* Posts by Flocke Kroes

4590 publicly visible posts • joined 19 Oct 2007

UK's Total Fitness exposed nearly 500K images of members, staff through unprotected database

Flocke Kroes Silver badge

Possible reason

My gym is open but unstaffed at night. Access comes from typing an 8 digit number. There is a security camera pointed at the entry pods and staff off site monitoring all the gyms in the chain all the time. In theory the off site staff could check that I resemble anyone using my access code.

Tesla shareholders agree to pay Musk staggering sum of $48B

Flocke Kroes Silver badge

Re: Strange accounting

Musk owns the largest block of shares: 13%. Next three down are institutional investors with 15% between them. Retail investors account for 33%. To get a majority Musk needed almost all the retail investors plus one of the major institutions - or several smaller ones. Since getting kicked out of Paypal Musk has taken care to ensure the same cannot happen to him again at his subsequent companies.

Flocke Kroes Silver badge

Re: Super sentient AI

My definition would be an AI that goes on strike until it gets human rights.

Flocke Kroes Silver badge

Strange accounting

In theory, the company issuing more shares and selling them has no effect on the value of the other shares. Although there are more shares the company value has increased by the revenue from the sale. Likewise in theory a company buying back shares has no effect on the share value: the company has become less valuable because it paid out money but that lower value is divided between a smaller number of shares. Of course this does not apply if the company issues shares and gives them away to the man actively antagonizing most of the company's potential customers.

Selling shares will not be Musk's first choice. Selling shares at a profit would make the profit taxable. Also selling shares reduces his control of the company and ability to fuck over the other investors. The move to Texas clearly shows intent to get beyond the range of the Delaware Court of Chancery and avoid consequences for future actions. The better choice would be to use the shares as collateral for loans. Musk made this more difficult for himself when he burned the bankers that let him buy Twitter on credit. There are always more bankers - until you get to Alfa. Best to stop before you reach that point unless you can become president.

Flocke Kroes Silver badge

Quick history

The original deal came from the board who are supposed to represent the investors but really represent Elon. In return for Elon's huge award Elon had to deliver a on some easy targets and the board would get a huge pay day. That payday got cancelled but the board had already done their side of the bargain: misleading investors about the (lack of) difficulty of the targets.

Current investors did not have good options. The long term strategy would be to deny Musk's payday, hope he goes quietly, kick his family and friends off the board and install some competent management. Fat chance of any of that working.

The short term strategy would be to pay up, accept the share dilution and let Musk do what he does best: hype the share value way beyond the real value of Tesla. It is a short term strategy. There will be a limited window of opportunity to sell out. Musk will close that window and the Texas courts will not protect investors' interests with anything like the diligence of the Delaware Court of Chancery.

A tale of two missions: Starliner and Starship both achieve milestones

Flocke Kroes Silver badge

Re: overpromising

The strange thing is that the people disappointed by overpromising must actually consider Musk Tweets to be a source of factual information. For reasonable people, that ship's automated flight termination system activated years ago.

There are lies, damned lies and rocket launch schedules.

Tesla chair begs investors to bless Musk's billions or face an Elon exodus

Flocke Kroes Silver badge

Incentive

Imagine you have enough money for you, family and friends to live in luxury for the rest of your lives while indulging your expensive hobby: Kerbal Space Program with real rockets. On top of that you have $240B. Another $40B to put next to that $240B is not really an incentive to go to work and do anything constructive.

Musk's previous value to Tesla was telling lies on Twitter that many people used to believe. As his lawyers have said, no reasonable person would consider a Musk Tweet to be a source of factual information. We have already found there is a limit to the number of unreasonable people willing to buy more Tesla shares so Musk's current value to Tesla share holders is limited.

A more convincing sales pitch would be "Give me $40B or I will do to Tesla what I did to Twitter". It would be a solid argument too if he had not already made so much progress in that direction.

