dgriffith

joined 3 years ago
[–] dgriffith@aussie.zone 2 points 33 minutes ago* (last edited 32 minutes ago) (1 children)

It's an accessibility issue, mainly. And very faintly a safety issue, and also owners of the cables risk getting them damaged by something with steel/solid wheels (like an industrial bin or delivery trolley) running over their cables.

Other councils have suggested cutting a groove across the footpath to lay a cable in, which seems like a fairly acceptable solution. 20 minutes work with a concrete saw, tap a channel with a flip top lid in, done. Or even just the groove - a 3/4" groove is enough for a cable and fine for wheelchairs and etc to pass over.

[–] dgriffith@aussie.zone 35 points 6 days ago (34 children)

What are you expecting to break down? A modern internal combustion engine plus continuously variable transmission is one the world's most complex pieces of engineering in common use today.

An electric drive system is trivial in comparison.

[–] dgriffith@aussie.zone 25 points 6 days ago (6 children)

WinXP ran quite nicely on 4GB of RAM. You could boot to the desktop in under 5 seconds if you had a solid state drive.

All the day to day tasks someone needs to do with their computers haven't materially changed in the last twenty years since XP held the crown. In that time the speed of processors, storage, gpus, and ram has increased by literal orders of magnitude.

And yet, somehow, things have become much shittier since then.

[–] dgriffith@aussie.zone 2 points 1 week ago (4 children)

DKIM/SPF/DMARC add extra legitimacy to your message, but you can certainly still end up marked as spam if Google thinks you're sending it.

One line emails, emails that just have links, emails with generic subjects, emails with just a subject and no text body, emails with no subjects - all that massively shoves Google's spam detection towards spam. But that generally holds true with all providers as all the big players are using some form of Bayesian filtering (and so am I, with SpamAssassin on my incoming). If you send spammy-looking stuff, you're gonna get marked as spam.

[–] dgriffith@aussie.zone 14 points 1 week ago (6 children)

You spend an hour or two setting up DKIM , SPF, and DMARC, so that Gmail/Hotmail/etc can validate that you are who you say you are.

[–] dgriffith@aussie.zone 31 points 1 week ago (5 children)

Once there is enough adoption, expect token costs to skyrocket and advertising enshitifaction to accelerate

Models on local hardware that give adequate performance and the availability of non-US models mean that path is doomed to failure. If they crank up the price then people will look elsewhere. While they might not get frontier-level performance out of the alternatives, there's plenty of stuff out there that's good enough for 1/4 to 1/10th of the cost. Or just the cost of power, if you're willing to pay for some hardware and are operating on a small scale.

The only play the AI companies have got now is twofold :

  • Regulatory capture, hence all the press releases of basic and frankly embarrassing sandbox failures and "we need to slow down" talk to try and get that rolling, and
  • Banning of those nasty, evil, non-US models in the name of national security.

And then what? The rest of the world will just move on to the next most useful thing on the list. There will be ~330 million US citizens to extract wealth out of, to service trillions in data centre debt. That shit's not gonna fly.

[–] dgriffith@aussie.zone 5 points 1 week ago (2 children)

All they need to do is get governments high on AI mass surveillance and they'll be golden

All they have to do is be the last AI company standing, then just absorb everyone else's hardware for cents on the dollar.

This recent "we should slow down" talk is basically just them stalling until the collapse. They've spent all they dare to try and be the one on top, now it's just waiting until the bubble pops.

Meanwhile China is out there being the black swan of the whole field. If they manage to get models with efficiency and/or utility above what these guys are doing, they're fucked and they know it. So it's a ticking time bomb, and there's not that much time left.

[–] dgriffith@aussie.zone 3 points 2 weeks ago (1 children)

Jesus Christ, Julia. Why is it that, in all these online ramblings, you never learned how to drip-feed your blog posts?

One a day, maybe. Don't blow through them all at once like some desperate gambler trying to strike it rich. Now all you've done is gotten yourself onto a bunch of people's blocklists and now you'll never be visible to them again.

[–] dgriffith@aussie.zone 20 points 2 weeks ago (4 children)

Lemmings.

Wolfenstein.

Doom.

Descent.

Unreal Tournament.

