Cognition (Devin AI) to Acquire Windsurf (cognition.ai)
300 points by alazsengul 4h ago 242 comments
Replicube: 3D shader puzzle game, online demo (replicube.xyz)
53 points by inktype 3d ago 9 comments
Lightning Detector Circuits
60 nateb2022 35 7/14/2025, 3:07:31 PM techlib.com ↗
https://www.lightningmaps.org/#m=ses;t=3;s=0;o=0;b=;ts=0;
triangulates lightning strikes detected by this method. Right now I see a front moving between the Triple Cities area and Scranton.
https://www.sparkfun.com/sparkfun-lightning-detector-as3935....
Can somebody provide clarity on if any of the boards really work?
It seems that bri3d has faith in the ScioSense.
If you want to know more specifically you need to have separate sensors that have with reasonably synchronized clocks so they can trilaterate or triangulate strike locations. Those sensors probably need to be further away from each other than most boats allow for. The further away the sensors are from each other, the tighter the time synchronization between them, and the more sensors you have will determine the overall accuracy your system will be able to attain.
I don't think that's feasible. These sensors work by detecting radio waves which propagate near the speed of light. Even if you had two sensors separated by 1 km, they'd both receive the radio signal within ~3 µs of each other. I don't think the sensors guarantee detection time to that level of precision, nor that it's feasible to keep clocks sufficiently synchronized that you can timestamp events at a sub-microsecond level.
With triangulation you only need the clocks accurate enough to correlate the events. This will largely be determined by the rate you see events, and you only need to distinguish specific events if you want to differentiate locations of individual strikes as opposed to a storm front. For a boating case you probably only care about the location of a storm front.
Two sensors measuring angles from a couple dozen meters apart can get you a pretty precise location of storm front especially when aggregating results from multiple detections.
Everything I did failed and I have never been able to get anything other than noise out of these sensors at best. YMMV, as I know these chips have a history of problems but supposedly do actually work. After I finally gave up on these things I ended up getting an Acurite 06045M, a cheapo SDR dongle, and a Pi Zero W running rtl_433 for like $60 and that has worked perfectly for years and is fairly straightforward to plug into Home Assistant over MQTT. It very reliably detects strikes and gives a pretty good estimate of distance (but NOT bearing), plus you get an outdoor temp/humidity sensor for free.
I think the 06045M might actually use the AS3935 internally so presumably you can get them to work and when they do work they work very well, but nothing I could do as a (pretty advanced and well-equipped) hobbyist could make them work, and I've heard others had similar issues with them. So maybe stick to buying something commercial with it inside?
I've been mighty tempted to try again with a newer one as I really want to build a compact portable lightning sensor linked to Meshtastic but I dread going through the whole ordeal again.
Posted this some time ago, but it got no traction :(
I was very excited about it over ten years ago and wanted to join the network. They are very weird about not releasing the designs and code, and instead want you to get it from them. Fair enough, but then they don't produce and ship units. I signed up for a waiting list like 10 years ago, with many others. They kept promising a new design, but never delivered it, and didn't let anyone else do so. I never heard back from them about the waiting list.
They are also restrictive about access to the data if you are not in the network, with reasons like it's too expensive. That might have been true 10-15 years ago, but today it's just silly.
It's a shame, I think there is lot of potential to make a better community and more usable free apps around this data. But you would have to start a new network from scratch because these people are not interested in that.
Maybe my browser just got better? Who knows :-)
By working right, I mean whenever I see flickering from afar, or hearing the first rolling of thunder, clicking on it without much fuss, maybe zooming out, shifting a little, zooming back in and seeing what's going on where instantly.
Which in the context of larger storms works even better in kachelmannwetter.com, because there it's embedded in other data, and projected where it moves/arrives in how many minutes in easily understandable symbology. But that's for Germany only. Useless for Colorado. Havn't found something comparable for there, so far. Would be mostly useless anyways, because weather at 2600+ m altitude changes faster. Or can.
Here it is discussed though by someone who tried it (and improved it):
https://www.angelfire.com/oh3/ebjoew/Charged_Cloud_Alert.htm...
(Ha ha, angelfire)
> We welcome offers of hosting a new WWLLN lightning sensor. All hosts receive all the world-wide data for their own research.
https://wwlln.net/#home
Pahaha pure Spinal Tap moment
https://en.wikipedia.org/wiki/Point-to-point_construction
https://entertaininghacks.wordpress.com/2020/07/22/prototypi...
I assume the author liked the idea and used the perfbaord/punchboard as a grid.