Hacker News new | past | comments | ask | show | jobs | submit login
NASA just sent a software update to a spacecraft 12B miles away (bgr.com)
555 points by quyleanh on Oct 24, 2023 | hide | past | favorite | 312 comments



This is incredible. Not just because of the engineering that went into such a long lived machine, but also because of the ingenuity of the teams that have looked after this mission and those currently working on it.

Half a century of operation. Bits of data, like gold dust, peppering our radio telescopes with telemetry from the edges of our protective solar shell, and beyond.

Not only is the technical achievement something to be celebrated, but the pushing of boundaries of our understanding is awe inspiring. It carries the symbolism of some of the best parts of humanity’s desire to explore.

This sort of news, at least for me, is an antidote to the darker side of our species. It reinvigorates my hope in what we can achieve when working together.

Thank you to NASA and the scientific community at large.


>Half a century of operation.

The stability of society (specifically American, specifically California/Caltech/JPL/Pasadena) must also be marveled at. That's 2 or 3 generations of engineers and scientists that were trained well enough to actually get things done. This is a marvel of teaching, technical communication, and societal infrastructure. Notably, NASA/JPL has maintained its prestige and funding for that entire time, as has Caltech.

There are many institutions (Sears), and world superpowers (USSR) from 50 years ago that seemed like they would dominate forever, but are gone.


Yes, indeed, when so many projects get started and stopped, this continuity is notable, and we need more of this.

And then, this puts me in absolute awe: https://en.wikipedia.org/wiki/Nalanda_mahavihara

A residential university that ran for ~750 years. What a marvelous achievement of human culture and spirit!

It survived multiple royal dynasties.


This is a really good point, especially for a country that is so young and that has grown so fast. Maybe the instability resulting from that growth is yet to come but it doesn't detract from what NASA and JPL have accomplished so far.


> especially for a country that is so young and that has grown so fast

USA gained independence 247 years ago. That puts it in a minority of countries that have been sovereign as long or longer.


IIRC a good chunk of the voyager team is "first generation"


The thought occurred to me some years ago that a key aspect of governmental institutions is their durability, particularly in the face of adversarial politics. Though the inefficiency of government services is quite often grossly exaggerated, there's much consideration which has to be made of the trade-offs between efficiency and durability.

This should be front-of-mind whenever someone proposes a start-up to address some long-standing social, societal, and/or political issues:

<https://news.ycombinator.com/item?id=12345506>

The other institutional forms which seem to tend toward durability are educational and religious institutions. Which often overlap considerably.


> The thought occurred to me some years ago that a key aspect of governmental institutions is their durability, particularly in the face of adversarial politics. Though the inefficiency of government services is quite often grossly exaggerated, there's much consideration which has to be made of the trade-offs between efficiency and durability.

Governmental institutions are durable even when they are detrimental to society, as long as the government has enough police and military might to prevent insurrections. The society must put up with grossly inefficient institutions simply because the government has a monopoly on some sectors.


Durability, as with virtually any other characteristic or technique, has two edges. However where it is useful to have, it is indeed useful to have.

As to the other substance of your comment, I suspect you would benefit from reading this and its references: <https://news.ycombinator.com/item?id=37366751>


Education on an ongoing project is far easier then rebuilding fresh, even with better science and engineering.

We often see this in infrastructure, like train electrification, tunnel boring and so on. Trams is a good one, so many cities got ride of their tram system, 50 years later they want to build it again and many of them issues.


You're going to be blown away when you discover people have been reliably passing down information to future generations for many many centuries.


"Reliably passing down information" includes a lot of stuff a lot less impressive than "reliably passing down the information required to keep operational a spacecraft mid-flight at the edge of our solar system."


Well the systems controlling Voyager are quite simple, being 1970s technology. Obscure by today's standards perhaps, but not terribly complicated. 70KB of memory, programmed in Fortran and probably some amount of machine or assembly code.

Edit: apparently, due to the post-Apollo budget environment in the 1970s, the Voyager program had to keep costs down so while some new systems were developed, they also reused technology from the Viking program, not even updating or enhancing it.

"The Voyager CCS and Viking CCS would ultimately have the same amount of memory (just under 70kB) despite the routines and programs for Voyager being much more complex. In-flight programming allowed for new routines and programs to be uploaded regularly in non-volatile memory and eliminated the need for large amounts of memory to be required onboard."

https://www.allaboutcircuits.com/news/voyager-mission-annive...


I didn't take the GP seriously because of this. Papyrus and clay and stone carvings do not equal operating a network of brittle, bespoke devices across the scale of a solar system. The existence of language, or writing, or even cathedrals or pyramids, do not eclipse this accomplishment. The odds of losing control of a system is proportional to both time and complexity of the system, and the GP ignores the latter factor.


How are you measuring reliability? I actually would be kind of blown away if this was actually true.


"Reliably" passing down information for future generations is a recent phenomenon. Just go and look at the history of mathematics and there were multiple periods where information was lost and rediscovered.


The mastery of reliably passing down information is likely older than anything you've read.

https://en.m.wikipedia.org/wiki/Vedas

> The Vedas have been orally transmitted since the 2nd millennium BCE with the help of elaborate mnemonic techniques.

That's at least 4000 years of lossless transmission.


We are talking past each other, I would not consider a few religions being able to pass down information about that specific religion to future generations as humanity as a whole being able to pass down information reliably. Especially given that the "Vedas" is wholly useless information while things like the Pythagorean theorem were discovered, lost, then rediscovered. We've only recently rediscovered how Romans created concrete?

Also given humanity has existed for many years longer than 4000 years and 4000 years really only represents 160 generations of humans I don't consider that impressive.


> Pythagorean theorem were discovered, lost, then rediscovered.

And documentation to an average software project is lost in 1 year.

> 4000 years really only represents 160 generations of humans I don't consider that impressive

But 3 generations at Nasa is?


We literally have clay tablets from ancient babylon and books that are thoisands of years old. We have traditional crafts like blacksmiths and pottery that were taught for thousands of years.

Conoaring this to 3 generations ia peak absurdity


This is a great read digging into the team that I loved. I wonder how many of the core team are still there.

https://www.nytimes.com/2017/08/03/magazine/the-loyal-engine...

>Over decades, the crew members who have remained have forgone promotions, the lure of nearby Silicon Valley and, more recently, retirement, to stay with the spacecraft. NASA funding, which peaked during the Apollo program in the 1960s, has dwindled, making it next to impossible to recruit young computer-science majors away from the likes of Google and Facebook.


> NASA funding, which peaked during the Apollo program in the 1960s, has dwindled, making it next to impossible to recruit young computer-science majors away from the likes of Google and Facebook.

This problem can be attacked in a couple of ways, one way is to increase compensation sure, but another is to make it possible to have healthcare and housing on the salary they do offer. The chasm between pay and cost of living is huge and growing.


I've long thought universal healthcare (& UBI) would enable more people to take entrepreneurial risks, but I hadn't considered that passion jobs like NASA could also be beneficiaries.


UBI and universal healthcare would enable _so_ many people to take risks that they can't today. We would see an explosion in the arts, in fundamental science, in passion projects like this. It would be absolutely incredible.

Except according to half of the people in charge it would mean The Wrong People would get money and all they would do is buy drugs. And that's why we can't have nice things.


Is this about the US specifically?

There are many countries with universal healthcare and a decent social safety net (Scandinavia comes to mind).

Why haven’t we seen this explosion coming out of them?


> Why haven’t we seen this explosion coming out of them

You might argue that there's more tech came out of the UK (which still has some sort of healthcare, and had free education until recently) than there should considering its population and economy size. E.g. ARM, HTML, cloned sheep... Also quite a bit of popular music came from the UK. And Harry Potter was written by a person on state benefits.


Nowhere in the world has UBI.

The Scandinavian social safety net is excellent and I think the positive effects of that are apparent. People can take larger risks than they can in the US, but still nowhere near as large as they could with UBI


Also very high taxes to pay for it. Very productive granted, but also taxed at a commensurate rate. Plus Norway now has a bottomless pit of oil money keeping it afloat for the foreseeable future.


> Nowhere in the world has UBI.

Does anywhere have something close that we could see partial "explosion"?

Or is the completely novel?


