The only 18650 headlamps that I know of, Hank and Firefly, both require flat tops. Other "enthusiast" headlamps like Skilhunt and Zebralight also require them, though that matters less to me since they're non-Anduril anyway.

Anyone know what the deal with this is? Is it just lumenitis? It also occurs to me to call that "lumen bloat", similar to Javascript bloat. I'm an experienced enough user to not care about that any more. I'd rather have fewer lumens and keep the ability to swap 18650's (including button cells) between my various lights and other devices. For a headlamp I prefer a floody beam since if I want to check out something in the distance, I can pull a handheld thrower from my pocket.

Interchangeable battery tubes (18350/18650) would also be a plus.

Thanks.

Left to right: Olight i3e with Glow Rhino Spark locator, Lumintop EDC01, Lumintop EDC15 with Wurkkos HD12 clip, Wurkkos TS10v2, Convoy T3.

Micro-reviews: i3e is nice but too bright (90lm/70 minutes). I'd prefer 30lm/3.5 hours. Skilhunt E3A is same idea, has high CRI led available if purchased direct, but is harder to find at US dealers. It's possible to change the brightness of both lights by swapping an 0603 sized resistor, i.e. you need SMT rework tools, but some BLF'ers have done it. I'll want to attempt this if I can get access to the tools sometime. I'll note that I tried a mostly-depleted alkaleak AAA (1.0V OCV) in the i3e and got no light at all. Some other lights are able to reach lower voltages and still make some light.

Spark locator works but the next bigger size (Ember) should be a lot easier to find in the dark. It does cost 2x more and both are sort of mismatched to these tiny cheap lights.

Lumintop EDC01 is ok, needs a pocket/hat clip. Comes with a GITD diffuser/end cap (not shown) which is kind of nice, and also a split ring / gate clip. Also has a GITD O-ring in the bezel, but that has such short persistence that it's almost useless. It has mode memory and levels are in H-M-L order. I'd prefer L-M-H and no memory, or maybe even just two levels (20 and 100lm). Beam is floody which I like. Uses one AAA cell, 10440 not allowed. The old Fenix LD01 felt more solid but is long discontinued. I haven't tried the current E01v2 which has a similar setup to this Lumintop.

Lumintop EDC15 is the lightest weight current production AA light that I know of at 20.5g, and it's pretty small too. It has 4 levels + strobe, uses AA or 14500, and came with a 14500 with a USB-C charge port, nice. (Product page says micro-USB but must be out of date). It also has mode memory. If up to me I'd give it L-M-H (3 levels), no memory, no strobe. It also came with a diffuser, gate clip, and split ring, but no pocket clip. The Wurkkos HD12 clip fits nicely. I haven't tried the TS10 clip on it yet.

Wurkkos TS10v2: probably familiar to most of you. I also have the TS10SG and I like the SG a little better, though opinions vary about this. I wish the series hadn't been discontinued. Note, this light is 14500-only, not dual fuel. The Emisar KR1AA is a few mm longer and dual fuel and I'll probably get one sometime, though I prefer the TS10SG's throwier beam.

Convoy T3: Pretty decent for this style of mix and match light. Host looks suspiciously similar to Sofirn SP10 Pro host not in photo. I got it with an 1800K LED which I'm still getting used to. I have the 12 mode group driver that I haven't yet taken out of the light. It has a reflector which makes a somewhat focused beam that still has good spill. The slightly smaller T5 has an optic that's more of a flooder. My main wish for these lights is to be able to reprogram the modes since this would be a very good "granny light" if it were only possible to configure it with 1 level of around 50 lumens. It's rather large for a 1AA/14500 light but that's fine. There is some work happening on BLF about reprogramming some of the Convoy drivers, which use a Sonix MCU with an 8051-like instruction set. But that will probably come sooner to 18650-powered drivers, plus will require removing the driver and maybe de-soldering an MCU pin, so it's much nicer to have Anduril and flash pads. On the other hand the Convoy tail clickie switch has positive tactile feel, guarantees zero parasitic current drain, and the lights are much more affordable than a KR1AA.

It's on sale for $60 with SBT90.2 led which is a quite expensive led. I generally prefer small lights but whenever Wurkkos puts an Anduril light on sale, I get FOMO since I think the light is being discontinued.

