A Deepness In The Sky | Hacker News Reader

A Deepness In The Sky

hn-books.com

There's a great scene where young Pham, a kid from a medieval world, learns to maintain a 5000-year-old legacy code base on a starship:

http://books.google.com/books?id=GUUvxumMf6kC&lpg=PA226&...

Sura gave up on her debugging for the moment. "The word for all this is 'mature programming environment.' Basically, when hardware performance has been pushed to its final limit, and programmmers have had several centuries to code, you reach a point where there is far more signicant code than can be rationalized. The best you can do is understand the overall layering, and know how to search for the oddball tool that may come in handy—take the situation I have here." She waved at the dependency chart she had been working on. "We are low on working fluid for the coffins. Like a million other things, there was none for sale on dear old Canberra. Well, the obvious thing is to move the coffins near the aft hull, and cool by direct radiation. We don't have the proper equipment to support this—so lately, I've been doing my share of archeology. It seems that five hundred years ago, a similar thing happened after an in-system war at Torma. They hacked together a temperature maintenance package that is precisely what we need."

"Almost precisely." Bret was grinning again. "With some minor revisions."

I love the terms "programmer-at-arms" and "programmer archeologist." I wish more people read this book, so I could use the term and not have people look at me with a blank look in their eyes.

The above excerpt is actually an accurate description of what I do at my day job! ("program archeology") My coworker, who's been working on and off with this code base for over 12 years, never fails to point out if any of my ideas was already implemented once before. So far, everything I've suggested since starting this job 3 months ago has already been done in the past.

My job title at a previous job was Software Archaeologist, inspired by ADitS. Though my boss then went on to explain it to a customer more as an Indiana Jones of programmers than as someone working on old and crufty systems.

My favourite part of the program archaeology is when he digs right down to the core and discovers the computers are still keeping time in seconds since the time when man landed on the moon (and then discovers it's really a time a bit more than five months afterwards).

I never did think of UNIX time as tied to the moon landings that closely, but it's a nice fact. It was still odd seeing small negative timestamps when we were working on spacelog.org, though...

I started getting excited but then quickly realized that if I had to read a sci-fi novel about coding (or maintaining a codebase) I would throw up.

I spend my days coding.

It's not what the book is about, just a few interesting/realistic scenes. The book itself is epic and brilliant and spans almost every genre.

Fortunately, not everybody hates their job.

My worry about coding/hacking scenes in books is that authors usually do something dumb. That makes it all the more rewarding when they sidestep it with something amusing, as in the given example.

Having read 'A Deepness in the Sky', it is good, interesting, and not about coding. You can probably read it safely without having flashbacks to your life.

The first time I read this, the middle of the book tried my patience and I wanted it to just get to the conclusion; I started skimming a bit more until I got to the end.

The second time I read the book, I had planned on just skimming the middle again, but this time without the drive to know the resolution (since I remembered it from last time), I discovered that the middle of the book is really the best part. It's very detailed and well-thought-out, and the second time I didn't even have a problem with the pacing. It's just a very big book, in a good way.

The only other book I know of that I have had this reaction to was the Lord of the Rings series. The final denouement feels like it's taking forever the first time, but subsequently, the only thing I routinely skip over is the foot chase through Rohan. (Which I remembered as being 50-60 pages long but turns out to actually be only about 12. Still manages to feel 50 pages long to me. YMMV.)

Strongly recommend, and despite being listed as a "prequel" you do not need to have read A Fire Upon the Deep first; only a handful of throwaway details will be mysterious as a result and none of them matter much to you. Vinge does not fall into any of the usual "prequel" issues that for instance the Dune prequels did ("Remember that expansive universe you loved that so carefully managed to feel like just a small part of the totality? No, it was the entire totality and the only people in the universe up to that point were the direct ancestors of the people you loved.") or other things. This is truly a prequel, a story that took place before a later story, and temporal ordering is honored.

> The final denouement feels like it's taking forever the first time

Final denouement? Man, I gave up trying to read it after what felt like seventeen chapters of "they walked through a field."

In Tolkien's defense, the man had some of the finest field-describing skills I have ever seen.

I believe he came closest any author ever has to walking someone else through their imagination.

Tolkien only wrote a story incidentally. What he did was show us the real Middle earth, the one in his mind.