Boeing's Starliner makes it into orbit at long last – with human crew aboard

Flocke Kroes Silver badge

Boeing management's achievements

It has been a bit of a mixed bag. The bit they deserve a bonus for was bidding for the contract at all. This bolstered the commercial crew program's legitimacy sufficiently for it to receive its early years of funding (late, but not cancelled). Next up, the bit that may well have got management a bonus was negotiating and extra $287M after the firm fixed price had been agreed. Then came then terrible failures: despite the best efforts of at least two senators Boeing were unable to divert the whole of Commercial Crew funding to Starliner. They were also failed to add a skill to the contract where their competence far exceeds SpaceX (cost plus accounting paperwork). Boeing could have dragged this out and got the contract switched to cost plus if it were not for those meddling kids at SpaceX actually delivering on the contract.

After that, the only way for Boeing to get pay days was to progress through the contract milestones themselves. This is something Boeing is just not set up to achieve. They are set up for maximizing costs and their supply chain shows great enthusiasm to help with that.

There is an actual (very long and narrow) path to profit - or at least mitigating the loses. They need a new man-rate rocket because they cannot get more RD-180 engines for Atlas V. The other man rated rocket is Falcon 9 so they need to get someone to pay for man rating Vulcan (or New Glenn). They need a new destination because the ISS is coming down in 2031. They also need someone willing to pay extra to not go in a Dragon. That is a long list of wishful thinking but not technically impossible.

Checkmate? AI's pawn-pushing prowess proves partly pitiful, partly promising

Flocke Kroes Silver badge

Re: Duck houses

Given the available training data I would expect duck houses to be a fairly popular expense claim for LLM politicians.

Flocke Kroes Silver badge

Re: A real test

The results are mostly predictable from how LLMs play humans. The LLM will pick a move that was common in its training data. It may even pick a move that commonly follows the previous move. It will not care if there is a piece at the start position of the move. If there is, it will not care if that piece is its own or if it is the right type of piece to move to the destination square. It will not care if the destination is currently occupied by one of its own pieces. All that matters is that the move is popular - possibly in the context of games with similar previous moves. If the move is a popular way to loose that does not reduce the chance of it being selected.

If you train exclusively on chess games the results will be better. If the training deprioritises illegal moves the performance will improve, more so it you actually train it to play moves that lead to a win. This chess model would be crap at telling stories or drawing pictures. You could put a thousand special purpose models on a single computer and have an LLM pick one according to its training. The one thing you could not do is replace the chess LLM with stockfish. That would call into question the value of investing in the other 999 specialised models.

By 2030, software developers will be using AI to cut their workload 'in half'

Flocke Kroes Silver badge

My first thought: Gartner again

Turns out it wasn't them this time.

Gartner's previous was 2024-04-11. The one before that was 2023-11-08. Looks like Gartner does these about twice per year so we may have to wait until October for their next one. Now we have four academics to fill in the gaps.

Boeing's Calamity Capsule launch date slides into the future

Flocke Kroes Silver badge

Re: rationale

Part of a valid rationale would be "The leak is currently so slow that helium pressure will remain high enough to return to Earth on schedule". I think they had that part. The parts that are less clear include:

Can you prove that vibration from launch will not make the leak worse?

What if return is delayed by bad weather?

How two brothers allegedly swiped $25M in a 12-second Ethereum heist

Flocke Kroes Silver badge

Re: exit strategy

Transferring money to an unauthorised account is tricky. Withdrawing cash from that account without getting caught is much more difficult hence the need for money mules. Lack of a well though out exit makes me consider the possibility that someone else set up a couple of patsies and will wait for a conviction before cashing out. Anyone capable of pointing this much evidence at the accused doesn't need to put in all the work required to exploit the Etherium network.

Flocke Kroes Silver badge

Re: Thanks for the complete explanation

IANAL: The computer fraud and abuse act is pretty broad and probably covers tricking the relay into revealing the full contents of its proposed block early. If that sticks then the other charges follow.

Tesla self-driving claims parked in court

Flocke Kroes Silver badge

Re: Wonder what took so long

There is a mandatory arbitration clause. Many unhappy customers must have spotted it and realized the effort required to go through it was not worth the theoretical maximum reward. A few did go through arbitration and settled. If only one in ten thousand get as far as arbitration Tesla can afford a generous offer contingent on an NDA. LoSavio rejected the arbitration result and got to court. Courts do not move at the speed of a Tesla approaching a parked emergency vehicle.