Go there and look at how they live; that takes inspiration. The benefits of their social welfare are apparent in their lifestyle compared to US daily life.

And the flip: Shouldn't Scandinavia be overrun with drug-addled do-nothings if social welfare is so bad?


I completely agree and am generally supportive of universal healthcare and the welfare state and suspect the Scandinavian model would be an improvement on the current US state of affairs.

I’m just skeptical of the parent poster claim that a Cambrian innovation explosion awaits us all if we would just welfare harder.


And reverse your flip: how do Scandinavians perform when they immigrate to other countries (vs the local population) ?


They spend a lot of money on education, and almost everyone is multilingual – Finns generally speak both Finnish and Swedish natively given they’re the state languages, and many I’ve met have very fluent English and German alongside. So I imagine they do pretty well. Look at Linus Torvalds, for example…

Oh btw, “Scandinavians” only includes countries in the Swedish axis of influence, so to include Finns alongside one has to say “Nordics”. Finland is Nordic without being Scandinavian (technically)


Sweden and Norway have about 50 Nobel Laureates between them out of 15M people.


Capital is concentrated in US, so you have to go there, make some good money and if that doesn’t work — fall back into safety net of your home country.


I was speaking of the US, yes. I should have been more specific.


> UBI and universal healthcare would enable _so_ many people to take risks that they can't today. We would see an explosion in the arts, in fundamental science, in passion projects like this. It would be absolutely incredible.

> Except according to half of the people in charge it would mean The Wrong People would get money ...

If you look at that 'half' as people representing the status quo power, they don't want an explosion in such things that cause change.


This is a tangental subject but I disagree. I think this is one of the many excuses used by people who feel pressure to be entreprenurial but in their hearts are not risk-takers. And that's OK -- the world needs a lot of people who work for others.

If you are a risk taker, especially a young one, lack of UBI or health insurance is not going to stop you. It never has in the past, when those things didn't exist.


I think there’s a substantial number of people who aren’t averse to calculated entrepreneurial risk specifically, where negative outcomes are more predictable/bounded.

These are your types who currently bootstrap themselves and grow slowly rather than seek explosive growth by way of external funding, and I think you’d see both see this group grow and become more bold if basics like housing weren’t something that was ever in question.


Your "especially a young one" is doing a lot of work.

I anecdotally know a number of people, both in and out of tech, who would prefer to be independent contractors or small-business owners - in fields that they know well, and stand better than even chances of ending up with higher total compensation - but have families. Their fear of losing healthcare benefits (temporarily, whilst they get their businesses off the ground; or longer-term, should their businesses fail) keeps them working for others.

The sibling post about "calculated risk" is correct. The conclusion these folks have made (not to strike out on their own, under current circumstances) is likely the correct one, but it's a continual drag on total economic potential. Worse, this is a cost which isn't captured by calculations of what's currently spent on health-care, nor can be off-set against what a more robust social safety net would "cost". It is, nevertheless, considerable.


But at the same time, rents and prices would go up so we’d soon be back where we started unless we made changes to capitalism alongside.


If that were true, we would be seeing far more entrepreneurship from countries with Universal healthcare (most of EU) or UBI (most gulf countries).


EU taxation and regulation plays into this more I guess, outweighing the benefits of healthcare and other social benefits?


If you feel that way about this, I would highly recommend "Carrying the Fire" for you as a tremendous read. Out of all the Apollo-era books I've read, I'd say this one does the best job of mixing in really interesting technical details to really get you to peek at just how much complexity there really was (and may I even suggest giving NASSP a go so you can even try it yourself) while also doing a fantastic job at connecting it to the human and meaning side of the whole endeavor. Just a fantastic read, and sounds like you'd be the perfect target audience for it!


> Not only is the technical achievement something to be celebrated

Only the _documentation_ effort must be monumental.


Lets hope the _archival_ effort can match or in 50 years time we’ll have another one of these:

https://en.m.wikipedia.org/wiki/Apollo_11_missing_tapes

https://en.m.wikipedia.org/wiki/Stolen_and_missing_Moon_rock...


Even with those problems, those are still far better than the (complete lack of) documentation I see in software workplaces these days.


In fairness to the workplaces that I've been in, documentation for a software stack, or even the product, from 18 months ago may as well be in Swahili for as much good as it would do any new member to the team

Just like the comparison to code quality I often see is any such NASA thread: it's a whole different ballgame when one is sending a spacecraft out, or carrying lives to space, than "I tried to send a Slack message and it did a sowweee,oopsiez"


The missing tapes are definitely one of the saddest engineering stories for me. Such an important moment in history and we just threw it away.


Yes, and JPL makes it available.

Voyager Telecom: https://descanso.jpl.nasa.gov/DPSummary/Descanso4--Voyager_n...

A lot more at the top-level site and the links on the right-hand side: https://descanso.jpl.nasa.gov/


I've been trying to "learn" the missions more thoroughly by flying them on Orbiter and NASSP, and just the sheer amount and depth of documentation that is accessible today is truly astounding. There's something awe-inspiring about sitting and my computer and looking at the mission plan, only to then not understand the "why" of this particular switch that I'm throwing, to then open a pdf with a "astronaut training booklet" which explains the overview of the thing I'm using, to then open another pdf with the detail technical documentation of it, to then a schematic of the equipment should I at any point (but prob not) ever try building one myself.


The care that must go into writing and testing software that goes out to a machine like this is incredible. I write software that, if and when it fails, it fails hard and it just comes back with no added cost to me while I go ahead and look at what went wrong and release a fix within the hour.

Not having that failure feedback loop is daunting. I really respect that kind of work.


Also the fact that this CPU is still running at all, after such long time in such hostile environment (radiation, temperatures)!


These spacecraft don't have CPUs, at least not the way we think of them, in the sense of a single microprocessor IC. Their "CPUs" are built from discrete logic gates on CMOS and TTL ICs, the way the earliest digital arcade games were.

https://en.wikipedia.org/wiki/Voyager_program#Computers_and_...


The concept of CPU predates micro-processors. They were cabinet sized originally.


As Oscar Wilde wrote, "We are all in the gutter, but some of us are looking at the stars".


Shared this with my son who really enjoyed reading that others share his optimism about our species. Very well said, thank you. You've impacted the outlook of a young man trying to figure out his place in the universe, who often times struggles with noticing the rare opportunity it is to be gold dust peppering a radio telescope.


"antidote to the darker side of our species." .. LOVE THIS!


This for summing up what I couldn't find the words for!


Before:

In 1994 JPL started working on the Remote Agent (RA), an autonomous spacecraft control system. RA was written entirely in Common Lisp despite unrelenting political pressure to move to C++. At one point an attempt was made to port one part of the system (the planner) to C++. This attempt had to be abandoned after a year. Based on this experience I think it's safe to say that if not for Lisp the Remote Agent would have failed.

We used four different Common Lisps in the course of the Remote Agent project: MCL, Allegro, Harlequin, and CLisp. These ran in various combinations on three different operating systems: MacOS, SunOS, and vxWorks. Harlequin was the Lisp that eventually flew on the spacecraft. Most of the ground development was done in MCL and Allegro. (CLisp was also ported to vxWorks, and probably would have been the flight Lisp but for the fact that it lacked threads.) We moved code effortlessly back and forth among these systems.

The Remote Agent software, running on a custom port of Harlequin Common Lisp, flew aboard Deep Space 1 (DS1), the first mission of NASA's New Millennium program. Remote Agent controlled DS1 for two days in May of 1999. During that time we were able to debug and fix a race condition that had not shown up during ground testing. (Debugging a program running on a $100M piece of hardware that is 100 million miles away is an interesting experience. Having a read-eval-print loop running on the spacecraft proved invaluable in finding and fixing the problem. The story of the Remote Agent bug is an interesting one in and of itself.)

The Remote Agent was subsequently named "NASA Software of the Year".

https://flownet.com/gat/jpl-lisp.html


There's some pretty depressing stories in the second half of that article. But more interestingly, the author articulates beautifully one of the main reasons I chose pure mathematics over a programmer career in the late 90s:

"One of the reasons I stayed at JPL for twelve years was that I was appalled at what the software industry had become. The management world has tried to develop software engineering processes that allow people to be plugged into them like interchangeable components. The "interface specification" for these "components" usually involves a list of tools in which an engineer has received "training." (I really detest the use of the word "training" in relation to professional activities. Training is what you do to dogs. What you should be doing with people is educating them, not training them. There is a big, big difference.)"