Most authors are the Lonely Planet Guide to their fantasy worlds. Tolkien... Tolkien was David Attenborough.

As someone whose imagination was shaped in large part by reading Tolkien's vivid descriptions of Middle Earth as a child, I couldn't agree more. You put it beautifully.

It works fine standalone, yes, but if you expect to read both then read them in publication order, because there's a powerful resonance/irony implicit in aDitS with aFutD as background: http://www.tor.com/blogs/2008/07/deepnessreview

Vinge is, in my opinion, the preeminent hacker SF writer. He was a CS professor at SDSU and both the two ZoT novels and Rainbows End show that he really grasps the end-to-end nature of technology and how hackers interact with the world. I also thought that Rainbows End was a great story about how hackers age, and it's just as well worth reading as Deepness.

> At one point I think Vinge showed his manuscript to somebody else, because after that he spends a bit of time defending his description of some of the aliens The author protests a bit too much. Because it's a prequel, there is just a bit too much detail.

Almost surely. He had a number of consultants and others go through _A Fire Upon the Deep_ (which I liked better), as one can see from the annotated edition: http://news.slashdot.org/story/03/09/18/0411259/Review-A-Fir...

I'm really curious what draws people to these two books (and I've read both books and the review). I found them both to be somewhat slow and about half-wasted with all the stuff about the "Tines". While I love Vinge's writings on the future of technology, his books (including Rainbows End and The Peace War) have never worked for me.

I have a pet theory that people who really like Stross' Accelerando (super-fast, super-speculative buzz-wordy, hard science fiction), Banks' Culture books or Morgan's Market Forces would have a harder time with A Fire Upon The Deep or A Deepness In The Sky. Yes/No?

I haven't read Market Forces, but Vinge, Stross, and Banks are my three favorite sci-fi authors, so I'm a counterexample to your theory. I didn't find the Tines boring...honestly, the Zones bothered me more, because it was so central to the story and so unscientific. But it still didn't really bother me, and I enjoyed both Fire and Deepness a lot.

I thought Zones were a great idea - what did you find "unscientific" about them? (Allowing for the fact that this is all fiction).

The fact that they were simply posited, with no explanation? There was no science there, just a mystical "this is the way it is".

There are a lot of hints about how the Zones work, spread through both books.

The overall picture seems to be this:

* What we think of as the "laws of physics" in any given Zone are actually a kind of VM layer, running on top of some (unknown) physics.

* The implementation is done through extraordinarily tiny machines. What we see as quarks, leptons etc (and the laws they follow) are in reality complex machines made of far smaller constituents.

* The implementation was probably done by some power-beyond-the-powers that lives in the center of the galaxy. It uses the Zones as a protective device - civilizations in the Slow Zone, the Beyond and the Transcend have only relatively limited abilities, unless they can break through the virtualization layer. (That, incidentally, is why the Blight is such a threat in AFUTD, and why the Zones change so much in response.)

If you read both books with this set of ideas in mind, a whole lot of otherwise puzzling scenes and one-off observations in the books make a great deal more sense. There's no reason to call the Zones unscientific.

Yeah, I agree except for the last parenthetical remark -- I don't recall any evidence for it in the text. Why would the powers-squared need us to invoke them for their own security? I suppose it was more about protecting the wildlife preserve, to the extent it makes any sense to us.

Can't remember where I read it, but another way to look at the Zones was that it breaks the idea and occurance of a technological singularity from a temporal to a spatial dimension.

Given what we know in the realm of complexity theory, I see no justification for arbitrary thresholds in computing power based on space. Nanomachines can change matter but they can't change math, and the fact of the matter is we know a reasonable amount about what we can compute (and how long it takes) and what we can't. Either P = NP or it doesn't, for example; we don't know which, but we're pretty sure the answer doesn't change dependent on where you're standing. Leaving the galactic plane isn't going to enable me to solve the halting problem.

If you change the underlying physics, computational power can change quite a bit. The best-known example of this is that quantum computers are widely believed to be more powerful than classical computers. An even more striking example, though less well known, is that certain non-linear variants of quantum mechanics allow NP-complete problems to be solved in polynomial time. This result was written up here: http://arxiv.org/abs/quant-ph/9801041

We could imagine, then, that what's going on is that the Universe really runs according to the laws of one of these non-linear variants of quantum mechanics, but the power-beyond-the-powers has arranged so that at the length scales that can be probed by the races in Vinge's story things look effectively like standard (linear) QM.