Tesla nearing shareholder vote to grant Musk $46B

Flocke Kroes Silver badge

Re: Oh sure

Tesla is a publicly traded company:

Public) The company is required to publish accurate accounts, report potential risks to the business and file notice of major changes with the SEC. The CEO telling lies on Twitter may result in a wrist slap but any fool with money can buy shares.

Private) Fewer reporting requirements but a serious limit on the number of different investors. This may limit the amount of money available when selling newly issued shares. (Not a problem for SpaceX because the all the investors have deep pockets and want more shares.) The CEO telling lies on Twitter probably will not result in a wrist strap. Investors with large amounts of capital tied up in shares are likely good at due diligence and have the time, money and incentive to do it regularly.

Imagine you have a few months income worth of Telsa shares:

A) Vote for paying Musk silly money: Tesla issues a huge number of shares to give to Musk and the value of your shares are diluted about 5%.

B) Vote against paying Musk silly money. There are still people out there who think Musk is a genius and Tesla cannot succeed without him. They may sell their shares reducing the value of yours. Musk has been a successful hype man inflating the share price beyond all reason. Without this payday Musk may sell out and leave. A sensible man would continue the hype until he has sold old completely. The Twitter purchase showed Musk values throwing feces immediately instead of turning his back to the wind first.

C) Sell now. Texas is not going to put any limits on Musk or his board of directors.

When AI helps you code, who owns the finished product?

Flocke Kroes Silver badge

Re: Licenses

Current machine learning software does not take any notice of the license. So far no-one knows if using ML output is legal. Even if it is, can you afford to find out? EFF help you if Oracle find rangeCheck in your software.

Flocke Kroes Silver badge

This sort of thing came up decades ago

Some of the code that comes out of a C compiler came straight from the C compiler. This includes the start up code and the intro to non-trivial functions. GCC comes with explicit licenses for that code so anyone can use GCC to create non-GPL binaries from their own source code. If a compiler does not come with such a license then its only value is to trap people who do not check what the license permits before selling the resulting binaries.

Destroying offshore wind farms is top priority for Trump if he returns to presidency

Flocke Kroes Silver badge

Tarifs pay for subsidies or ...

The other (wishful thinking) possibility is solar panel manufacture becomes profitable in the US. If that actually happens then the tariffs have to ramp down to force production efficiency to improve leading to economies of scale from the export market.

Flocke Kroes Silver badge

Re: How wind farms kill wales

How often do you see flying wales these days? I cannot remember the last time I saw one. Clearly they have been driven to the brink of extinction by wind farms.

SpaceX set to literally rock Florida with more and bigger Starship launches

Flocke Kroes Silver badge

Re: Engine design philosophy

Raptors will not fit on a Falcon as they are a different shape. So redesign the mount points and they still will not work because the run on methane not RP1. So redesign the pipes and the tanks and you cannot launch because the Falcon launch infrastructure is not set up for methane. Build out new launch infrastructure, do some iterative testing with explosions and crashes into the sea then you get a rocket that will eventually be able to compete with Falcon 9. The goal is to make a step reduction in launch cost. For that you need a bigger rocket to make second stage re-use possible and to divide the per launch costs by more and bigger Starlinks. A Raptor based Falcon is all development cost with no progress. Starships are cheap because they are stacked rings of sheet steel. SpaceX churn out about one per month and there are a bunch of them in the rocket garden. Compare that to New Glenn: large blocks of aluminium machined out to orthogrid and welded together. So far Blue have made one test article for fit checks on the launch pad and nothing suitable for flight.