We just "solve" the problem by training tens of thousands of dogs. Now we have a gigantic dog circus running 24x7. What's mysterious is no matter how many shocks and surprises ever growing complexity produces it doesn't collapse.


What is the most Lisp-like type safe language?


Shen has type checking that you can turn on and off. When it's on, it's very much like an ML such as Standard ML.

https://shenlanguage.org/

This probably isn't the "most Lisp like," though, even though it does have s-expressions and cons cells. The other person who suggested Coalton is probably right, since AFAIK it's just Common Lisp with static types.


Carp is a statically-typed Lisp: https://github.com/carp-lang/Carp


Coalton which runs on SBCL


Typed Racket probably.


To put the 12B miles in an easier grasp, Pluto's average distance is 3.7B miles. So Voyager 2 is a hair under 3.3 Plutos away, and Voyager 1 at 15B miles is slightly over 4 Plutos away.

Another point of reference is the heliosphere, at 100AU. 1AU (earth to Sun) is 93M miles, so the heliosphere is around 2.5 Plutos away.


On some other YouTube channel, I recently saw a diagram where the width was 0.01 light-year. That’s “only” about 59 billion miles; surely we can’t be too far from the moment we can conceive of Voyager 1’s distance in light-years.

Nope. At its current speed, we’ve got another 129 years. That’s an underestimate because the spacecraft will continue to decelerate a bit as it pulls farther away from the sun.

That’s about 60,000 years to the nearest star.


>That’s about 60,000 years to the nearest star.

Assuming it doesn't hit a rock or eventually fall into something. How long could it physically last with cosmic rays etc hitting it? will it eventually just an unrecognizable metal lump?


> How long could it physically last with cosmic rays etc hitting it?

Likely a long, long time.

Neanderthal cave paintings have lasted upwards of 60,000 years.


Surely there's a difference between cave paintings on a planet with an atmosphere, and a spacecraft travelling void?


Yes, the former has many more destructive forces.

So I expect Voyager to be more or less structurally similar in 60,000 years


the one they updated is apparently 18.40 light hours away. So it will be quite some time before we're talking about light years.


I thought I read more recently that they're no longer sure about where the heliosphere ends, because of the data returned from the V'ger probes.


Voyager 2 is over 50,000 average Earth/Moon distances away, Voyager 1 is coming up on 63,000.

I don't know if that makes it any easier to grasp, but I have a hard enough time relating to Pluto's distance, let alone the Voyagers.


50k Earth-moon distance may be easier to grasp for some. For me, I'm not sure how to understand the earth/moon distance enough to make 50k earth-moon a more meaningful measurement. Pluto is in the Kuiper belt, which for many, signifies the physical bodied edge of our solar system. So for me, scaling that to 2.5 to know where the heliosphere is and 3.3 & 4 to where the Voyagers are is more understandable.


Muricans will use anything but metric.


Voyager 1 will be 4 Plutos away whether we use metric or not.

Putting large numbers in context helps understand them. "Muricans" know it, and so do Europeans at the European Space Agency. Here's what they write about the International Space Station:

> At 28 800 km/h it only takes 92 minutes for the weightless laboratory to make a complete circuit of Earth. Astronauts working and living on the Station experience 16 sunrises and sunsets each day.

They could have left it at 28 800 km/h - which is metric - and call it a day. But they went an extra mile (or should I say an extra 1.6 kilometer) to put that number into perspective.


Until someone explains it to me in Olympic swimming pool volumes, I'm clueless.


I'll also accept football fields.


I don't know how to measure things in anything but capybaras.

https://www.jpost.com/science/space/article-761322#asteroid-...


Metric doesn't let people grasp large numbers.


4 Plutos sounds pretty close and I doubt most grasp the distance of Pluto


Only the US, Liberia, and Myanmar have a complete grasp of large numbers like, for example, fifty-seven.


Damn it, internet and rage bait culture. "Americans will use anything but metric" in refernce to using pluto distance as a metric is what I replied to, not metric vs mi.

Freaking HN.


19.312.128.000 km is a pretty big number, people can grasp that.


Metric is neither better or worse than anything else, in that regard.

Describing a distance as "4 Plutos" isn't that helpful when Pluto is too dim to see with the naked eye, and still only a dot with a good ground-based telescope.

Even smaller distances like "AU" are too big for a human mind, with the Earth itself being less than a pixel across if the orbits were drawn to scale on an 8K monitor.


> still only a dot with a good ground-based telescope.

That gives a pretty good feel for how far away it is!


What? No it doesn’t, you need to know how big it is for that, and most people don’t have a sense of scale that encompasses either the full size of Earth (how far have you tried hiking, even over multiple days? If it was the Pacific Crest Trail congrats but you’re unusual) nor the size range of planets.

Most stars (and galaxies) can be described with the same words I just used for Pluto, but stars and galaxies are thousands to trillions of times further away than Pluto.


It's not at all quantitative, obviously, but I think knowing that

- Pluto's the size of the moon, more or less. (Or even assuming it's ~Earth-sized), but... - ... can barely be seen from Earth, even with fancy equipment, and - ...The probe is 4x farther away than that.

Gives good "distant" vibes in a way that large numbers don't.


And yet literally the entire world aside from United States, Liberia, and Myanmar is able to grasp it perfectly.

So such a statement would either insinuate that Americans are genetically inferior somehow, or you're just culturally attuned to an outdated British system.


What are you talking about? Look at the context before you get outraged. The comparison to metric here is "plutos" kind of like measuring with bananas. Saying a million km is not the same as saying "one moon distance".


So genetically inferior that they are the only ones with a craft that is 4 Plutos away :)


Could have been 5 plutos if they used metric.


NASA uses metric


NASA are still Americans. I'm not arguing about metric or imperial being better.


Parent claims genetic inferiority for not grasping metric, NASA does, so they are excluded in the first place.

Same with "military" time.


NASA didn’t begin to standardize on metric until 1990. Far after Voyager was developed and launched.


NASA already used metric internally for the moon landings

https://ukma.org.uk/why-metric/myths/metric-internationally/...

And I doubt that calculations for the voyager mission where done in imperial. The output and for contractors yes, but internally it's metric.


How many bananas is that? (/s)


A more relevant measure (for software updates etc.) would be light-years or light-seconds etc.


What got me intrigued was how did they send the update and what was the size of the update. Sadly no mention of those aspects in the article. Is there any way to find out ?

I was feeling so low today that I just switched to HN to tune things off of my mind, saw this headline and right off the bat, am in a better mood.

What an incredible feat of engineering !



If you want to learn about how the Voyager flight data system is designed, read Nasa's historical document Chapter 6: Distributed Computing On Board Voyager and Galileo [1].

The whole read is gripping, but the most interesting part for me is the section "Developing Voyager's Flight Data System Computer " describes the choice to use CMOS memory and DMA from instruments.

[1] https://history.nasa.gov/computers/Ch6-2.html


It was a different world in 1975:

  After determining requirements, Wooddell examined possible hardware and software tradeoffs. In an insightful memorandum, John Morecroft explained the concept of "soft logic" as a complement to the "hard logic" in the Flight Data System44. Writing in 1975, when the actual flight software began to be prepared, Morecroft pointed out that the program for the computer was actually a soft representation of hard-wired circuits. Conceptually, the memo stands as an explanation of the essential meaning of firmware in general.


Damn, that's cool. Do you have a link to the whole document?


To the memorandum? No! That would be fun to track down.


Looks like this, although a quick search doesn’t find it:

43. J. Wooddell, "Design of a CMOS Processor," pp. 2-3, files of J. Wooddell. 44)J. Morecroft to K. Frewing, "FDS Programming," January 14, 1975, files of R.J. Rice, Jet Propulsion Lab.