The thing that bothers me a bit about the Zones: how is it that humans think? We are, after all, just machines. So if AI can't be built, then how is human consciousness possible?

I'd be interested to know what (if any) response Vinge has to this. A possible-but-not-very-satisfactory response is that life on Earth is really an artifact from another Zone, and one that goes beyond ordinary Slow Zone capabilities. What prevents us from building AI is that we can't reproduce those capabilities. Seems a bit of a stretch.

I'm no expert on the field, but I believe the optimality of Grover's algorithm strongly suggests that the upper bound on the speedup from quantum computing is quadratic (and I apologize if I'm fudging something here). I have no idea about non-linear QM - really over my head - so I suppose I'll have to concede that point.

I agree with you that having conscious carbon but not conscious silicon is arbitrary and silly. Once we understand the brain better we'll surely be able to build one. Again, not something that should depend on location, nanomachines or not.

The optimality of Grover's algorithm doesn't in any way suggest that the upper bound on speedup is quadratic. That result only applies to one very particular class of problems: unstructured (oracle) search problems. There are other oracle problems for which the speedup is known to be exponential. For non-oracle problems the situation isn't yet clear, although many people believe the speedup will be exponential. That will be true if factoring isn't in BPP, for instance.

If you push any science fiction hard enough, you will find something impossible in the real universe. If this is your criterion, stop reading science fiction. You need to cut them at least a little slack.

Further, I don't think you are correct it your interpretation. My interpretation was: What changes is not what is computable and what is not, but what bandwidth the components are capable of communication with each other at. Bandwidth limitation limits the power of distributed computational systems, and when you're limiting things at the atomic level that includes the brain. And it is implied that it was constructed this way.

Yes, that's probably impossible. It's probably also impossible that there is a computational threshold above which you can do FTL.

You might be right about the bandwidth thing - been a while since I've read it - but it doesn't change the fact that crypto algorithms suddenly breaking (for non-brute-force reasons - again, you're saying it's bandwidth and not computing power) or AI suddenly becoming possible once you're saturated in a different mixture of nanomachines doesn't make sense.

I cautiously disagree with this, by the way:

>If you push any science fiction hard enough, you will find something impossible in the real universe.

What in Accelerando is impossible? Even more generously, what in it violates our current understanding of physics?

> What in Accelerando is impossible?

Faster than light travel. We do not actually know how to build computronium and it probably won't have the exact parameters given in the novel; real computronium may be many orders of magnitude less powerful. We do not know that living on Saturn is feasible from any engineering point of view. We do not know that there is any level of technology that will enable us to build a living adult human-like body in a matter of days as described in the book, as the ape manifestation of Manx was created. We do not know the limits of nanotechnology. We do not know that there is a path from humanity to post-humanity that still includes entities that were at one point human. We do not know what the limits of intelligence are. Is "Accelerando" really possible, or is it merely hiding in the unknowable more cleverly than older science fiction authors?

Bear in mind I often cite Accelerando as an example of sci-fi that is mostly possible (excepting the faster-than-light bit), but it's more that we can't currently prove it impossible than that it is guaranteed possible. We can't eliminate it on scientific grounds but there will be engineering limits in the real universe, too, and there's no guarantee that anything like that is actually possible. And the probability that the social implications of everything Accelerando laid out are correct is effectively zero, no offense to Stross, as nobody can really work it out.

We further do not know that we are not actually living in a simulation in which the "magic" parameter is merely set to 0, but switchable tomorrow on the whim of the Great Simulator. If you do not like the Zones of Thought today, well then the Accelerando-computronium can simulate it adequately enough to meet the criteria of the story one way or another. Given that this actually happens in Accelerand itself, I'm not even stretching; consider A Fire Upon the Deep as Accelerando fan fiction if it makes you feel better.

I'm advocating being a little more careful what nits you pick not because I'm a hippy who just wants to bask in the glow of the work, but because I've gotten to be a master nitpicker over the years. Nothing's perfect.

I thought the zones being unknowable was part of the story.

The Magellanic Cloud aliens created (or at least can manipulate) the zones to stop transcendental "things" like The Blight from destroying all sentience (and then becoming a problem for the Cloudies?).

The zones are two things: a few layers of "play nice, you guys" and a cheap way of avoiding singularity issues muddling up the plot.