There's some discussion here: https://old.reddit.com/comments/1t5exay

I'm holding off from the partly comparable TS26 because of its driver issues (moon mode probs) and it's LEDs, though who cares about moon mode in a monster light anyway?

Also, what is so great about the SB90.2 LED? Convoy and Hank charge quite a lot extra for it.

The TS30S blurb claims 295K candela in turbo mode (6K lm) though that seems suspect since 1086m throw is claimed, which would be more like 100k cd.

The Noctigon K1 with same led has comparable throw and output and is of similar size but 2x the price.

In practice idk what I'd do with such a big light. My TS11 probably fulfills my needs for a thrower.

Spoiler: Nutria.

https://archive.ph/bSQhD

This seems to be about recovering push notifications of incoming messages from the phone. Still not good.

HN discussion: https://news.ycombinator.com/item?id=47703573

I'm wondering if there is a decent affordable cell welder. Purpose is rebuilding tool batteries and that sort of thing on an occasional basis. K-weld has been recommended but is pretty expensive and might be overkill. https://diy500amp.com/ has a bunch closer to my price range but might be underkill. Any thoughts? Thx

The FTC said that “OkCupid provided the third party with access to nearly three million OkCupid user photos as well as location and other information without placing any formal or contractual restrictions on how the information could be used.” OkCupid “did not inform consumers or give them the chance to opt out of such sharing,” the FTC said.

cross-posted from: https://lemmy.ml/post/45257951

I don't like this retailer but this is a very sparse category (headlamps under 1 ounce) and I happened to be across from one of their stores today, so I picked up 4 of them (two red and two teal). Weight is 27g all up (light, headband, and batteries). The light separates from the backing clip and the light+batteries weigh 18g so the rest is about 9g. It might be possible to golf away a few grams by replacing the headband with some shock cord, but treating 28.35g (1 ounce) as the target weight, this light comfortably hits the target and I can skip further optimization.

Pluses: 1) the light runs on included lithium primary batteries. They won't leak and have known low temperature performance and long shelf life. 2) Cheap ($1.97 at the store mentioned). The nearest competitor that I know of is $11 (UST Tight Light and harder to find). 3) Very simple UI: press on, press off. No modes. Large, easy to use pushbutton. 4) smooth flood beam, fine for up-close repair work or walking to the loo in the dark. No real throw but this is more like a keychain light than a caving lamp. Maybe a light to stash away in case something happens to both your "real" light and its backup.

Minuses: 1) Large pushbutton might be easy to press by accident. If stashing the light away just in case, I'd leave the protective tab installed in the battery compactment, which also ensures no parasitic battery drain. Then pull the tab out before using the light. But it would be good to have a software timeout or lockout. 2) The headband is too short, maybe intended for kids, or maybe I'm just swell-headed. I can wear it when adjusted to the maximum but it's tighter than I prefer. I might sew in an extra inch or two of fabric, or replace the headband with shock cord. 3) While it's very small for a "proper" headlamp, it's much bigger than a Photonlight or similar. 4) Not that waterproof, particularly because of the slot left unsealed by pulling out the battery tab. Of course you could cover that with a bit of tape, but that could be awkward if you pull out the tab in the field. 5) Need Philips screwdriver (a 4mm driver bit is probably enough) to change batteries.

I think the use of CR2032's is a win for this light specifically, on the assumption that you'll use the light only very occasionally or as a backup for a backup, like a Photon light. The stated runtime is 3 hours and I can believe that. Two 2032's have about as much energy as a charged AAA Eneloop so there's plenty there. I'd expect a long tapering "tail" as well, but haven't yet tested. I will do a brightness and runtime test once I get some kind of measurement setup in place.

The 4 leds are SMT with the usual crappy tint. Maybe you could replace them with high CRI leds if you were nerdy about it. Someone actually mentioned that. It might even be possible to replace the 2032's with a small lipo pouch cell.

Main more serious alternative in the sub-1-oz category that I know of is the Nitecore NU05 v2, which unfortunately has just 1 hour runtime (40 lumen output that can't be adjusted downward). I haven't yet gotten to try one of those so can't compare directly. Maybe soon though. Black Diamond's 2032 light ($20 iirc) is discontinued and the Petzl e+Lite ($30+) might also be. The Petzl is of far better quality than this thing, but the cost is outlandish unless you're really trying to prove something.

My main immediate modification will be to staple a piece of Countycomm GITD tape to the headband, as I do with almost all my headlamps, to locate the light in the dark. Put the tape at the "far" end of the headband so you can use the LED's to charge up the tape. The staple is because the tape's adhesive isn't that strong under flexing, and it can fall off if you just stick it onto the headband. You could sew it instead of course.

3.5 stars out of 5 given the niche category and the low price. Main deductions are for the short headband, the waterproofing issue, and the easy accidental activation. This all could have been fixed very easily during design.

I weighed this with a crappy kitchen scale with 1g resolution and poor consistency, but will use a more precise scale when I get a chance.

I don't like this retailer but this is a very sparse category (headlamps under 1 ounce) and I happened to be across from one of their stores today, so I picked up 4 of them (two red and two teal). Weight is 27g all up (light, headband, and batteries). The light separates from the backing clip and the light+batteries weigh 18g so the rest is about 9g. It might be possible to golf away a few grams by replacing the headband with some shock cord, but treating 28.35g (1 ounce) as the target weight, this light comfortably hits the target and I can skip further optimization.

Pluses: 1) the light runs on included lithium primary batteries. They won't leak and have known low temperature performance and long shelf life. 2) Cheap ($1.97 at the store mentioned). The nearest competitor that I know of is $11 (UST Tight Light and harder to find). 3) Very simple UI: press on, press off. No modes. Large, easy to use pushbutton. 4) smooth flood beam, fine for up-close repair work or walking to the loo in the dark. No real throw but this is more like a keychain light than a caving lamp. Maybe a light to stash away in case something happens to both your "real" light and its backup.

Minuses: 1) Large pushbutton might be easy to press by accident. If stashing the light away just in case, I'd leave the protective tab installed in the battery compactment, which also ensures no parasitic battery drain. Then pull the tab out before using the light. But it would be good to have a software timeout or lockout. 2) The headband is too short, maybe intended for kids, or maybe I'm just swell-headed. I can wear it when adjusted to the maximum but it's tighter than I prefer. I might sew in an extra inch or two of fabric, or replace the headband with shock cord. 3) While it's very small for a "proper" headlamp, it's much bigger than a Photonlight or similar. 4) Not that waterproof, particularly because of the slot left unsealed by pulling out the battery tab. Of course you could cover that with a bit of tape, but that could be awkward if you pull out the tab in the field. 5) Need Philips screwdriver (a 4mm driver bit is probably enough) to change batteries.

I think the use of CR2032's is a win for this light specifically, on the assumption that you'll use the light only very occasionally or as a backup for a backup, like a Photon light. The stated runtime is 3 hours and I can believe that. Two 2032's have about as much energy as a charged AAA Eneloop so there's plenty there. I'd expect a long tapering "tail" as well, but haven't yet tested. I will do a brightness and runtime test once I get some kind of measurement setup in place.

The 4 leds are SMT with the usual crappy tint. Maybe you could replace them with high CRI leds if you were nerdy about it. Someone actually mentioned that. It might even be possible to replace the 2032's with a small lipo pouch cell.

Main more serious alternative in the sub-1-oz category that I know of is the Nitecore NU05 v2, which unfortunately has just 1 hour runtime (40 lumen output that can't be adjusted downward). I haven't yet gotten to try one of those so can't compare directly. Maybe soon though. Black Diamond's 2032 light ($20 iirc) is discontinued and the Petzl e+Lite ($30+) might also be. The Petzl is of far better quality than this thing, but the cost is outlandish unless you're really trying to prove something.

My main immediate modification will be to staple a piece of Countycomm GITD tape to the headband, as I do with almost all my headlamps, to locate the light in the dark. Put the tape at the "far" end of the headband so you can use the LED's to charge up the tape. The staple is because the tape's adhesive isn't that strong under flexing, and it can fall off if you just stick it onto the headband. You could sew it instead of course.

3.5 stars out of 5 given the niche category and the low price. Main deductions are for the short headband, the waterproofing issue, and the easy accidental activation. This all could have been fixed very easily during design.

I weighed this with a crappy kitchen scale with 1g resolution and poor consistency, but will use a more precise scale when I get a chance.

I wonder if someone here with suitable test equipment could run a graph on a chemical glowstick, the ordinary 12 hour kind, not a special fancy one. I'm basically hoping for a rough estimate of the lumen output and its decay over time. I did a bit of web search and got drastically varying numbers. I guess I could make an eyeball estimate, but would want to use a reflector of some kind for the ceiling bounce.