Contents [0]

Part I : Manned Spacecraft Computers [1]

Part II: Computers On Board Unmanned Spacecraft [2]

Part III: Ground Based Computers for Space Flight Operations [3]

[0] https://history.nasa.gov/computers/contents.html

[1] https://history.nasa.gov/computers/Part1.html

[2] https://history.nasa.gov/computers/Part2.html

[3] https://history.nasa.gov/computers/Part3.html


I just realized, there's an "index page" linked at the bottom of your OP's link. I didn't notice it initially.

https://history.nasa.gov/computers/contents.html


I hope this update doesn't come with ads and data collection to "improve my experience"


VoyagerOS Updated yesterday 529B

Bug fixes and improvements, watered the plants, took out the trash.

Keep VoyagerOS up to date to always have the best experience!


Depressingly accurate. How I hate these useless "changes" paragraphs accompanying YouTube and the daily Netflix app update on iOS and Android (why would a video player need so many updates to do exactly the same thing for years: show a selection, permit selection of a video, scrub, pause?!)


Sometimes I wonder if the Win11 start menu search is running on something 12B miles away. Lol.


Feels like a webpage from the days of AOL with blinking images and marquees.


Evolution of long-delay error correction in Voyager also evolved a bit, given their 32KB firmware size constraint.

NASA went down to k=7 (from k=9) for convoluted coding.

Not only were Golay and Reed-Solomon ("a special subset of [Bose-Chadhuri-Hocquehem] codes" which are "a powerful class of Hamming code", Brown, Elements of Spacecraft Design) used, convolutional coding was also used.

Sources:

- Figure 3 - https://i.stack.imgur.com/z5F1p.png

- paper -https://repository.arizona.edu/handle/10150/615317


From Wikipedia:

> There are three different computer types on the Voyager spacecraft, two of each kind, sometimes used for redundancy. They are proprietary, custom-built computers built from CMOS and TTL medium-scale CMOS integrated circuits and discrete components, mostly from the 7400 series of Texas Instruments.[34] Total number of words among the six computers is about 32K.

32K memory to manage such a complex device for 46 years .. Nowadays the size of an OS is measured in GB's.


> 32K memory to manage such a complex device for 46 years

Word size was 18 bits, so 32k words is an enormous 72 kilobytes.

Google also gives me 69.63 kilobytes, for example in https://www.reddit.com/r/technology/comments/15iaox2/nasa_ha...

https://history.nasa.gov/computers/Ch6-2.html says the CCS had 8k words (“Along with the new chips, the memory changed with an expansion to 8K”)

I guess the confusion is because the hardware has multiple (¿3?) computers.


I don’t understand why we are not sending these off in every direction constantly. The price can’t possibly be that prohibitive considering what we can accomplish with 40 year old technology.


"Every direction" isn't what you think it is.

Even if we had a voyager travelling the "opposite" direction, once outside the solar system they'd appear to be travelling at almost the same speed with respect to the orbit of the solar system around the milky way.

The solar system is estimated to orbit ~200km/sec around the centre of the milky way.

Even with a perfect ejection, at the estimated speed of Voyager you'd end up with one probe going ~180km and the other ~220km, which isn't a huge difference.

And that's without the fact that getting Voyager to that speed required some unique planetary alignment.


On a related note, starting from a circular orbit, it takes

1/(2sqrt(2) - 3) ~ 5.83

times more energy to throw an object to the center of a gravity well than it does to throw it out of orbit.

It doesn’t matter how close your initial orbit it, or whether you’re orbiting around the Sun or Sagittarius A*. The ratio is always the same, though gravitational assistance from other bodies often affects the calculations in practice.

I expect we’d send a satellite outside the Milky Way long before we get one anywhere near the center of the galaxy. Not that either mission seems particularly feasible with current technology.


Doing a % comparison makes little sense. Everything in this arm of the Milky Way is moving along with us.

The key metric is that their locations would spread by ~40km/sec. Whether that is an interesting amount of distance in an interesting amount of time is debatable, but that's the number.


You can take a look at the New Horizons to get an idea of what a moderately capable trans-Neptunian probe costs today (~$700M). If you made several, economies of scale would of course lower marginal costs somewhat. But the money isn’t even the main bottleneck (although it absolutely is a huge one) – power is. Specifically, the RTG radioisotope generators that are our only reasonable way of powering things at distances beyond Jupiter. Synthesis of radionuclides (eg. Pu-238) suitable for RTG use was originally a nice side effect of nuclear weapons production, but these days availability is very low even though the DOE recently restarted production after a long pause.

In the 21st century NASA has launched exactly three RTG-powered missions: New Horizons (2006), Mars Science Laboratory/Curiosity (2012), and Mars 2020/Perseverance. The next will be the (very awesome) Dragonfly mission to Titan, planned to launch in 2027. So given a cadence of roughly seven years, and one out of four missions targeting trans-Neptunian space, you could expect to launch an interstellar probe every 30 years or so, barring significant advances in power technology. The next one will most likely be a sorely-needed Uranus or Neptune flyby mission that would possibly launch in the mid to late 2030s.

There are maybe a dozen reasonable mission proposals submitted for every mission that actually ends up at the launch pad. The competition is fierce, and any proposal has to argue its return on investment incredibly convincingly to get selected for implementation. We now know many interesting Kuiper Belt objects that would definitely deserve a closer look – but we also know many other things that are much more accessible and would also deserve a closer look!

No amount of extra money would realistically allow NASA (or any space agency) to fly missions simply "because we might discover something" because there are always many, many other mission proposals in the pipeline with vastly better justification than merely "because we might discover something".


Because there are windows to visit certain planets, especially with modern rockets. You can't just point them in the direction of a planet and yeet them into space; there is physics involved.


We did yeet New Horizons at Pluto. It got an assist from Jupiter, but could have reached Pluto without, six years delayed.


What about electrical propulsion methods?


The only reason why we got these probes as far and as fast as we did was because we were able to fly by each planet and use their gravitational influence to skip to the next one, increasing velocity after each slingshot.


> use their gravitational influence to skip to the next one

You are correct in that the Voyagers' launch window was unique.

Nit pick: gravity assist doesn't fundamentally rely on gravity. It extracts energy from planets' motion about the sun [1]. If we could somehow "bounce" a spacecraft off a planet's surface, it would impart momentum similar to that of a gravity assist. (In my opinion, it's a badly-named term.)

[1] https://science.nasa.gov/learn/basics-of-space-flight/primer...


> If we could somehow "bounce" a spacecraft off a planet's surface, it would impart momentum similar to that of a gravity assist.

When my kids were little, I showed them by holding a basketball in one hand and a tennis ball in the other, with the tennis ball on top of and touching the basketball. When I dropped them simultaneously, they fell together and bounced off the driveway — and the tennis ball took off like a rocket (no pun intended).


> gravity assist doesn't fundamentally rely on gravity

I'm not sure how you can make that claim. If you were to bounce off the surface, that would likely not be considered a "gravity assist", no?


Good point, it’s a bit like calling hitting a baseball with a bat an “electric assist”. It’s not wrong, it’s just not very helpful.

It’s also a good illustration of why we call some forces strong and others weak.


I used to think this until I learned how our solar system moves around the galactic core. Truth is, space is wild. We're like the equivalent of stone age people standing on the shore of the atlantic looking out wondering if we can make it out there. Except the ground beneath us is also moving through the ocean at fantastic speeds, causing huge waves in our wake.

This is a huge oversimplification by a layperson, but just as ancient seafarers had to escape the currents caused near their shore, we have to figure out how to escape our solar system first, into interstellar space. You're not safe there, but at least you've exited the wake of our sun. Just like we had to exit the graviational pull of earth to reach the moon.


The Voyagers took advantage of a particularly good planetary alignment for gravity boosts.


The cost is the fuel. It takes an insane amount of energy to leave Sol. The voyager probes were able to save 30km/s worth of Delta-V by performing a ton of gravity assists during the "Grand Tour" launch window planetary alignment that occurs once every 175 years.

For context, getting something to low earth orbit is about 9.4km/s of Delta-V. From LEO to mars is another 4.3 km/s of Delta-V. Getting from LEO to any solar escape trajectory adds is about 18km/s.

As you can see, 30km/s of fuel savings is extremely significant, and that required the 175 year Grand Tour alignment.


The equivalent of about 4 billion dollars in modern day money for both of them. That, plus waiting for some very opportune launch times so that we can gravity boost off of everything.


Those gravitational boosts are vital.