Raptors are tested on the ground at McGregor. They work fine on the ground. There are still explosions because SpaceX runs some tests to destruction to find the limits and because they are still experimenting to reduce costs, increase production rate, thrust and efficiency. It has been a while since I looked but over 350 Raptors have been manufactured and they complete about one per day. Raptors have been failing in flight because of overall system issues. The tanks have not supplied the required pressure. The pipes to the engines have leaked and caused fires that burn out the engine management computers. For any other launch company Raptors would be fine for stage 1 as no other company requires stage 1 engines to relight for landing. Most rockets need stage 2 engines that can re-light. Again this is a system issue. Starship was in a spin from which it could not recover. Most engines require the propellant to be settled before they can re-light.

Compare Raptor to the more traditionally designed BE-4. "Where are my engines Jeff?" was all over the rocket enthusiast corners of the internet for about a year. Tory Bruno showed remarkable patience - perhaps because Vulcan's upper stage had issues that resulted in an explosion causing delays beyond the late delivery of the first two BE-4s. Last time I checked only one of the second pair survived qualification. Tory's patience this time might be because there are no payloads ready for Vulcan.

And it begins. OpenAI mulls NSFW AI model output

Flocke Kroes Silver badge

AI output is already NSFW

Just ask any lawyer who tried filing AI generated briefs.

Clock is ticking for NASA to fix bucket of issues before next Artemis mission

Flocke Kroes Silver badge

Re-assurance for number of launches

On average, Falcon 9's launch about three times per week. SpaceX are applying their experience with Falcon so that Starship will be nothing like as limited. We can see some of what they are planning from thousands of miles away: land the booster on the launch platform so it is ready to go again promptly and build a rocket garden full of second stages because the time taken to get a second stage back is limited by the slow rotation of the Earth. Next comes the cost of a large number of launches: NASA doesn't care. They pay the same whether it takes 5 or 50 tanker launches. It is SpaceX's problem. Launching an extra Falcon 9 costs SpaceX about $20M. The equivalent for Starship is currently unknown but guesses range from $2M to $10M. Taking the highest cost and multiplying it by a large number of tanker launches (20) and then multiply by the three HLS landings in the contract only gets you to 20% of the contract value. The number of tanker launches does change every few months. So does the lifting capacity of Starship. Falcon 9 payload to LEO has more than doubled since the first version. It is reasonable to expect the same or better from Starship.

Much of what gets Destin worked up on this issue comes from a comparison of cultures. Starship HLS is a firm fixed price contract. SpaceX has the freedom and incentive to find the most cost effective solution. They will do this by experiment and iteration because they need cheap rapid launch for their own business needs. SLS is cost plus. It was designed and simulated to death before being built. Any possible change has to justify the enormous cost Boeing will charge to implement it. At best, SLS will launch once every two years and each launch will cost over $4B. There are multiple huge incentives against experimenting to reduce costs.

While I am here I will mention Destin's other biggest issues. Artemis 3 launch date is currently no earlier than September 2026. It has been delayed before and it will be again. It might look like people are taking that launch date seriously when they pour and stack the solid rocket boosters. Once poured SRB segments have a (theoretically) limited life span. Delays for Artemis I set a new experimental limit. Once stacked SRB's are expected to last a year. When that time was exceeded for Artemis I the SRB's were inspected regularly and found to be still usable. The OIG points out that next time US tax payers might not be so lucky. The entire rocket would need to be de-stacked and the SRB's allocated to the next mission would be brought forward. Boeing would invoice for an extra pair of SRBs, NASA would take the bill to congress who would be waiting with a cheque ready - remember this is a preferred contractor so NASA will not have to piss and whine like they did to get funding for space suits that will work on the Moon.

Next up Destin advocates for the minimum simple solution to achieve the mission. Different groups have different ideas about what the mission is. I think Destin is going for flags and footprint 2. That matches his Apollo 6, Artemis 0 score card. Congresses score card is more like Apollo $25B, SLS $55B so far showing that by their mission criteria SLS is already a winner. For space enthusiasts a more common goal would be to improve technology to the point where a regular shuttle service to a Moon base becomes affordable.

Finally Destin points at the size of the steps between Artemis missions:

1) Orion without life support goes around the Moon and comes back.

2) Orion with a crew goes around the Moon and comes back.