[–] dgriffith@aussie.zone 3 points 3 weeks ago

English is rather notorious for having these phrases

English is essentially just a whole bunch of corner cases that its speakers agree on. Mostly.

[–] dgriffith@aussie.zone 2 points 3 weeks ago

I have a QNAP box at home, with Forgejo running as a container on it. It's pretty much a standard docker install straight from Forgejo's repository. It can run tests and a build on each push, which is not strictly necessary, but nice. Took me about half an hour to setup Forgejo, and about 4 hours of swearing setting up CI for a few repos, because I've never done it before.

This doesn't easily scale to a whole organisation's worth of repos like OP wants, but it easily fits a small business/single dev situation.

 

Hi all,

In an effort to liven up this community, I'll post this project I'm working on.

I'm building a solar hot water controller for my house. The collector is on the roof of a three-storey building, it is linked to a storage tank on the ground floor. A circulating pump passes water from the tank to the collectors and back again when a temperature sensor on the outlet of the collector registers a warm enough temperature.

The current controller does not understand that there is 15 metres of copper piping to pump water through and cycles the circulating pump in short bursts, resulting in the hot water at the collector cooling considerably by the time it reaches the tank (even though the pipes are insulated). The goal of my project is to read the sensor and drive the pump in a way to minimise these heat losses. Basically instead of trying to maintain a consistent collector output temp with slow constant pulsed operation of the pump, I'll first try pumping the entire volume of moderately hot water from the top half of the collector in one go back to the tank and then waiting until the temperature rises again.

I am using an Adafruit PyPortal Titano as the controller, running circuitpython. For I/O I am using a generic ebay PCF8591 board, which provides 4 analog input and a single analog output over an I2C bus. This is inserted into a motherboard that provides pullup resistors for the analog inputs and an optocoupled zero crossing SCR driver + SCR to drive the (thankfully low power) circulating pump. Board design is my own, design is rather critical as mains supply in my country is 240V.

The original sensors are simple NTC thermistors, one at the bottom of the tank, and one at the top of the collector. I have also added 4 other Dallas 1-wire sensors to measure temperatures at the top of tank, ambient, tank inlet and collector pump inlet which is 1/3rd of the way up the tank. I have a duplicate of the onewire sensors already on the hot water tank using a different adafruit board and circuitpython. Their readings are currently uploaded to my own IOT server and I can plot the current system's performance, and I intend to do the same thing with this board.

The current performance is fairly dismal, a very small bump of perhaps 0.5 - 1 deg C in the normally 55 degree C tank temperature around 12pm to 1pm, and this is in Australia in hot spring weather of 28-32 degrees C.(There's some inaccuracy of the tank temperatures, the sensors aren't really bonded to the tank in any meaningful way, so tank temp is probably a little warmer than this. But I'm looking for relative temperature increases anyway)

Right now , the hardware is all together and functional, and is driving a 13W LED downlight as a test, and I can read the onewire temp sensors, read an analog voltage on the PCF8591 board (which will go to the NTC sensors), and I'm pulsing the pump output proportionally from 0-100 percent drive on a 30 second duty cycle, so that a pump drive function can simply say "run the pump at 70 percent" and you'll get 21 seconds on, 9 seconds off. Duty cycle time is adjustable, so I might lower it a bit to 15 or 10 seconds.

The next step is to try it on the circulating pump (which is quite an inductive load, even if it is only 20 watts), and start working on an algorithm that reads the sensors and maximises water temperature back to the tank. There are a few safety features that I'll put in there, such as a "fault mode" to drive the pump at a fixed rate if there is a sensor failure, and a "night cool" mode if the hot water tank is severely over temperature to circulate hot water to the collector at night to cool it. There are the usual overtemp/overpressure relief valves in the system already.

All this is going in a case with a clear hinged cover on the front so I can open it and poke the Titano's touchscreen to do some things.

Right now I am away from home from work, so my replies might be a bit sporadic, but I'll try to get back to any questions soon-ish.

A few photos for your viewing pleasure:

The I/O and mainboard plus a 5V power supply mounted up:

The front of the panel, showing the Pyportal:

Thingsboard display showing readings from the current system:

Mainboard PCB design and construction via EasyEDA:

view more: next ›