The closest in speed so far would be 'New Horizons'. But even then NH is going at 14.5KM/sec compared with Voyagers 18KM/sec. There are no probes currently that would catch up to Voyager.


18km/s is almost understated in terms of speed, so I looked it up to more familiar measures of speed: 65,000kph or 40,000mph. The only thing I really learned here was how even in the thousands of miles or km per hour, it’s just inconceivable how fast that is. Km per second was a better measure after all.


> it’s just inconceivable how fast that is

And even more inconceivable is how slow it still is in terms of human spaceflight outside of the Solar System.


What is crazy to me is that even at those speeds, it is over 18,000 years from reaching one light year distance. The closest star is 4.2 light years away. So to ever reach it we would need multi generational ships or some way to go much much faster.


Round the world in 40 minutes

Travelling from New York to LA quicker than playing the song "From New York To LA"


But the more copies you make of something, the cheaper they become. The R&D cost is the same and is only paid once whether you're building 2 space probes or 2000.


The probe itself is a relatively insignificant cost. The launch costs and timing windows are the limiting factor.

The tyranny of the rocket equation is the fact that to carry more fuel to orbit, you need more fuel to get it into orbit, which itself needs more fuel. That's why gravity assists are so essential for escaping the solar system on a budget.

The planetary alignment that enabled the Voyager launch windows only occurs once every 175 years.

We should definitely be sending out more probes, but we can't just do it whenever we want, at least not without a lot more expense.


The specific alignment for the Voyager missions was to permit flybyes of all the gas giants: Jupiter, Saturn, Uranus, and Neptune, with only two spacecraft. Both V1 & V2 visited Jupiter and Saturn, with Voyager 2 continuing on to both Uranus and Neptune. All visits were flybyes, that is, the probes were within the vicinity of the target planets, but did not orbit them for any prolonged length of time.

More conservative trajectories giving fewer flybyes but relying on more probes would be possible, and are in fact what we've accomplished with the Gallileo & Juno (Jupiter), Cassini-Huygens (Saturn), and New Horizons (Pluto) spacecraft. There are NASA (U.S.) and Chinese proposals for a return visit to Uranus and Neptune, though those remain in planning stages.

<https://www.space.com/nasa-uranus-orbiter-and-probe-mission-...>

The orbiter missions had a much longer time-on-station. For Gallileo, roughly 7 years 9 months, Cassini-Huygens 13 years, and Juno 7 years to date, with another 5 or more possible.

The Ulysses solar probe also flew past Jupiter, using the planet for an orbital assist to both slow it down to approach the Sun more closely and incline its orbit above and below the planetary plane so that the Sun's poles could be observed directly.


> But the more copies you make of something, the cheaper they become. The R&D cost is the same and is only paid once whether you're building 2 space probes or 2000.

The problem with that idea is, except for a few geeks, no one wants to spend that kind of money on space probes.


The probes are handcrafted from the unobtanium by the very expensive engineers, and designed individually due to this.

We need a heavy lift platform flying for a few years before we could start making much more crude larger probes, with cheaper or even massmarket components. Then it may work.


Because the inverse square law applies to probes too. We can send them out by the millions, but by the time they get far enough out to see something interesting the probe density will be practically zero.



How do either of these still have any propellent left? If they were originally scheduled to last only four years, how did they manage to get launched with enough propellent to last decades and billions of miles past their original mission?

According to this article from 2003, the probes have enough hydrazine to last until 2034 and 2040:

https://slate.com/news-and-politics/2003/11/what-voyager-1-u...

Edit: this "Voyager Backgrounder" paper from 1980 is terrific:

https://ntrs.nasa.gov/api/citations/19810001583/downloads/19...

The probes were launched with 104 kg (230 lb) of hydrazine. But I'm still not clear why they were so obviously overbuilt for a four year mission.


They usually have the mindset of "let's put enough propellant so that even if our estimates are orders of magnitude off the mission will live long enough", and normally their estimates are correct so they end up with tons of fuel


I believe the mission length is just how long it takes to complete the mission goals. It isn't necessarily an estimation of the spacecraft's lifespan.


Possibly, they underpromised and overdelivered.


It takes around 22+ hours for 1 way communication to Voyager if I recall correctly. It takes 8 min for light to reach earth from the Sun. The closest star is 4.24 light years away. These kinds of distances are hard wrap your mind around.


Not distance related, but here's another terrifyingly fun fact:

The sun accounts for ~99.9% of total mass of the solar system.


“It’s called space for a reason.”

We live on what barely qualifies as a dust particle, orbited by a much smaller dust particle. The next dust particle over, Mars, is far enough to require something like 12 minutes for light to make a round trip.


* 99.8%

Don't underestimate yourself!


Is this considered OTA given space is a vacuum?

Also, I've had some scary deployments in my time but this would take the biscuit!


Well, "over the aether" works ;-)


According to the sibling link to the Deep Space Network <https://news.ycombinator.com/item?id=37995866>, the updates did in fact travel through our air first :-D


These two probes are mind blowing achievements. It would be lovely to read about the code and the various things it does, along with a description of the computing hardware (which machine code or assembly language or other language was used), etc.

> The update, which took almost 18 hours to complete, was transmitted to help Voyager 2 avoid the same problem that its sibling, Voyager 1, experienced last year.

> …

> If there aren’t any issues, it will trigger the Voyager 2 update on October 28.

Correct me if I’m wrong. The wording is a bit confusing. I presume the first paragraph above, when it says “to complete”, refers to just transmitting the update (and acknowledgment that it has been saved). The patching with this updated code will be triggered on October 28 after some verification checks are done.


Were unit tests a thing back when the probes were built?

I wonder how new engineers learn about the intricacies of such ancient tech, without breaking them.


Sadly the "megabytes of JavaScript to 'hello world'" culture of today could not even begin to comprehend how this even works


It funny, usually people on HN diss NASA et all hard for being those big, inept governkent agencies, with SpaceX being so so mich better because SpaceX get SW"engineering". And then NASA does something like this, and all of sudden it so cool! Coupd it be that those orgs aren't even nearly as incompetent as people have a tendency to think?


> It funny, usually people on HN diss NASA et all hard for being those big, inept governkent agencies, with SpaceX being so so mich better because

Many of NASA’s problems aren’t the fault of NASA, they are the fault of Congress. NASA didn’t come up with SLS+Orion all by themselves, the high-level design was dictated to them by Congress and NASA was left with the job of fleshing out the details and making it all work somehow. SpaceX needs NASA as a customer and a source of valuable expertise, it just wants it freed from being forced by Congress to spend billions on super-inefficient projects such as SLS+Orion, when SpaceX has a solution which can do more for less (Starship+SuperHeavy) and is likely to be available soon (bureaucratic delays due to the FAA and environmental agencies such as the FWS being under-resourced being the current biggest obstacle to that)


So SpaceX wants NASA to be able to spend all those billions with them instead of a competitor? Shroud business move, that says nothing about capabilities.

Just one question, what will be avaiable first, FalconHeavy and a fully certified Starship (which has to launch without blowing up for that), Tesla FSD or Tesla's humanoid robot?

Regarding SLS, they launched successfully last year and aim for a manned launch in 2024. SpaceX has to beat that timeline, assuming this whole thing being a race, which it isn't.


Falcon Heavy has been launching commercial, government, and military payloads successfully for years now. [1] I'd strongly encourage checking out their first successful test launch of it, from 5 years ago. [2] Skip to about 29 minutes in for the money shot. Still gives me goose bumps!

So let's compare the two! Falcon Heavy was developed with completely private funding, at a cost of ~$500 million. The prices of the launches are listed in Wiki. They start at $117 million, as NASA paid in 2023 to launch their Psyche craft. The SLS has been in development for more than a decade now, fully paid by the taxpayer, at a cost just about to break the $30 billion mark. The nominal cost per launch is still unknown but will almost certainly top a billion dollars. The last NASA administration gave an estimate of $800-$900 million per launch in 2019 dollars. And SLS cost estimates only go one way.

And what do you get for $30 billion, and then another billion per launch? Ideally up to almost twice the capacity of a Falcon Heavy. The reason people are so cool on the SLS is that it's already a very poor value against existing technologies. And will be completely obsolete by the time Starship comes to pass which should not only be able to launch for a fraction of the cost of Falcon Heavy (which is already a fraction of the cost of SLS), but also launch well over the ideal launch capacity of SLS.

