit card. A thing that should have been scan the QR code and click one button to mobile pay became a five minute ordeal.

In theory we could have gotten back in the car and looked for parking elsewhere but the curtain was in 30 minutes and who's to say they didn't have a contract with the city that covered all the parking?

Overall a very hostile experience and I was annoyed about how little I felt I had a choice in it.

mikepurvis··on Small Models Have Arrived
But isn't that really just about giving "front end" models more access to specialized tool libraries, which include models tuned to specific tasks? Like the first model says ah, we're being asked to code something, oh and we've been provided with some example code, let me invoke a tool call to my model the recognizes many languages, that model says that we're looking at ocaml. Okay, I better pass this off to my ocaml model which will decipher the supplied code and make a plan for what we do about the user's intent. The ocaml model recognizes that there are tests in the supplied code, let's have the special testing model have a look at the testing strategy and see how that fits in with what we just implemented, etc etc.

And perhaps at the end it all gets a single pass by a god-tier model for overall sanity and congruence, but the actual work, planning, coordination, and even user interaction was done by cheaper and faster agents of much more limited capability.

mikepurvis··on Tailcat – Like netcat, but over Tailscale’s data plane
I suspect some of this goes the other way too, that the people who see the value in Nix and/or are willing to tolerate its idiosyncrasies in exchange for what is on offer there are are a lot of the same kinds of folks likely to take an interest in rust, ocaml, elixir/gleam, etc.
mikepurvis··on Tailcat – Like netcat, but over Tailscale’s data plane
Yeah, Nix flakes can be a static URL path, but they inherit your VCS refs directly, so

    nix run github:tailscale/tailcat
Gets you the default branch. Tailcat doesn't yet have any tags, but if you want a build of the main tailscale repo at a particular tag, it's just:

    nix build github:tailscale/tailscale/v1.102.3
But because this can be any ref, you can specify an arbitrary branch or tag, so like if you're trying to get a colleague to check out your work it's as easy as:

    nix run github:tailscale/tailcat/my-fancy-branch
And obviously a flake can reference other flakes with the same flexibility, so your home manager config or your company's private app flakes can transparently pull the not-yet-merged version of a change you're proposing upstream without having to make special internal "patched" releases with tortured 1.2.3~actually.1.4.6 type version numbers like in Debian. And of course that same flake reference policy applies all the way down, so you also don't need to worry about the private apt repo as a big piece of shared state that's going rev forward and break your older code branches or Dockerfiles, since each branch of your app flake has its own flake.nix and accompanying lockfile controlling which deps it is pinned to.

It really is an awfully nice system, and shockingly modern in its sensibilities considering how old the Nix fundamentals are.

mikepurvis··on Tailcat – Like netcat, but over Tailscale’s data plane
Maintaining and providing a deb is pretty easy too; the real killer for a nix flake I think it's how dirt simple it is to provide infinite testing builds for different branches.
mikepurvis··on Tailcat – Like netcat, but over Tailscale’s data plane
I enjoy that they supply a nix install/environment, similar to the main tailscale/tailscale repo. Is nix widespread or the standard dev environment at tailscale, or is it like a 10% option and most people just use Docker or whatever?
mikepurvis··on New Mac Studio with M5 Max and M5 Ultra
I switched back to a desktop tower a few years ago as well. Notably, I don't use external displays with the work-issued laptop, but rather set up the laptop on a stand alongside the dual 1440p displays and have synergy for mouse and keyboard sharing.
mikepurvis··on New Mac Studio with M5 Max and M5 Ultra
Unrelated to disk, I've definitely been thinking about this more on the network side recently. I upgraded to 1.4/900 home fibre a while back and then got a modern Unifi UCG/U6 wifi setup, and while I do get 1.4GB down from my hardwired workstation in an explicit speed test I have yet to get anything close to that on real-world workflows. Torrents, container-pulls, backups, all of it seem to cap out around 20-50 MB/s.

So while have having this massive almost-symmetric fibre pipe is cool on paper, I haven't felt a huge need to install 2.5G gear all over the house.

mikepurvis··on New Mac Studio with M5 Max and M5 Ultra
I bet this is a lot of what is driving it on the DC side, and just in general bringing disk closer to RAM as far more conventional database type workloads.
mikepurvis··on I ran Photoshop on a £0.60 computer chip
uCLinux was a real thing for a good long while, though I think at this point real microprocessors are cheap, simple, and widely available enough that there hasn't been much need.
mikepurvis··on Why aren't smart people happier? (2022)
Parenting philosophy rooted in growth mindset advocates less for "I am smart and therefore I can do this" but rather "the hard work that it will take to do this is worth investing, and failures along the way are opportunities to learn rather than judgments of whether I'm good enough or should continue."
mikepurvis··on Pressed Penny Machine Map
Fridge magnets are this for me, but having a way to display a whole collection of smashed pennies would be pretty great too.
← PreviousPage 3 of 34Next →