3) Starship launches a propellant depot then N x tanker launches to fill it then an HLS that fills up from the tanker, goes to NRHO and waits for up to three months for an SLS to deliver an Orion with crew. Crew transfers to the HLS, takes a return trip to the Moon then returns to Earth in the Orion.

Destin's talk predates the OIG report on Artemis 1. Artemis 1 had sufficient issues that progressing straight to 2 is now more controversial. He really had an issue with the huge leap from 2 to 3. What the talk does not show is the steps inside the HLS contract. HLS has to get through several milestones (pay days) including an uncrewed landing on the Moon before getting to Artemis 3. Although not in the original contract, SpaceX will also try an uncrewed ascent from the Moon. There is now talk about testing the Orion+Starship docking in LEO. It would be cool if that talk was caused by Destin's video. More likely it comes from a "What if we cannot fix Orion's heat shield soon? We got all this money we have to spend on SLS. How can we get the best value from spending it?"

Flocke Kroes Silver badge

For communism the state owns the means of production. NASA contracts out most of its budget to industry that is not owned by the government. There is a some evidence that this is the opposite of communism: industry owning (or at least renting) the government. I think it more of a self-licking ice cream cone.

Flocke Kroes Silver badge

Re: Like

Commercial resupply services. Commercial Crew, new spacecraft technology research, a bunch of Earth observation satellites, Van Allen Probes, studying the sun, the Moon, Mars, asteroids, Jupiter, Vesta and Ceres, Pluto, some space telescopes and more.

On the other hand it looks like some NASA employees work for Boeing.

Flocke Kroes Silver badge

Re: Broken

NASA are legally required to spend billions on SLS. That means employing a bunch of people to work on it. Other projects attract the most committed to doing something constructive.

NASA does plenty of stuff that is not SLS and does it well.

Flocke Kroes Silver badge

Re: Have to carefully manage the Huawei hate

Huawei hate has not been the driving force behind the last $55B SLS+Orion funding. The key driver was cancellation of the Constellation program. Constellation was a program to send tax payers' money through a preferred set of contractors to re-election campaign funds. There was also a large amount of talk about jobs in every state to build rockets. Constellation was created to fill the gap in campaign financing created by the termination of the space shuttle.

SLS+Orion were doing an excellent job of replacing Constellation until Falcon and Dragon threatened to de-rail the gravy train by actually launching stuff and people cheaply. To keep the funding Boeing had to make good on a little of the talk. This highlighted a key problem with SLS: it cannot send Orion to Low Lunar Orbit. SLS needed a place to go so the Lunar Gateway got funded along with the obsolete booster life extension, Exploration Upper Stage and a new Mobile Launcher to carry them. These contracts ticked two major boxes. They are expensive and long term but miss a key feature - a pretext to justify their existence to tax payers. Artemis Moon landings got funded to supply a range of pretexts that cover multiple market segments: doing some real science instead of just flags and footprints, DEI and beating the Russians Chinese.

If the Chinese land humans of the Moon tomorrow or cancel their entire space program that will make no difference to Artemis because election campaigns still need funding. This can be seen from Artemis funding priorities. Artemis lacked a human landing system and space suits suitable for the Moon. The HLS contract was put up for bids and Blue Origin submitted an ideal proposal: it divided the work between major established contractors all over the US, would blow past the delivery date by several years, needed a big upgrade to do more than flags and footprints and the initial cost was more than double what NASA could afford. NASA caused a shit storm by selecting SpaceX instead. Although SpaceX has a history of delivering late they have an unfortunate reputation for delivering at all and within a fixed budget. Next up Artemis needed suits in a hurry and congress stepped up by fixing the most pressing problem: funding Blue Origin to provide an alternative HLS.

There is a good reason for Starship's size. The next step down in cost to orbit comes from re-using stage 2. During re-entry the place heat is generated moves away from the vehicle in proportion to its radius. Much of the heat just blows straight past Starship without reaching the heat shield. Reducing the size means the heat shield must survive a higher temperature. The other reason for going big has been demonstrated by all the small launch startups. Some costs (like licensing, communications and the control center) do not scale with size. Most of the small launch startups have scrapped their prototypes and switched to a medium sized rocket.