[1] - https://en.wikipedia.org/wiki/Falcon_Heavy#Launches_and_payl...

[2] - https://www.youtube.com/watch?v=wbSwFU6tY1c


We do not have any reliable financials from SpaceX, just advertized launch costs and PR releases. SLS on the other hand has, as does e.g. Ariane-6, public budgets and a lot more bureaucracy.

For Starship to make things like SLS obsolete, it has to be available first. Until then, it is just pure speculation. Is SLS more expensive than Starship development wise? Yes. But it is not the goal of a government to be profitable, nor is it to hand technologocal monopolies to eratic billionaires. Supporting alternatives is always good, and it maintains know-how and technology at different companies, which is also a good thing.


Starship's a red herring. The SLS is already so far behind the game (before even being finished) that it lacks any rational argument even against the Falcon Heavy, which has been in service for years. And nobody's being handed monopolies. Government space contracts have a competitive bidding process. If Boeing/Lockheed, Blue Origin, or any of the other less known aerospace startups can come up with something that can reasonably compete against half a decade old SpaceX tech then I'd love to see it as much as anybody else. But for now that's simply not happening, and just corruptly tossing tens of billions of dollars towards Boeing/Lockheed isn't bringing us any closer to that happening.


Assuming Starship succeeds, hopefully at that point other players will adopt the basic design, iterate and improve, then forcefully compete. I'm still surprised that, all these years after Falcon 9 obviously works, there isn't a "clone" on the market. (I am aware there are couple of small "new space" companies working on something similar.)


Agreed. I think you can even take it further than other new small space companies. Boeing and China would both absolutely love to replicate SpaceX's tech, and both have (relative to SpaceX) practically unlimited resources to do so, yet both are still nowhere near them yet, and not for lack of trying.

From an outsider's perspective it's quite difficult to understand what the big barrier is.


> But it is not the goal of a government to be profitable, nor is it to hand technologocal monopolies to eratic billionaires.

So instead it should award contracts - without any competition even - to big established aerospace firms such as Boeing, Lockheed-Martin and Aerojet? SpaceX, by contrast, won its most important NASA contracts (ISS cargo and crew, HLS) by open competition.

> Supporting alternatives is always good, and it maintains know-how and technology at different companies, which is also a good thing.

If NASA wanted to do that, they would have spread the SLS/Orion money around newer and more innovative players (not just SpaceX-Blue Origin, Rocket Lab, Axiom, Relativity, etc), rather than giving billions to Boeing/Lockheed/Aerojet/etc to keep alive technologies from the 1970s. Of course, Congress wouldn’t let NASA do that


> We do not have any reliable financials from SpaceX, just advertized launch costs and PR releases. SLS on the other hand has, as does e.g. Ariane-6, public budgets and a lot more bureaucracy.

We know how much SLS and Ariane 6 have spent, but that's not any kind of reliable indicator of how much they're going to cost.

> For Starship to make things like SLS obsolete, it has to be available first. Until then, it is just pure speculation.

Only if SLS is ever going to be available. If Starship is further along in development, that's pretty suggestive.


> Just one question, what will be avaiable first, FalconHeavy and a fully certified Starship (which has to launch without blowing up for that), Tesla FSD or Tesla's humanoid robot?

In order: Falcon Heavy (because it's been around for years already), then I recon Starship, the robot, then FSD.

But this is on the pair of assumptions:

(1) that the robot will initially launch with a very limited role and set of environments it can work safely in — a lightweight mobile pick-and-place robot is much much easier than a fully general humanoid that you can drop into any role a real human would do and expect it to not accidentally dismember or decapitate itself or others.

(2) it only counts as "FSD" when they can ship the cars without including a steering wheel. If the current marketing of a nice-but-limited cruise control, lane changer, and occasional automatic break counts as "FSD", then it too is already out. (I would not count this myself).


Shrewd business move?


Personally, I'd tend to give most credit for something like this to the era that created it. I mean building something that can be reasonably expected to operate and maintained in some of the most hostile conditions known to man, for half a century? It's just borderline magical. And NASA of that era did everything just perfectly, nailing all the gravity assists (obviously using computational/telemetry tech of the time), and more.

And that applies to everything from that era. We hadn't even put a man in orbit in 1962. 7 years later, we would be walking on the Moon!!! The degree of competence and capability of that era of NASA is something that I think is probably unmatched in the entire history of our species.


Agree with you here.


Most of the discontent I see aimed in NASA’s direction has little to do with their autonomous missions (which have been consistently excellent) and more to do with how Congress has misused NASA’s crewed efforts as yet another cash pipeline for traditional aerospace companies, which has for decades prevented crewed spaceflight from becoming progressively cheaper, more sustainable, and more common, instead keeping it artificially suspended in its mid-late 70s state.

I think it’s a massive error for Congress to have anything to do with any remotely technical decision. That should be at the sole discretion of NASA. Congress’ involvement should go no further than deciding what slice of the taxpayer funded pie they get yearly, and should not include the ability to attach strings of any kind to the funding.


>It funny, usually people on HN [insert tribal "sports team" logic], and then all of sudden [actual technology discussion]

I sure hope this doesn't come to be considered 'usual' on HN.

If so, it just becomes a race to the bottom.


It definitely doesn't, maybe there are some jerks that think badly of NASA, but I doubt they make the majority of us. They do a lot of engineering and science, but have been underfunded for so many years, with the government siding with private companies.

For many, like me. It would be a dream to work at NASA, whatever the salary or bureaucracies it might involve.

It was my childhood dream, but given I wasn't American, that would never happen.


Depending on where you live, ESA would be an alternative. The quite constantly look for people at their Oberpfaffenhofen site near Munich, if memory serves well they do some Galileo related work there.


I'll keep my eye on it. ESA is awesome as well.


the delay of James Webb 10+ years is the epitome of gov bloat and incompetence. Also the 6x or so budget bloat. Thats why NASA has a bad name.


JWST has at least justified itself with the excellent performance and better than expected predictions for its lifetime (since it is limited by the cooling and station-keeping fuel).

SLS has no such justifications forthcoming. It's too rough of a ride to launch any scientific probes on, it's too expensive for private launch, they can't make them fast enough to do more than 1 launch per year (without pouring more untold billions and decades into it), and the only reason it's even involved in Artemis is because Congress insists on it. Even putting Starship aside, between Falcon Heavy, F9, New Glenn, Vulcan, Neutron and maybe Ariane 6, the West has so much lift capacity available or coming online (mostly drastically cheaper than SLS) that SLS is kind of irrelevant.


At least it makes it easier to spot the inept and incompetent in the crowd.


You're a barrel of laughs aren't you :)


Unit tests — probably not.

They likely have a 'test article' though. See: https://en.wikipedia.org/wiki/Test_article_(aerospace)


That guy in the suite making sure he doesn't leave any dust speckles... but that mirror looks very dusty in that photo. I verified it is not dust on my screen :)


I hope they will publish firmware/source code at some point in the future. It would be definitely interesting to look into.


JPL has this neat guide to Voyager Telecoms with loads of details. https://descanso.jpl.nasa.gov/DPSummary/Descanso4--Voyager_n...


I suspect they can’t. It is quite possible security was not on anyone’s mind back then and the last thing they want are rouge signals from some amateur fucking up generations worth of work.


Sure if u have a 70 meter dish laying around and this is near peak cold War tensions so I'd say security was on people's minds then.

>Because of their distances, the Voyagers can only communicate through the largest 230-foot (70-meter) dish antennas in NASA's Deep Space Network, or by arraying multiple smaller antennas together to detect the faint signals coming from the spacecraft


> I suspect they can’t. It is quite possible security was not on anyone’s mind back then and the last thing they want are rouge signals from some amateur fucking up generations worth of work.

> Sure if u have a 70 meter dish laying around and this is near peak cold War tensions so I'd say security was on people's minds then.

Don't just think it's amateurs, nation states aren't above mischief either. If Russia has one of those laying around, there's a reasonably high chance they'd use it to send a big "fuck you" to America in today's geopolitical climate (e.g. command each Voyager to burn all its remaining fuel to spin on an axis perpendicular to Earth.

