fghorow

I am old enough to be retired -- career spent as a research geophysicist. I have about 2-3 scientific projects/papers/codebases left in me before I go be a hermit in a cave somewhere.

LLMs appear to be able help me get through that. I'll take the help.

BUT JEEZ! Dealing with "Rain Man"[1] LLMs week-in and week-out is wildly annoying, frustrating, and anger inducing.

If they all would just STOP sounding so authoritative when spewing nonsense!

[1] After the Dustin Hoffman character in the movie of the same name. "Ah but you know I'm an excellent d̴r̴i̴v̴e̴r̴ programmer."

It is a nice field they have.

It happens to be right next to the DOE sponsored Frontier Observatory for Research in Geothermal Energy (FORGE [1]) site that established the temperature and ability to create permeability at depth using EGS.

[1] https://utahforge.com/

Honestly, to do anything like justice to this topic would require a publishable academic paper. That is something I simply do not have time for.

There are (at the very least) 2 reasonable places I can think of off the top of my head in which to look around for answers -- both are organizations that run annual meetings on the topic of geothermal energy.

One is the Geothermal Rising professional organization [1]. There is a claim on the website that the papers from the annual meeting (roughly a month back) are available for free download by anybody. I can't find the actual head-link for those papers. A search engine is your friend. (I would not trust an LLM for this particular job.)

A second one is the Stanford Geothermal Workshop [2]. That is also an annual meeting. Again, use a search engine.

Sorry to not be more helpful, but this is a pretty large topic.

[1] https://www.geothermal.org/2026-geothermal-rising-conference... [2] https://geothermal.stanford.edu/events/workshop/past-proceed...

Disclaimer: I work in this field.

This is a very brief, high-level overview. It weirdly dives deep into some less-relevant topics and glosses over a LOT of things.

It is not irrelevant, but not terribly reliable either. The "Mathematics" in the title is self-evidently clickbait.

YMMV.

I know. I know. I grew up in the '60s. Feel free to unfollow me (or whatever it is that one does on HN).

"Danger Will Robinson!"

Fair point.

Preach!

The only possible second-or-third order effect of red cross-traffic arrows is to decrease the downstream traffic on the cross road. Mostly, the places where I encounter them have a low traffic cross road. IMHO, in those situations, red cross-traffic arrows are simply a failed "traffic engineering fad".

(Edited to fix a typo and generalize "red left-arrow" to "red cross-traffic arrow" to account for left-side-of-the-road driving systems.)

Does anybody know anything about "Cliff Compaction"?

I need it yesterday!

(At the moment, I'm using the OSS version of headroom. That was a real PITA to configure for local-only, but I think it is working reasonably well now.)

Cliff Compaction though???

Check out proxmox. Debian base with Debian apt repos available for standard stuff, with ZFS built into their kernels. Easy to install. And, if you insist on running TrueNAS (or whatever) just stash it in a container or KVM box.

(Edited to add: I forgot! ZFS on root available from the installer -- encrypted even if that's your jam.)

Thanks. I'll kick the tires on OMP a little harder...

I've been using Claude Code Extension in VSCode (no phone-home configured), backed by DwarfStar on a LAN local MBPro 128GB M5. The context bloat is horrendous, leading to 5-10 minute prefills.

I've recently been exploring tools like headroom to help manage context, with some limited "success" (for some definition of success). What do others with similar setups do?

(I kind of hate to abandon Claude Code, as it seems to be the most capable coding assistant of the limited set of tools I've tried. But that horrendous context bloat is really painful!)

It's getting harder and harder to tell these days when El Reg is being serious. I think this is a case when that is so, but I'm not sure...

Written by "A worried Belgian researcher".

Perhaps this approach should be known as "The Method of MOOments"?

Agreed!

Also, I just found this (total hack, but cool)! [1]

[1] https://github.com/nsuan/TelescopeFilter

Yes indeed! Spain, next year! (And Australia, the year after that! ;-) )

I'm also wondering about the feasibility of taking these gizmos on airplanes, what with the Lithium battery packs, etc.

What we really need (for some definition of need) is a robot-filter-remover, timed to deal with the two critical times (first and last instant of totality; 2nd and 3rd contacts from memory). Anybody out there interested in building something like that? It would be purely a passion project, because it could only be used a few times a decade...

I'm interested in your eclipse experience. I tried capturing totality with a seestar S50 back in the 2024 eclipse. What I found is that at the very moment of totality, I had not removed the solar filter fast enough, and the bloody thing lost lock on tracking the Sun. Rather than fiddling with the equipment, I immediately started to ignore the 'scope and just experienced the eclipse.

That was the right thing to do!

But removing the solar filter right at the moment of totality is such an edge case (both for the software and my wetware which was kind of overwhelmed) that I wonder how others dealt with it?

On a related note, there is a reasonably well known commentary on an adjacent technology by Ruth Schwartz Cowan (1985) titled "How The Refrigerator Got Its Hum". It describes a 1930s market battle between heat-driven ad/absorption chiller technology ("Kelvinator") and work-driven compressor chiller technology ("GE"). Tl;DR GE won.

The ab-chiller is likely best known these days as the "propane fridge" in RVs.

Sadly, the only online resources for Cowan's work I can find are either Scribd paywalled [1], the entire book in which Cowan's piece is a chapter on the Internet Archive [2] or a pointer-to-a-pointer from a piece in the Guardian [3] to a boingboing piece [4] that seems to be offline.

(Background: I work in Direct-Use geothermal. Such ad/absorption chillers are one major tool in the toolbox for low-T cases where it is uneconomical to turn geothermal heat into electrical energy -- mostly due to Carnot's Law considerations.)

[1] https://www.scribd.com/document/527479452/Cowan-How-the-refr...

[2] https://archive.org/details/socialshapingoft0000unse_g5b0

[3] https://www.theguardian.com/science/political-science/2014/f...

[4] https://boingboing.net/2012/08/20/how-the-refrigerator-got-i...

It's not exactly straight out of one of the labs -- it's heavily quantized -- but have a look at [1].

[1] https://github.com/antirez/ds4

...or a station wagon full of magtapes.

Testing:

The union of a lot of necessary conditions is not a sufficient condition. But it might be good enough for software.

I'm curious too!

One obvious thing is that the memory requirements for this are substantially smaller than DwarfStar-- which AFAIK can only start to be used at 64GB ram and upwards. Another obvious thing is that antirez is pretty obsessed with making sure that DwarfStar passes all of DeepSeek V4 Flash's generating tests (loosely). I suspect that is also true of DwarfStar's implementation of GLM5.2, but I don't use that.

Probably 42, but I could be mistaken...

Counterpoint, I'm just a happy user of this local LLM [1] from antirez (of Redis fame).

[1] https://github.com/antirez/ds4

District heating is a mature technology. Direct Use geothermal heat is the one I am personally most familiar with -- as a geophysicist. However "waste" heat utilization is a definite thing for people with mechanical engineering/heat and mass transfer training.

(Edited to add: there are several examples of public swimming pools being heated with Low T geothermal heat in the Perth metropolitan region of Western Australia.)

True. Chena Hot Springs [1]. They are famous in the "direct use" geothermal community.

A lot of the thermal energy is not used for electrical generation. Although a small portion actually is -- made possible by the \Delta T rejecting heat at a low annual average atmospheric T.

Most of the rest of the heat is used to run an absorption chiller to maintain the ice "palace" in the summer.

(This info might be slightly outdated. It was true about 2018 or thereabouts when I met the owner of the resort at a geothermal conference.)

[1] <https://www.chenahotsprings.com/>.