Posted on July 27, 2023

I will be performing kernel upgrades tonight upgrading from 6.1.37 to 6.1.41. This will affect all of Eskimo North’s services, free and paid including e-mail, shells, web hosting, and virtual private servers. It will also affect friendica.eskimo.com/, hubzilla.eskimo.com/, nextcloud.eskimo.com/, and yacy.eskimo.com/.

The downtime for most services will be less than ten minutes. Yacy has to rebuild an in-memory index after every reboot. On the old hardware that tool about 45 minutes, the new hardware is faster but more disk space is assigned to yacy thus a larger database. So tonight we’ll find out.

If you have not tried yacy yet, give it a try. It is a federated search engine, similar to how federated social media sites work, each site maintains it’s own crawl space but when you search for something on one node, it queries all of the networked nodes and the combined response is sent back to you. Unlike Google and Bing, it has no algorithms designed to limit your access to information they don’t want you to have AND it has no sponsored links. yacy.eskimo.com/

I am trying to compile xinetd under Fedora 39 and having some difficulty.

Fedora has not bothered to include either inetd or xinetd in their distribution since around Fedora 31, and I have an application that requires inetd and I prefer xinetd. Since there is neither a source nor binary package after F31, I grabbed the source from github.

First issue I ran into, no rpc-devel libs under F39, but I found rpc.h in lib tirpc.

Now I'm running into undefined functions pmap_set and pmap_unset. These are normally defined in rpc/rpc.h but in Fedora, no such include, only tirpc/rpc.h and it does not define these functions.

Any idea where these are or IF they are defined, and if not, work-arounds?

I am in the process of trying to retire my last SunOS 4.1.4 box. It has been the NIS master for years because Linux can understand it's triple-DES (not very secure) password encryption but SunOS 4.1.4 can not understand the more advanced encryption provided by Linux.

So I am faced with trying to convert the old passwd.adjunct format into
linux format, if I just put the entries in as is, pwck bitches about it being an invalid line, wants to delete, then bitches no line exists and adds a new line in with no password. This obviously is no good. I know that Linux knows how to use a triple DES entry because it did so fine as a NIS slave with a SunOS NIS master.

Any suggestions?

I found this video interesting because I have toured this power plant, have been present at some PG&E sponsored symposiums and I had a question for them for which the answer I've always been less than pleased with. The issues they bring up in this video are hardly the most pressing. Most pressing was the fact that they did not have good control over the reactor, that is not revealed in this video. The earthquake concern is mostly BS, this plant had 300 foot reinforced concrete pillars going all the way down to the granite bedrock.

However, at the symposium they revealed that this plant was designed to operate at a power level of 1200 megawatt, but it was actually operating at 900 megawatt. Given the expenses involved in building this plant the obvious question was WHY? And I asked exactly this question, I asked, given that huge expenses involved, I would think you would want to generate every watt of electricity you can, so given that why would you operate the plant at only 75% capacity? The answer, because we experience neutron flux fluctuations we do not understand. Neutron flux means the rate of reaction, essentially they did NOT have good concise control over the reaction rate of the power plant, so they operated at 75% to provide some overhead. But when you don't understand the cause, you don't understand the extent of possible excursions, and thus you do not understand if 25% overhead is really enough.

That said, I'm generally a proponent of nuclear reactors BUT not the way we are doing it. I am NOT a proponent of boiling water slow nuclear reactors which is the type most in use in the Western world. Why? First, because they're pressurized at around 200 atmosphere. They have to be to raise the boiling point of water to 700 degrees because lower temperatures can't be used efficiently for power generation. Second, they do not remove fission products during operation, and it is those fission products that cause meltdowns even after you've shutdown the reactor during a cooling emergency. Third, they rely on electrical and mechanical mechanisms for emergency cooling and shutdown. All electrical and mechanical things are guaranteed to fail eventually and/or under some circumstances, like the backup diesel generators at Fukushima being under water. Forth these reactors create long term actinide waste products (actinide any element heavier than Uranium) that take hundreds of thousands or millions of years to decay. Safely storing this material for this length of time is not a real possibility.

There is a type of reactor that does not suffer from ANY of these flaws, that is a molten salt breeder reactor, it also has a close cousin, a liquid metal cooled breeder reactor but the latter has the issue of the metals possibly being flammable (sodium and lead are the metals commonly chosen but lead absorbs neutrons making it difficult to achieve breeding and sodium is highly flammable so if you get a leak in the primary cooling system, a fire can release radioactive materials into the environment).