Edit: and it looks like Russia does have a dish like that:

https://en.wikipedia.org/wiki/Yevpatoria_RT-70_radio_telesco...:

> The Yevpatoria RT-70 radio telescope (P-2500, RT-70) is an RT-70 radio telescope and planetary radar at the Center for Deep Space Communications, Yevpatoria, Crimea. ... It has an advantage in comparison with other large radio telescopes in the fact that the complex includes powerful transmitters that allow active space experiments.


I highly doubt an amateur has the means to contact the probes. Heck, I'm not even sure any organization besides NASA has antennas big enough for the job.


You don't think a modified DirecTV antenna will do it?


Over 20 billions of kilometers away?

I'm not a RF engineer, but it's a fair guess that synthesizing the signal isn't going to be the problem, having the probes receiving it is another story.


If it's sarcasm I'm failing to detect with my l/4 antenna but no

And it should beg the question: why do you think DirecTV managed to use the smaller antennas ;)


The Russians, or the Chinese.


Maybe, but what would they gain by this exactly?

Aha! Now you'll have to wait another 40 years to collect information from deep space! Take THAT, Americans!


> Maybe, but what would they gain by this exactly?

It's called the Space Race for a reason.


It'd be one thing if those probes remained in near-space. There are of course advantages to a country having its own exclusive information strategically and possibly for future space-mining.

Deep space is quite literally as far away from that as it gets. There's no realistic mission that far out in our lifetime or the lifetimes of our great grand children that will have any realistic return for a nation beyond pure exploratory interest.

If Russia or China are really interested in objects far outside our solar system, they're probably better off sitting back and letting NASA do it, given that they don't have to wait decades to do so. Something tells me that neither of them actually cares, though.


That is true now, but maybe once the probes go completely dead with no chance of revival and there is no risk of tampering?


NASA have a history of reusing bits and pieces when they develop new vehicles and systems. If someone learns how to control Voyager there may not be much left to figure out if they want to control New Horizons etc.


That is a fair call. If that is the case, robably best to just leave it.


I am fascinated with routines they used for imaging. All parameters to make an image had to be sent from Earth. No auto exposure or anything.


I could be able to find out the code is a mix between assembly and fotran. I wonder how percentage of it is for sensor, navigation, communication




Anyone else disappointed they didn't mention the actual number of bytes in the update?



It's mind boggling to think the first of these launched when I was ~5 years old - and still operational, at these distances.


Can recommend the documentary It's Quieter in the Twilight https://www.imdb.com/title/tt17658964/ that follows the engineers working with voyager 1 and 2


What will be more ... (I don't know the word in English) is, once Voyager 2 completed the update it sends the same update to Voyager 1 (I read online it has the same computer system).



it really illustrates just how incredibly empty space is that these probes have survived operational and unmolested for this long.


I guess their code for signal correction for the case of package loss or data loss on in transit has to be fascinating. Or it is some simple, yet always working method. Anyway, would be cool to read about the details of that logic/implementation.


Recent and related:

NASA's Voyager Team Focuses on Software Patch, Thrusters - https://news.ycombinator.com/item?id=37963826 - Oct 2023 (48 comments)


How much is that in metric units?


1931256466260489.5 centimetres or, if you prefer, 3167950339119.439 giraffes.


I doubt that giraffes are metric. But maybe they are some obscure SI unit?

Fun fact. 1 billion km is almost 1.1 light hours. 1 billion miles is almost 1.5 light hours with less error margin. So you can easily calculate that Voyager 2 is about 18 light hours from Earth.


So hwo it "just" took 18 hours for the update to complete, according to the article? Shouldn't be at very least 36 hours, given that you need feedback from Voyager?


It took 18 hours to transfer the data to the probe.

According to the article the update hasn't been activated yet but they are still checking if everything is okay.

> NASA scientists are carrying out a readout of the AACS memory to ensure that it is in the right place for the update. If there aren’t any issues, it will trigger the Voyager 2 update on October 28.


Yeah, but how many football fields for us common people?!


Metric or freedom football?


Rugby football, which went metric some decades ago.


Yes.


Soccer, football? And how much is this in Empire State buildings, depending on the average soze of a giraffe?

Either way so, that SW update of Voyager is some seriously impressive stuff!


How about earth equator-meter as a unit? or Equatometer?


How many minicoopers filled with pingpong balls, if all pingpong balls were put in line?


It's in space, so we can say 129 Astronomical Units.

Easier to read than 19300000000km


I wonder which checksum algorithm they use at those ultra low latencies. Amazing work


There are some papers discussing this, e.g.:

"Polar Coding for Forward Error Correction in Space Communications with LDPC Comparisons", Naveed Namipour, Haleh Safavi, & Harry Shaw.

<https://ntrs.nasa.gov/api/citations/20190032244/downloads/20...>

Also "Design of Low-Density Parity-Check (LDPC) Codes for Deep-Space Applications" <https://ipnpr.jpl.nasa.gov/progress_report/42-159/159K.pdf>

The first specifically discusses early missions.


I'm amazed at the sheer sensitivity of technologies man can create. An object billions of miles away is able to sense and pick out our communication from the EM radiation noise hanging out in the vastness of space...


Still not accepted we do not have v3 and v4 every decade or a few decade.

https://news.ycombinator.com/item?id=37983710#37983731


Assembly software, Launched in 1977 with link speed of a few hundred bps


Firmware update in progress, do not unplug your computer...


To a spacecraft and to whoever else was listening.


And yet. I cannot get good cellular service next to Tmobile HQ lol. Always on wifi calling. Amazing.


They're about the same age as me, and they're aging better than I am.


KnowledgeWorks update most likely.


Does anyone else say “V-ger” in their head when they read about these missions?


I say “Voyager” in my head personally.


People jest the quality of our current software, but it's a real concern. The programmer of the '2023 update' is conditioned to bundle too much unneccesary bloat to the update that could eventually slow down and obsolete the device.


You don't know that. And this is one device that it really isn't possible to "bloat". Every byte and every bit in the voyager system needs absolute awareness. Bloat happens when programmers are unaware.


And it stresses me out just submitting code to github.


I really have to hand it to NASA. Voyager 2 was launched in 1977 and it's still getting updates but we're supposed to be okay with the cell phones that launched 5 years ago no longer being updated.


Long-term updated devices might be OK for your wallet and the environment, but they're not good at making profits. They also slow down "innovation", and what would we be without 100s of useless ("necessary") functions in our modern phones? We'd have to (gasp) talk to each other at restaurants or also use proper dedicated devices like a camera and would miss some selfies!


Old mobile phones aren’t transmitting priceless scientific data.

Long-term updates to old devices are a drain on the people updating them. You start a hardware company and let’s see how long you and your little team of really expensive engineers keep updating last year’s product.


>Old mobile phones aren’t transmitting priceless scientific data.

No, they just have (Apple 2022 iPhone revenues alone vs 2023 NASA budget) 10x the amount of money of NASA's budget (or 3x, if we stick to net profit), and around 1000x more money are depended on them.

>You start a hardware company and let’s see how long you and your little team of really expensive engineers keep updating last year’s product.

As long as the law mandates updates and bugfixes for older products, which, in a civilized society that actually did care for waste and the environment would be decades.


> Old mobile phones aren’t transmitting priceless scientific data.

Aren't they though? Clearly it's not on the same scale, but a 5 year old phone is still perfectly capable of transmitting very valuable data and providing utility to the user.


If you have Voyager 2's budget, I'm sure someone would be quite happy to sell you a phone that receives updates for decades.


It's not just an issue of money. Apple made a lot more money last year than NASA did.


And Apple would have made a lot more money if you spent the equivalent of Voyager 2's cost on your iPhone.

It is just about money. Cheap out and buy a budget phone and you will also get budget support.


I think it's better to say - if your phone was 12 billion miles from Earth and used to send sensitive scientific data - it would get updates as well.


Not when you only paid on the order of $1,000 for it. For so little, nobody is going to bother to support your device for decades, no matter how near or far away it is from you or what you happen to be doing with it.

The original Voyager budget was $865 million (~$5 billion today), and has paid out even more for maintenance and operation since. Spend that much money on an iPhone and not only will updates be available for decades, but someone from Apple will come and personally install the updates for you.


