I'm open to having a monthly thread about this but it shouldn't be on the first of the month - that's a traffic jam of monthliness because there are 3 (already too many!) Who Is Hiring threads on those days. As a result, these threads aren't really achieving liftoff: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
If that were as well as they could do, we'd have to call it off, but there should be more potential than that so I'd be willing to put some support into it, at least for one or two tries.
Here's one option: why don't you try the 3rd Saturday of each month? A weekend day should bias things in the direction of side projects and curiosities, which is probably more interesting (in HN's sense). If you do that, email [email protected] and we'll manually place it on the front page, see how it does, and maybe take it from there.
I truly believe that the increasingly growing awareness around "supply chain" security presents a fantastic opportunity for FOSS maintainers to gain financial independence and sustain their projects by selling security assurance to the consumers on software contents, packaging, etc. I'm building this marketplace. Would love to get feedback/recommendations from my fellow FOSS devs here.
I'm still working on the latest prototype for my hurricane engine. I finally made the fixture to hold the compressor, ducting, and turbine together. It's kind of a kludge but I ran the compressor up to full power and nothing blew up, so that's good. I'm hoping to be done with rewiring the turbine generator by the end of the weekend, so I can run some recondensation tests later this month.
It's basically an engine that runs on the same process that powers hurricanes, where hot air and water vapor carry entropy between the warmth of the earth's surface and the cold of the tropopause, making a heat engine. But in a much smaller package, using compression to act like an additional layer in the atmosphere. At least that's the idea. I'll see how it actually works.
Also do you happen to know of any good visualizations of how a hurricane works? I can’t figure out why the eye is calm. Shouldn’t that be the rapidly rising hot air?
Sorry, I don't know of any good hurricane visualizations, but I'm sure there must be some online. The air in the eye is sinking colder air, which is clear because sinking colder air does not create any clouds or precipitation. It is basically where the air goes that gets displaced by the rising warm air that makes the clouds and rain.
Not sure what else to say about my project. If it goes well, I'm sure I'll have a lot more to say. If you have any questions, I'll see if I can answer them.
No, not at all. It's more like a wind turbine that runs in a constant wind generated by the evaporation/recondensation process. More water is evaporated than condensed so that the entropy balance is always positive. It sucks in heat locally and converts some to work and sends the rest to the upper atmosphere. You can think of it as a form of solar energy that doesn't need the sun to be shining at the time or place that the engine is running, as long as the sun keeps the troposphere warm.
It can incorporate a heat pump (or an equivalent phase change cycle in some alternate designs) to increase specific power, but it is definitely a heat engine, not a heat pump. Running it backwards to make it like a heat pump would be wicked hard, and besides, no one wants to try to pump heat out of the upper atmosphere.
I’m working on a math problem generator. While there are many similar websites out there, I wanted to create one for myself that caters to my needs and doesn’t come with the usual overhead such as ads, settings, and registration.
I'm building a Starlark-based build system that operates at the level of building multi-container environments, rather than normal code assets: https://github.com/kurtosis-tech/kurtosis
It's kind of like Docker in the sense that it packages up an application in a reproducible, composable, and portable way...but instead of working at the level of one server, it works at the level of multi-server applications.
In the back of my head I'm very curious about making an application that just messages me on WhatsApp whenever its the birthday of one of my close friends because I never remember those
A spaced repetition vocabulary app for German, which allows parsing words from text (correctly handling conjugations and trennbarre verben) and generates different sentence examples using GPT.
On the less technical front, rebranding and repackaging my SME coaching (10-100 employees) business. I explain that AI means I bring Actual Intelligence, but increasingly I’m getting hands on implementing some actual AI tools to free up some time or cash for bigger strategic projects.
I’m also slowly making baby steps towards replacing myself with an AI coach I’ve helped build. If I don’t, someone else will, and I think I’m well-placed to give it a crack!