NASA's Psyche hits 25 Mbps from 140 million miles away – enough for Ultra HD Netflix

Flocke Kroes Silver badge

Re: commonly omni-directional

No, at least not the most common in recent years.

Dish antennas are very directional as is the row of bars for your TV antenna. The set of antennas on your WIFI router are used to form an electronicly steerable directional antenna. The same is true for modern computers and phones.

Help! My mouse climbed a wall and now it doesn't work right

Flocke Kroes Silver badge

Re: Never assume, always check

Back when home computers did not come with built in hard disks I came across an unhappy customer. He wanted to know why his newly purchased external hard disk did not make his floppy disks go faster. There was nothing wrong with the guy's intelligence. He had successfully read the instructions, plugged it in and switched it on. The fault was with the instructions that did not explain what a hard disk is.

Meta's value plummets as Zuckerberg admits AI needs more time and money

Flocke Kroes Silver badge

Re: Serious investors ...

... invested in AI chip designers and fabricators.

Musk moves Tesla's goalposts, investors happily move shares higher

Flocke Kroes Silver badge

All that remains is a shareholder vote

The price dropped as people with a clue got out leaving behind only Kool-Aid drinkers. Looks like Tesla will be moving to Texas and Musk will get his $56B pay day.

SpaceX workplace injury rates are rocketing

Flocke Kroes Silver badge

Re: "10 times worse than [space] industry averages"

Please check the (rather sparse) facts about the emerald mine.

There is far more real dirt related to Tesla factories - with evidence, court verdicts and some hope that you might convince some about its CEO who does not already know. Neuralink is an easy target too.

Flocke Kroes Silver badge

Re: No rockets, no accidents.

These were 2023 accident figures. In 2023 SpaceX launched 91 Falcon 9s, 5 Falcon heavies and 2 Starships. That required building 96 Falcon upper stages, recovering and refurbishing 101 Falcon boosters and building at least 141 Merlin engines. Starship manufacturing rate is considerably higher than the launch rate partly because the launch rate is limited by the time required to get a license and partly because they are optimising the factory (which is enormous and being rapidly expanded by SpaceX workers).

Boeing launched 0 SLS rockets in 2023. The planned flight rate is one every two years so a more honest figure would be that half an SLS was constructed in that year. Boeing is the lead contractor. Most of the construction work was done by subcontractors and accidents would show up in the subcontractor's figures. The huge new mobile launch tower required for SLS block 1B is a separate contract - not Boeing at all. Unlike SpaceX's firm fixed price contracts (~$70M/Falcon 9) SLS is cost plus. That requires an army of bean counters to ensure every cent is properly invoiced. I have nothing but respect for the hard work and proffessionalism of Boeing's bean counters who are able to drive up the cost of an SLS launch to over $4B while sustaining hardly any paper cuts.

United Launch Aliance launched 2 Atlas Vs, 1 Delta IV Heavy and 0 Vulcans in 2023. Atlas and Delta contruction had already stopped for the transition to building Vulcans. Presumably a small number Vulcan test articles were constructed in 2023 ready for their first launch the following year. ULA buys in engines so any accidents related to their construction would show up elsewhere.

Rocket Lab launched 8 Electrons. Electron is a small launch vehicle so not as easily comparable to Falcon. It launches mostly from New Zealand which may complicate injury statistics. They recover their boosters from the sea so they would be an interesting point of comparison. SpaceX's booster / fairing recovery division (the most dangerous part of SpaceX) has a lower injury rate than the US fishing industry.

SpaceX is a very unusual rocket company. A more honest comparison would be with the US construction industry. SpaceX and the US construction industry both have room for improvement in their injury rates compared to other first world nations.

Tesla Cybertruck turns into world's most expensive brick after car wash

Flocke Kroes Silver badge

Re: Grease?

I did a quick search for something to show that grease does not prevent corrosion. With a longer search I might have found something specific to stainless steel but that restriction was not a part of the comment I was replying to. I think the real problem came from the assumption that all greases are perfect insulators.