>I thought the zones being unknowable was part of the story.

I'm prepared to call unknowability unscientific.

>a cheap way of avoiding singularity issues muddling up the plot.

Yeah, exactly. They weren't an interesting idea so much as a cheap hack.

They're heavily implied to have been created by an ancient civilisation, either as an weapon, or as a nature-reserve for less advanced species.

A definite No I'm afraid - I love most of Stross' work (particularly Accelerando, which has an insane idea density), the Culture books in all of their socialistic contrarian splendour but I also love all of Vinge's work - particularly A Fire Upon the Deep (for its sheer sensawunda) and Deepness.

Another counterexample here. I've loved everything of Vinge's to date, and even like the stuff Charlie wrote before he got published (it's pretty damn freaky stuff for idea density). Don't like all the Culture books, but I do like many of them.

"socialistic contrarian splendour"

They're really post-scarcity; they have no need for capitalism or socialism.

I loved the Vinge novels, and I'm also a fan of Accelerando. I looked at a couple of Culture books, but really didn't like them. So here's one data point saying that things don't break out that way.

Never heard of the Morgan books before. Looking them up, it looks to me like the author is playing on popular conceptions of capitalism, without a solid grounding in real economic theory. Is that a fair impression?

Yeah, both a Deepness in the Sky and Fire upon the deep didn't do a whole lot for me.

The Aliens in DitS were too "human", which was a problem for me because he went to great lengths to describe what their planet was like, and their physiology and how and when they reproduce. Oh, and they have a major world war, and discover the exact same technologies in the exact same order as humans on earth!

Additionally, Vinge has very litte "reward" in his stories. Very few deep insights.

For my money, Neal Stephenson is a much more entertaining SF writer. Every time I read Anathem, it changes the way I see, well, everything, for weeks after I'm done.

Initially, I also thought the spiders were a bit too "human"... however, in thinking about the way Vinge portrays them, it was important for me to realize there's a point later in the story where he reveals that the focused translators are heavily anthropomorphizing everything they relate to the rest of the crew. I think this is a really strong suggestion that everything told from the point-of-view of the spiders is passing through the same narrator. There's a point where he specifically mentions that the translators seem to know about events and conversations they shouldn't, and we can infer this is because they've been covertly communicating with the Underhill for a long time and he's given them the back-story.

Also important, I think, is that a huge contributing factor to events and technologies developing similarly to those of human civilization, is that the humans were actively manipulating information and events on the planet from the moment they arrived.

I agree with you about Stephenson being fantastic; his books make you think. While I've enjoyed what I've read from Vinge, in some sense it feels much closer to the fantasy genre.

I'm curious as to what folks here thought of The Baroque Cycle. I loved Snow Crash, Diamond Age, and Anathem... but Quicksilver seemed to drag, and I didn't have the heart to pick up the next one in the series.

There are at least three more different hints about the translators than you've mentioned -- Vinge is so great at embedding little insight-bombs in the narrative, like Godel, Escher, Bach. In fact the whole translation subtheme strongly reminded me of another Hofstadter book, Le Ton beau de Marot.

The Baroque Cycle was among my favorite books of its years, FWIW.

I couldn't even finish Quicksilver when reading it as a book. However, the unabridged audio-book version that Audible has of the Baroque Cycle is extremely good - I love listening to these books!

> I think this is a really strong suggestion that everything told from the point-of-view of the spiders is passing through the same narrator.

This is explicitly hinted at when, uh, Vinh finds papers in Trixia's room that are identical to the first chapter from the point of view of the spiders. The point is that everything written about the spiders/from their perspective is actually her "novelization" of what happened.

> Every time I read Anathem

shudder

I've liked some Neal Stephenson, and I wanted to like Anathem, but just could not get "into" it. Maybe I'll have to try it again.

It may start slow (though personally I loved the start), but it sure picks up towards the last third. It blew my mind into many, many pieces.

I'd give a strong recommendation to stick with it - at first all of the strange invented terms are a bit jarring but after a while they become natural and you get to enjoy what I would regard as one of the finest work of fiction of recent years.

I read Vinge but not Stross. Can't begin to tell you how much I hate buzzword fiction.

I'm surprised at the lack of Stross--Accelerando describes a fairly Hansonian singularity.

I love everything I've read of Vinge and can't stand Stross, so that theory sounds reasonable.