Working on getting two virtual servers running (Windows Server 2022 and Ubuntu Server 22.04), so I can learn more about Server and System Administration. I've bugged the SysAdmins at my workplace to death with my questions, but now I get to dig deep and setup AD on Windows and Security/Hardening on Ubuntu. (Using VirtualBox to easily get snapshots, so I can rollback if I bork it up.)
I've been thinking about some code annotation tool recently, where you can quickly plaster a code base with (linked) notes. This could be used to outline a new feature for a junior to implement, but also when trying to understand a new code base.
There's a few tools similar, but not exactly how I imagine it to be.
Has anyone found a tool for the use-cases mentioned above?
I'm working on a online course that involves gamification with elements from movies and video games to make topics easy to understand and take people from "passive learning" to "active learning".
added some logic to visual-code-annotator to grab code scope text to this type of visual comments to make chap questions and approve for answers from open ai
Being somewhat older than most in the computing world, I remember well the days of 8-bit computers and in particular the Atari XL/XE range, but it's getting harder to link up these ancient plodding beasts to modern hardware, which really wants an HDMI link over an RF modulator, for example... So...
- Design an interface board[1] to the bus that pokes out the back of the computer, this is a simple level-translator and 36-way cable to "the box" (see below).
- The cable is a Mini-SAS type[2], because they're much more appealing to look at than ribbon cable snaking everywhere. This board is intentionally simple so that by swapping out the simple board, the rest of the hardware can interoperate between different machines.
- Design "the box". This has an FPGA on it (Efinix by choice since they're low-cost) which has a bitstream that understands the bus protocol of whichever machine it's plugged into.
- The FPGA has local memory (PSRAM) which it can interface to the 8-bit - indeed, the 8-bit will have its RAM requests be serviced from the PSRAM on the FPGA entirely under most circumstances.
- The FPGA also links to an RP2040 which acts as a sort of system-controller (really I'm using this for the boot loader, to make it easy to upgrade in the field). However, there is also a "slots" interface which will define part and SPI interfaces that can have data transferred bi-directionally to the Pi, and also "dma" directly into the PSRAM that hosts the 8-bit memory.
- And finally the FPGA links to a raspberry Pi - still not sure which one (it depends more on graphics capability) over the SMI (Secondary Memory Interface) bus - which takes pretty much all the GPIO but gives me about 500-600 mbits/sec data throughput from FPGA to Pi, with DMA to userland code happening in about 50uS.
- The FPGA reproduces the bus traffic on the 8-bit into a marked-up set of data streams, and batches them over to the Pi over the SMI. The FPGA will specifically interpret things like video into something easy for the Pi to understand without billions of bit-shifts.
- On the Pi, there's a (currently QT, possibly SDL2/GPU, maybe a.n.other library) application which scales up the incoming video to an HDMI signal, and vends it out over the HDMI port.
- Given this is an Atari, and to provide a few more features, I plan to implement a GEM interface (akin to the ST, but maybe better looking :) to allow a desktop environment for files that can be "downloaded" into the 8-bit's (PSRAM) memory from storage on the Pi.
- If all this actually works, at some point it'd be nice to integrate the Cyclone[3] 68k emulator with that GEM VDI/AES and actually run Atari ST programs on a 1920x1080 HD truecolour display.
- I was planning on implementing the GEM service akin to an X-server, where there's a fullscreen window, and clients (including the desktop) connect to it over a pipe. That way multiple apps can run, and it should even be possible to have a design/development interface with the XL/XE running in a window, and code being compiled/downloaded on the same HDMI display. I can dream :)
Anyway, what makes this a lot more feasible is the low-latency high-bandwidth SMI interface on the Pi, that really ought to have had more attention on it, given it's on every Pi shipped, and it's perfect for an FPGA interface.
If that were as well as they could do, we'd have to call it off, but there should be more potential than that so I'd be willing to put some support into it, at least for one or two tries.
Here's one option: why don't you try the 3rd Saturday of each month? A weekend day should bias things in the direction of side projects and curiosities, which is probably more interesting (in HN's sense). If you do that, email [email protected] and we'll manually place it on the front page, see how it does, and maybe take it from there.