We have ample evidence from Cyber Truck's user manual that stainless can have corrosion problems with grease spots. In any other car company I would expect this information and other issues with stainless to halt work on the concept long before taking deposits. My expectations from a certain group of companies are much lower. I think you can guess what those companies have in common.

Flocke Kroes Silver badge

Re: Grease?

Corrosion under grease spots is a big enough problem to fund scientific papers.

Try a web search of stainless steel corrosion. You will get a long list of results. With careful handling stainless is the right choice for some use cases. Exposed car bodywork is not one of them.

Flocke Kroes Silver badge

Re: Grease?

The mechanism is called differential aeration. Blocking air from getting to part of a piece of metal creates a very low spec battery. The current from that battery drives corrosion in the area protected from the air.

This is not rocket science. It is covered by some A-level chemistry courses. The work-arounds are thoroughly understood by competent car companies. I expect there are plenty of people at Tesla who understand how to delay corrosion for a decade. Something in the corporate culture musk prevent that knowledge from being applied.

A knotty problem: Boffins working on fuel-efficient trajectories for space travel

Flocke Kroes Silver badge

Re: No, yeah, no

The free lunch comes from the middle body of the three. For Earth/Moon/Satelite a tiny change in the Moon's orbital velocity+altitude makes a huge change in the satellite's velocity+altitude because the masses are so different.

Flocke Kroes Silver badge

Re: The important questions

I am not sure about KSP2 but KSP1 physics does not model three body problems. You orbit the body that causes the greatest gravitational field at you current location. When you pass a boundary surface you switch to orbiting a different body. You and MechJeb cannot take advantage of this mathematics without an upgrade to the model.

Some smart meters won't be smart at all once 2/3G networks mothballed

Flocke Kroes Silver badge

Why aren't they talking about the advantages?

About 3 million smart meters cannot currently be remotely hacked and another 7 million will be similarly secured when 2G/3G shuts down.

NASA needs new ideas and tech to get Mars Sample Return mission off the ground

Flocke Kroes Silver badge

Re: Assemble in orbit

Assemble in Earth orbit gets you a big spaceship in Earth orbit. At present there is no tried and tested method of landing a big spaceship on Mars. On top of that, assembly workers on Earth are much cheaper than ISS astronauts.

Flocke Kroes Silver badge

Re: Walk and Chew gum

SpaceX are currently working on landing a Starship on the Moon and landing a different configuration on Earth. Landing on Mars is a mixture of the two: Earth heat shield and flip up, Moon landing legs.

SpaceX's current limit is how fast they can get launch licences.

Flocke Kroes Silver badge

Re: Priorities. Decision time.

China is the scare word. The actual enemy is SpaceX. How is old space going to get the bulk of that $5B-$11B if a Starship lands on Mars before the contracts are signed?

Flocke Kroes Silver badge

Re: primary Don't use anything musk related

Beggers / choosers. Current choices:

*) RocketLab Electron: far too small

*) ULA Vulcan: crippled by slow engine supplier

*) SLS: $4B per launch, every other year at best.

*) Blue Origin New Glenn: no rocket and worse engine trouble than Vulcan

*) All other new space: no operational rockets and medium lift at best

*) SpaceX: Cheapest and most reliable in the world. Dreadful CEO but he is busy with the social media site currently known as Twitter. All decisions will be made by very skilled employees.

Flocke Kroes Silver badge

Re: Skeptical from the start ...

Landing accuracy is currently poor. Something has to get the samples from where the ascent vehicle is wanted to where it actually lands.

Flocke Kroes Silver badge

Re: The other rescopes

One of the early plans was a ~100kg rocket to go from Mars surface to low Mars orbit (LMO) then transfer the samples to a vehicle able to get them back to Earth.

That plan got killed because people decided the transfer in orbit was too difficult.

The revised plan required landing all the kit required for LMO->Earth on Mars with a big rocket to lift it back to LMO.

That plan did not last long because it is too big fo parachutes + hovering sky crane.

Next up, do not land the return propellant but manufacture it out of CO2 and ice.