Does anyone know if reflectors like the below are still being made? Of course I can DIY one if necessary. Web search didn't find anything quickly, but it's difficult to keep the results relevant.

https://www.worthpoint.com/worthopedia/nos-vintage-glow-stick-lantern-1960s-4706797024

I'm quite dubious of chemical light in this day and age, but they come up in discussions sometimes, so I'm interested in a concrete comparison against flashlights. Thanks!

Added: I'm happy to reimburse the cost of a few glowsticks if you're up for running this test. Or it's possible that I could order some and have them sent to you (Home Despot free delivery). PM if you want this.

Not sure what has happened, whether it's physically broken or something went wrong with the software. So the machine overheats fairly quickly now. This is running Debian 11 (bullseye) and /proc/acpi/ibm/fan says:

status:		enabled
speed:		0
level:		auto

so I don't know if the heat is being recognized. The CPU does clock down as it gets hotter. I'm not getting any type of overtemperature alerts though, and I haven't found any place in the acpi tree to read the temperature. That's annoying since there must be some sensors in there.

It looks like there is a program called "thinkfan" in trixie so I might try to upgrade the machine tomorrow. I can only do so much at a time before the box gets too hot.

Any help? Thanks.

They were able to de-anonymize posters from collections with 10,000's of users, tested with scrapings from Hacker News, LinkedIn, and Reddit. Not good from a privacy perspective.

They are making LEDs less than 100nm across, for use in ultra high dpi displays. That's way smaller than the wavelength of light that they emit!

The new program, called “masked engagement,” allows homeland security officers to assume false identities and interact with users—friending them, joining closed groups, and gaining access to otherwise private postings, photographs, friend lists and more.

A senior Department of Homeland Security official tells me that over 6,500 field agents and intelligence operatives can use the new tool, a significant increase explicitly linked to more intense monitoring of American citizens.

A tiny keychain knife with an Exacto style blade, nice for precise cutting but too delicate to be called general purpose. I just got two of them, pretty cool. Dimensions about 50mm long, 14mm wide, 4.5mm thick with the slider bulging up another 1.5mn or so. It's a pretty no nonsense design unlike some fancy and expensive ones I've seen in similar formats. Photo of the back side below:

Added: another good alternative, Derma-Safe folding razor, 7.6g, lacks a lanyard hole. I'm not sure if there's a good place to drill one. Review. The Derma-Safe is too long to fit into an Altoids tin "crossways" while the Ghost EDC will fit that way, if that matters to you.

Abstract: Life evolved under broad spectrum sunlight, from ultraviolet to infrared (300–2500 nm). This spectrally balanced light sculpted life’s physiology and metabolism. But modern lighting has recently become dominated by restricted spectrum light emitting diodes (350–650 nm LEDs). Absence of longer wavelengths in LEDs and their short wavelength dominance impacts physiology, undermining normal mitochondrial respiration that regulates metabolism, disease and ageing. Mitochondria are light sensitive. The 420–450 nm dominant in LEDs suppresses respiration while deep red/infrared (670–900 nm) increases respiration in aging and some diseases including in blood sugar regulation. Here we supplement LED light with broad spectrum lighting (400–1500 nm+) for 2 weeks and test colour contrast sensitivity. We show significant improvement in this metric that last for 2 months after the supplemental lighting is removed. Mitochondria communicate across the body with systemic impacts following regional light exposure. This likely involves shifting patterns of serum cytokine expression, raising the possibility of wider negative impacts of LEDs on human health particularly, in the elderly or in the clinical environment where individuals are debilitated. Changing the lighting in these environments could be a highly economic route to improved public health.

"Quantum theory provides a foundation for describing systems that are probabilistic, interdependent, and evolving (Busemeyer & Bruza, 2012; Haven & Khrennikov, 2013). Translating these ideas into tourism produces a model that explains how behaviour, feedback, and innovation interact across cognitive, relational, and systemic levels. This complements entropy reduction in tourism (Li et al., 2025), which conceptualises tourism as an open system that shifts between stability and disruption. While entropy theory focuses on energy and order, the quantum perspective explains the structure of uncertainty: how multiple possibilities, relational ties, and networked feedback generate adaptation and innovation."

Annals of Tourism Research Volume 117, March 2026, 104115 (nothing about April 1). No mention of Sokal in the article or its references. Not the Onion. I'm at a loss.