In your scenario there also needs to only be 2 iPhones in existence and making more would cost significantly more than the first 2 did, and you would have to wait decades for them to do anything noteworthy that the first 2 units aren't doing right now.

Instead, iPhones are like the basic bitch of devices and are mass produced to the point they are ubiquitous and still 99.999% of use of the device is pointless drivel.


Apple made more than 100 billion on iPhone sales in the first half of 2023 and the marginal cost of an update is essentially 0, so what your saying doesn't make sense.


And the vendors who supplied the Voyager program may have made quadrillions across the entire customer base for all we know. But there is no reason for a vendor to use the income from one customer to subsidize another.

If you only pay $1,000 for an iPhone, you are going to get $1,000 worth of support. Don't be such a cheapskate if you want greater service, I guess?


Economies of scale are not a subsidy.


That's right. Now, if you'd like to return to the thread...


The point is that they are getting updates, not because NASA paid 865 million in 1970s but because it is doing scientific work that no other spacecraft is able to do currently. If Voyager was orbiting around Earth it would get no updates at all, no matter how much NASA paid to get it into orbit.

So my point stands - if Voyager ran on an iPhone it would get those updates.


The point is: It gets updates because people are being paid sufficiently to provide updates. There is incentive given to see them get done.

There is no incentive for Apple to provide longer term support because nobody is ponying up the money to see that support happen. People cheap out and are only willing to pay for short term support – probably because they know that they can afford to buy a whole new, better phone (most likely for less than longer term support would cost!) again in a few years, so they don't care to brunt the cost of longer term support.

NASA knows they can't afford to buy a new Voyager every 5 years – no amount of money can place a new one to replace what is out there that far out in space – so they are buying long term support instead.

Your point only stands if NASA paid Apple enough to see them agree to update an iPhone for decades or more. Simply buying a $1,000 iPhone off the shelf and sticking it into a Voyager probe would not make Apple care or change their updating tune. And if that is your point, what is the point? You would be just restating what was originally said, which seems rather pointless.


You're right, I misspoke. I should have said it wasn't about the costs. NASA can keep updating their hardware after all this time with far fewer resources than apple has, so it isn't that it's prohibitively expensive to keep supporting old devices, it's just that ultimately "It is just about money" and cell phone companies would rather have people throw out their old devices and buy new ones.


Apple updates their phones for LOOONG time compared to Android


If NASA had the means to swap the voyager 2 with a new version, they would do that instead. If you have the means to swap the phone with a new version, you would do that instead.

--- open source enters the chat ---

Opensource: if you had the means to swap the code with a new version, you should do that instead.


This is simply dope! I know this comment doesn't add value, but still, Pluto is "only" 4.5 billion miles away from sun at its farthest point and it takes 247 years to loop around the sun.


Pluto is 3.7 billion miles away, not "light years". New Horizons probe wouldn't have taken "only" 11 years to reach Pluto otherwise.


Yup, effect of late night posting.


Pretty sure you got one of those measurements horribly wrong


I'll have you know that commenter was units lead on the Mars Climate Orbiter[0] mission!

[0] https://en.wikipedia.org/wiki/Mars_Climate_Orbiter


I think you mean 4.5 billion miles.


i wonder if NASA is the first "company" ever that implemented OTA updates ? (ok, not exactly air, mostly, space)


That's cool. My Redmi phone makers won't even send it a few miles across this same planet!


It's unbelievable.


Tested on staging?


Sent to Voyager 2 first, since Voyager 1 is further out so its readings are more valuable.

So... testing in production in a secondary market.


They have staging too (an on-Earth copy of the Voyager probe, with all relevant parts fully duplicated).


The real question is did they do it with a java applet?


They are doing it all wrong. Who runs on-prem hardware these days??? They should put it all in the cloud, they can replicate and have many instances of virtual Voyagers, not the measly 2. CDN should help with latency too, I hear their latency is measured in many hours, and is getting worse every day! Also, software update, big deal - use CI/CD!

They probably don't have a Kanban board either... Dinosaurs!


They are hard at work putting it in the Oort cloud.


Super low throughput and shard size, but robust vacuum gap security to defend against physical attacks.


Their latest patch was actually converting the entire app to use npm. Let's hope the poor probe has enough memory.


With npm, 64GB ought to be enough for everyone.


A rust rewrite from scratch is already in process


Great, I'll take the Web Assembly version.


A Go rewrite from scratch soon to follow.


A rewrite in Scratch is in the early planning stages.


then Microsoft Power Automate, basically the same thing as Scratch but we can get a support contract.


Actually a linux is in the pipeline. Everything has to run linux for once. Even in o… cloud.


Lol


I’m sure the AWS Ground Station[0] team is already on it. Soon you’ll be able to provision deep space probes via API and deploy your workloads in containers. (I hear the latency for bringing one up makes S3 Glacier look like a rocket sled though.)

https://aws.amazon.com/ground-station/


Run it in the nebula!


I was gonna say "run it in the stars", but yours is funnier for some reason :-)


I bet it doesn't even run Kubernetes.


I think whatever they're running is being run far beyond our cloud infrastructure at this point. That's rather high level stuff, but this is billions of miles higher level than that.


if they had built it with a modern stack it would have done well to last 43 days, let alone 43 years.


Actually no, they were first to have code successfully run in (space dust) cloud for so long


I both love and hate it at the same time, well done.


Love the Voyagers story.

Just get me thinking: Why there is not more missions like this ? It takes significant time to cross "somewhere there", every year one launch could not hurt ?

Edit:

*Obligatory video: "Uptime 15,364 days - The Computers of Voyager" by Aaron Cummings https://www.youtube.com/watch?v=H62hZJVqs2o

Also here on HN: https://hn.algolia.com/?q=voyager


Because the program literally depended on the planets aligning a certain way that they won't align again until about 2150.

https://en.wikipedia.org/wiki/Voyager_program#History


Make sense, thanks, fuel, gravity, not easy feat


Start getting the probes designed now.


> Start getting the probes designed now.

That makes absolutely no sense. Even someone who thinks technological progress will slow greatly in the future would acknowledge that something designed today will likely be obsolete and unbuildable in 127 years. I mean, probably *every* *single* electronic component in a design made today will almost certainly have been EOL for over a century by the time 2150 rolls around.


I say it in jest, but I am speaking to the experience of the JWST. Initial designs were proposed starting in 1996 and didn’t get completed until mid-2010s after many, many delays.

It would be no surprise to anyone to say the 2150 probes will have technologies that haven’t been invented yet, and won’t for some time.


Another constrained resource is communications: the Deep Space Network dishes are shared between all of the current NASA missions and already has rationing. So before upping the pace of new deep missions factor in the time for getting Congress to back expanding the DSN (not ludicrously expensive, but also not so sexy as launches, boots-on-the-moon, etc)

https://spacenews.com/increasing-demands-putting-pressure-on...


The answer is money.

Starship will both greatly reduce the cost to orbit, and the cost to design spacecraft (far wider mass budgets). I do not know how much C3 it will have, but very likely a kicker stage will come along for interplanetary and possibly interstellar craft.


The Interstellar Probe is something the Johns Hopkins APL has been proposing. I don't think its been funded yet.

https://interstellarprobe.jhuapl.edu/Interstellar-Probe-Exec...


The difficulty of long range communications is nicely demonstrated there: "X-band with 5-m fixed antenna capable of sufficient downlink (~10 Mbit/week) at 1000 au using Next Generation Very Large Array or equivalent resource" ... "Two Next Generation Radioisotope Thermoelectric Generators for 300 W (electric) at end of mission"

So with twice the area of radio dish and a power supply that after 50 years is still double what Voyager launched with, the data rate will be two bytes per second.


That's still a lot faster than FT8 or WSPR & both are perfectly usable for things like 16g solar powered balooon probes with a RPi Pico and a GPS, flying around the globe. :)

https://www.zachtek.com/post/a-new-wspr-transmitter-for-ball...

Should be similar out there & also likely not much stuff to take photos of.


Windows 11


Sir, we've just had a response for Betelgeuse! They say...can they have the old Start Menu back?


Ah, so that’s what Betelgeuse dimming was about, a few years ago


The cumulative patch.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: