HNHacker News
TopNewBestAskShowJobs

dpeckett

390 karma · joined April 6, 2015

Contact email: damian@pecke.tt

Recent Projects: https://github.com/apoxy-dev/icx

submissionscomments
dpeckett··on Fly.io Postgres cluster down for 3 days, no word from them about it
Couldn't agree more, I think Hetzner is probably Europe's best bet on a hyperscaler. One of the more telling indicators IMO is their growing market share outside of the EU/DACH.

To add on to the comments about Hetzner building their own custom hardware, they also custom built their own software stack. They rejected the hype that was OpenStack and worked diligently on their own hypervisor platform (that they are incredibly secretive about) and that appears to be paying off in spades for them. Most sovereign cloud plays end up being suffocated by the complexity, and incoherence, of the OpenStack ecosystem. It just becomes impossible to ship.

For a fascinatingly different take on how to build a datacenter: https://www.youtube.com/watch?v=5eo8nz_niiM

* Edit: remove speculation about Kubernetes and Hetzner, that was based on hazy memory.

dpeckett··on Honda Monkey Breaks World Record Covering 4,183 Km on Single Tank
Was reading about "The Quiet Achiever" the other day, the first transcontinental crossing in a solar powered vehicle (interestingly also 4000km). Kinda crazy looking back 40 years how far we've come!

https://www.youtube.com/watch?v=jMmTRI6giSU

dpeckett··on Many in the AI field think the bigger-is-better approach is running out of road
Sparse networks are the future, there's definitely a few major algorithmic hurdles we'll have to cross before they end up a real option but long term they will dominate (after all they already do in the living world).

All our current approaches rely on dense matrix multiplications. These approaches necessitate a tremendous amount of communication bandwidth (and low latency collectives). This is extremely challenging and expensive to scale O(n^2.3).

The constraints of physics and finance make significantly larger models out of reach for now.

dpeckett··on Opinion: Geoengineering Is Shockingly Inexpensive
Oh that's well underway, most people just haven't noticed yet.

https://www.youtube.com/watch?v=Dp2GHanfkOc

https://www.youtube.com/watch?v=ITh4GUhA_B8

https://www.youtube.com/watch?v=mEXzzlDCkoo

dpeckett··on 432-Core RISC-V European Processor Designed for Use in Space Taped Out
I hope someone picks up the architecture and brings it to the wider market. Would be great to see some opensource hardware AI accelerator options.

Looking more closely at snitch, unfortunately it looks like the two largest contributors have already left the project. Perhaps they can keep momentum but industry poaching remains an issue for ESA.

dpeckett··on Ask HN: What projects are you working on now?
I've got a great use case for horust, or at least a container init process capable of running multiple child processes (kubernetes nodes inside containers).

Consider me interested. Contact me via the email listed in my profile.

dpeckett··on Building an automatic chicken coop door opener
Vision sounds like a really tough way to solve it, challenging lighting conditions just to make things trickier.

I know that in some contexts, RFID tags are used for livestock tracking. Don't know if they would comfortably work on birds though.

dpeckett··on Building an automatic chicken coop door opener
Fair enough, I got the impression the target audience was more IOT focused given the site it was hosted on.
dpeckett··on Building an automatic chicken coop door opener
I have no qualms with the merit of an automatic chicken coop door. Obviously the benefits can be considerable, and the project certainly fits the bill.

The point is, almost all maker projects jump straight to let's throw an Arduino at it, with no consideration to the tradeoffs at all. This can be quite a disservice.

There are tradeoffs, eg. power management? (those batteries aren't going to last at all), fault tolerance? (arduino crashes due to condensation shorting out the timing crystal), maintenance? can you instruct an untrained individual to perform any meaningful repair work (eg. bend the arm on a damaged limit switch).

One benefit IOT can bring, is the detection of fault conditions etc (door took too long to open fully something stuck in the tracks), but implementing that kind of handling in a meaningful way is a complex problem.

These things rarely matter for a diy project, but you see this stuff everywhere. Consider whether a problem can be solved in an analog or mechanical manner, it's likely to lead to a simpler and more robust solution.

As for ensuring chickens are inside of a coop, that's a tough problem and I'm interested to hear your thoughts on that?

Eg. for the above chicken in coup problem, something as simple as a weight triggered switch on the ramp (or a push bar near the coop opening) is likely to out perform anything more complex.

dpeckett··on Building an automatic chicken coop door opener
Great project and all, and definitely something fun to work on, but this is a pretty questionable application for IOT (don't wish to discourage).

What's the key metric here? The chicken coop opens and closes at the appropriate time of the day? How does IOT help solve that objective? There's some great parallels here to an MVP in the software world.

As for my two cents, an alternative and more robust solution, could be put together with something like an aliexpress ambient light switch, limit switches, a recycled car battery, a trickle charge solar panel, and some relay logic.

dpeckett··on New Zealand’s War on Rats (2017)
I wonder if the drug would work on felines, combined with the technology/approach here https://www.youtube.com/watch?v=AnpJBI_502U you may have an ethical way of dealing with what's a huge problem in Australia. I've been keeping an eye on this technology for a while now.
dpeckett··on A Feasible Approach to Human Stasis for Long-Duration Deep Space Missions
Hypothetical question, if gene editing in humans became an accepted thing. I wonder how preserved our metabolic pathways and genes for long term hibernation are.
dpeckett··on An Augmented Reality Microscope for Cancer Detection
A lot of that cost comes from all the supporting hardware and certified software. Plus a lot of the machines have to be designed for high throughput applications.
dpeckett··on Microsoft built its own custom Linux kernel for its new IoT service
The Cortex M series is prolific but lacks a proper memory management unit which makes running Linux an iffy business. FreeRTOS is by far the most dominant player in this space.

Different markets really, sounds like Microsoft is targeting items like internet connected cameras (think Mirai endpoints). Where as Amazon is targeting low power prolific devices, eg. IoT buttons.

dpeckett··on Testing FIDO U2F security keys
Yeh it's a coprocesser arrangement for sure these days. Probably the only way to do it remotely safely, sure doesn't provide a ton of flexibility though and definitely adds to the total BOM cost for the devices.
dpeckett··on Testing FIDO U2F security keys
Broadly based on abusing USB ethernet and browser origin policies. Relies on the fact browsers are used to treating the network layer as a hostile environment.
dpeckett··on Testing FIDO U2F security keys
A while back I looked into building my own hardware two factor key. I wanted to experiment with the technology. I've since backed off the idea due to several serious challenges with designing secure hardware for this purpose.

* By necessity there's a pretty large attack surface, you've likely got a vendor specific bluetooth stack and coprocessor. The same goes for wired solutions with hardware USB peripherals (though atleast significantly less complex). Being sure the hardware peripheral doesn't have any major memory safety issues is a complete bear to test, and that's not even touching on sidechannel analysis.

* Running public key crypto on microcontroller (excluding specific hardware crypto support) is a little more novel than a desktop machine, the leading libraries are no where near as extensively battle tested. And there's architectural concerns about sidechannel attacks and often the lack of any memory protection units (bigger chips solve these issues, but power budget concerns are tricky).

* And given all software has bugs, how do you update the firmware? Signed firmware patches as part of driver update? Who's to say somebody hasn't already owned your bootloader.

I think the solution to all this, is to abandon the idea of additional hardware, what if your smartphone could act as a two factor hardware token over Bluetooth? The big problem is pairing really, and how to make this an easy process for users, I never figured that out. On the other hand I did come up with a scheme that would allow usb based hardware tokens to work without additional drivers / software and on all current major operating systems and browsers. I really ought to work on it, but as above I don't trust embedded hardware (embedded hardware is such an antithesis to, move fast and break things).

dpeckett··on Cinnamon II an Apple ][ Compatible Wrist Watch
I guess we'll have to see what the internet makes of it all :D. Short production runs are inherently expensive. And I think all too often kickstarter projects fail due to not properly pricing in externalities.

As for contributing/collaborating, the entire source code and interpreter for the watch will be made open source. At the end of the kickstarter. Plenty of opportunities ahead!

dpeckett··on Turning the Arduino Uno into an Apple II
Author here, this is one post out of a series http://dpeckett.com, if this stuff interests you, feel free to follow my twitter: https://twitter.com/damianpeckett.
dpeckett··on Turning the Arduino Uno into an Apple II
Awesome work, your avr computer is really impressive! So similar and yet so different! Emulating the Apple II was relatively easy, not a lot of extra hardware.

Haven't tried cross compiling anything yet, sounds really cool. Nice work on the Mandelbrot demo!

dpeckett··on Turning the Arduino Uno into an Apple II
Author here. I meant that in earnest, getting the emulator 6502 compatible was a large part of the project. But it did come together in about ten days, on the order of 70 hours.

There's something about these old systems, you can fit the whole system into your mental working memory. Just can't do that with todays abstractions.

← PreviousPage 4 of 4