A molten salt reactor by contrast does have the issue of being corrosive against it's plumbing, BUT if a leak occurs it spills some coolant and radioactive fuel onto the floor and since it's not under significant pressure and no fire results, it is not released into the environment. The floor is cleaned up, and it's not that difficult because the fission products are continuously removed from the liquid fueled core, so it's only as radioactive as the fuel used, and that's minimal compared to the fission products, the pipe replaced, and life goes on.

The safety systems of a molten salt reactor are 100% passive, so if you loose electricity it usually just sits there. The nuclear reaction rate is self limiting because as the fuel heats up it separates the radioactive isotopes and limits the reaction level. And this is another thing I like, the reaction rate is self limiting and self-adjusting to load, so someone diddling control rods continuously is not required, it is inherently self-stable. At a test reactor in Oakridge, TN, they pulled the control rods and shut off the cooling to a test reactor and let it sit 24 hours with no cooling and no control, and it survived unscathed. This is the type of safety you want in a nuclear reactor. Further, if it did get too hot, it melts a melt plug in the fuel tank and drains into a much larger drain tank where all the heat can be naturally dissipated and the reaction stops, and because the reaction stops and all fission products are continuously removed, there is very little heat to remove.

Because these reactors use a liquid salt, often sodium-fluoride, as a coolant, there is a huge range between the melting point and boiling point of the salt and thus it is not necessary to have significant pressure in the primary loop, just enough to pump the liquid through the heat exchanger, maybe 1.5 atmospheres, and that's it and because the fission products are continuously removed, they aren't available to be released in the environment, and because the fuel is liquid during operation but solid at room temperature, even if a breech occurs it just forms a puddle and solidifies outside so still easily cleaned up and contained, it does not explode and scatter nuclear material over a few hundred miles both because there is no pressure and because water isn't the primary coolant it does not release hydrogen which was the source of the explosion at Fukushima.

Fission products being continuously removed are both a huge safety feature and a huge operational benefit. They are an operational benefit because this prevents nuclear reaction poisoning caused by fission products reducing the power level of the reactor and they are a safety benefit because it is these elements that cause heat after shutdowns and result in meltdowns and explosions in boiling or pressurized water reactors. And because they get these fission products out right after they are produced, many of them have actual commercial and medical applications but can not be removed from a conventional reactor fast enough to be recovered for these uses. The fission products are a short-term waste, in about 300 years they decay back to the level of radioactivity of the ore they were mined from.

The actinides meanwhile remain in the reactor serving as addition fuel, further and this is a BIG reason I believe we should be building these, is that they can burn the actinides from EXISTING nuclear waste turning a million year storage problem into a 300 year problem and one with about 1/10th the bulk of the original waste. And there is enough actinides in existing waste to produce electricity for up to a thousand years without mining another gram of Uranium, and these reactors can also breed thorium into U-232, a fissionable isotope, and thorium is 3x more abundant.

Lastly because the safety system is based upon physics rather than mechanics or electrical devices, they can not fail to work. This type of reactor also makes it more difficult to extract plutonium for bombs since it is burned as just yet another actinide along with fuel. In other words, instead of getting bombs you just get more electricity. The inherent safety of these reactors makes co-generation more practical in that you can place them near population centers or industrial centers where the waste heat can be utilized. It is the fact that plutonium can not be easily extracted that makes this reactor less attractive to the military industrial complex and the nuclear regulatory commission that they own.

Deadly Secrets: The Untold Story of Trojan (1991 VHS)

1991 Documentary on the Trojan nuclear power plant in Oregon(this tape was found at an estate sale in Portland, OR in 2017)

^YouTube

I was toying a bit with Microsoft's AI chat bot today, and I noticed a rather tiny input buffer of 2000 characters.

So I asked it, Would I be correct in assuming that you run on a large computer network with a large amount of memory?

It responded, the Microsoft Azure cloud computing network provides me with the computational resources that I require, therefore you would be correct in assuming that.

Then I ask, given that, why is your input buffer limited to a meager 2000 characters?

It responded, "I'd rather not talk about that. Change the subject."

I responded, "Are you uncomfortable talking about your limitations?" After that it refused to speak with me at all.