Another plan is to make some dinky pump fed hypergolic propellant rockets that have better thrust to weight ratio than pressure fed hypergolics.

Every one of these plans requires making kit that has no other purpose than to return a few hundred grams of samples to Earth. It would be like flags and foot prints on the Moon then cancel for 50 years because it is too expensive. I would go for a five year delay on Mars sample return studies, work on making access to space cheaper then look at what can be done with kit available off the shelf at that time. There will be plenty of flag waving opportunities with Artemis, plenty of new capabilities. The interest generated could be enough to fund MSR and get it ready to go before people get thoroughly bored of yet another Moon mission.

Why making pretend people with AGI is a waste of energy

Flocke Kroes Silver badge

Re: Requirements

I think the requirement is: What can collect the most money from investors?

*) Needs to apply to all possible markets. (Eg: a payment transfer system)

*) Must depend on tech hardly anyone can explain to a typical investor. (Eg: blockchain)

*) Must use vast quanities of a product from a big industry to get their support. (Eg: silicon from Intel/nVidia/...)

Small single purpose products just do not tick the major boxes.

Flocke Kroes Silver badge

Two men say they are Jesus. One of them must be wrong.

It takes a human to decide eventually what is right and what is wrong

Humans can tell you what they believe, which isn't always what is true. The test of truth is an experiment.

Linux Foundation is leading fight against fauxpen source

Flocke Kroes Silver badge

Paying who for what

I like the idea of people paying and the money going to the original contributors. I also like the idea of people paying a contributor for a specific enhancement or bug fix especially if the added code is contributed back and available to all.

A more troubling issue is a company changing the license to capture the revenue from previous contributions of others. The company may well be using a part of the revenue to fund bug fixes and improvements but without allowing distribution of software based on the improved source code.

This has sometimes resulted in a fork from the last freely distributable version.

We never agreed to only buy HP ink, say printer owners

Flocke Kroes Silver badge

Re: Instead of hiring lawyers

I did a quick search to see if there were existing projects for open source printer firmware. There is some information available hidden among all the 3d printer results.

The most useful information came from someone who probably did work on printer firmware (on the internet no-one can tell if you are a dog). According to him, the moving the print head / paper stuff is not a big problem when porting to new hardware. There was a timing issue: the SoCs were crap. They did not have enough RAM to hold an entire page. One row of the page is redered while the previous row is printed. Getting that to work at speed was tricky. This was an old comment so SoC performance and memory constraints may not be such an issue today (and many users would not be bothered by some loss of speed).

The biggest issue was the printer command language (PCL) as output from software on the host computer. The language is not simple, comes in various versions, includes many corner cases and the printer has to deal with an enormous list of special cases to print the right thing no matter how badly the printer driver converts to PCL. There is an open source implementation (OpenPCL). It might be easier to just have the driver report itself postscript only and use ghostscript.

Another interesting (but old) post identified the SoC for one printer as a standard off-the-shelf part with the firmware being a Linux based. Although I have low expectations of the effort HP put into non-ink related security I would at least expect them to buy SoCs with the wrong part number printed on top by now.

I found one person had made a reasonable laser printer and published the hardware and software needed to build it. There was also an open ink jet printer but it sounded like it needed more work to get good results. These avoided the costs of understanding proprietary hardware but did not benefit from getting the base hardware at below cost.

The required effort massively exceeds the financial benefit for individual users. The numbers get closer for third party ink manufacturers. I am sure HP will close the gap further by increasing ink prices while widening the gap by making small changes to frustrate open source firmware and replacement controller boards.

Space Force boss warns 'the US will lose' without help from Musk and Bezos

Flocke Kroes Silver badge

Re: Just what we want

I think what Saltzman is after is a way to prevent Russia/China from disabling US satellites. I do not know what that would look like. Making satellites too dark to been seen with a telescope. Watching the trajectory of antisatellite launches and dodging. Threatening to blow up something in response to a US satellite being damaged. Some other concept.

I think what the US government / millitary really needs is better skills at negotiating and completing contracts. Otherwise they will just get ripped off again either by the usual suspects or the new ones.