You pegged me perfectly. I've been reading Ian M. Banks since I was a kid. Love most of Stross's novels, not just Accelerando. I thought Rainbow's End was weak writing at best. Fire was decent, as was Deepness, but overall Vinge is hit or miss with me.

I love Stross and Banks (have not read Morgan yet), but also have a profound love for both of the Vinge books you mentioned. Other Vinge's material, not so much. :( I hope he'll continue with the zones of thought saga some day.

Number three is nearing publication - this year! (http://news.ycombinator.com/item?id=1706449)

No. Vinge fits rather nicely into the Banks/Stross/Reynolds/Gibson/Stephenson fold of decent recentish SF.

no, not read any of morgan, but i love all three of stross, banks and vinge. 'accelerando' felt like a far more minor novel than, say, 'player of games' or 'a fire upon the deep', though. if you want super-fast, super-speculative hard-science fiction, check out baxter's xeelee sequence ('vacuum diagrams' makes an excellent starting point)

> I'm really curious what draws people to these two books (and I've read both books and the review).

As far as ADitS goes: it's the epic arc of human civilization over time and also the clash of different civilizations that really brings out the hidden assumptions we make based on our human condition.

Especially memorable was the history of the Qeng Ho, and how the customer planets die off as civilizations rise and fall.

Accelerando was all right, but unexciting if you already read of most of the ideas back in the 90s (in Extropy, etc.). Similarly Vinge's _Rainbows End_ was a much bigger deal to me than Stross's _Halting State_. I do read Stross, just without the same avidity.

I like Banks; he's a first-rate writer, as a writer. Not really in the same category for me of writers who hit my hacker sweet-spot: Vinge and Stephenson.

Haven't read Morgan.

A bit of a Blub paradox here wrt Stross and Vinge. My author can beat up your author?

I'm going to have to disagree with you, all of those are some of my favorite author/books

Just want to point out, that the reason he cites is NOT, at all, the reason why most sci-fi sucks.

The real reason that most sci-fi (and fantasy and romance and most 'genre' novels) suck is that they focus too much on the "stuff", and not on the things which actually matter in narrative fiction writing.

The things that really matter in a story are characters, strong motivations, good dialogue, plot, conflict, etc... techniques of the writing craft. These are what make a book good or not, but they are kind of an afterthought to all the "STUFF" in Sci Fi books.

I also recommend _A Fire Upon the Deep_ which is "space opera" written as intelligently as possible by "someone who should know better." (Vernor's words)

You might find this interesting. An alternate fan-fic ending of A Fire upon the Deep (and what comes after for the direct participants) by Eliezer Yudkowsky.

http://www.fanfiction.net/s/5389450/1/The_Finale_of_the_Ulti...

That's fundamentally fanfic of Permutation City rather than AFutD.

It may be a different medium, but I was impressed with the technobabble in Mass Effect. It actually made sense.

Every single detail, from astrobiology to weapon systems and shields, had a plausible explanation if you cared enough to visit the Codex. The whole game seemed like it was expanded from a giant "what-if element zero".

Maybe I'm deeply wrong about all this, but I think the game is not receiving enough praise for this side.

All my physicist/video gamer friends love it for that.

I think the entire game was a labor of intense depth. The internal fantasy was watertight, from the technology, to the politics, to the way the story changed dynamically depending on your choices.

What most impressed me about ADITS was how I was brought to feel great empathy and even affection for the spiders, not till the end do we see just how truly alien they and their environment are.

I agree, the 'golden age' narration was a great inversion, where the spiders seemed more human than the humans did.

So, what do people think of "focus"? (A drug treatment regime used by the "emergents" to turn people into monomaniacs.)

Vinge goes out of his way to show how it could be abused --- but I sometimes get the feeling that people crunching in early-startup mode would willingly use it, at least for a time, if they had it...

It's called adderall, I'm a fan.

Greg Egan has a similar concept ("neural mods") in his book Quarantine:

http://en.wikipedia.org/wiki/Quarantine_%28novel%29

What I thought was extremely interesting was the way the way the some focused were used... as components behaving as the intelligence behind computational tasks. The creator of Mechanical Turk undoubtedly read this book... I sure thought about creating something like that after I read it.

If I had a timer on it, I'd sign up for Focus in a heartbeat.

The concept of focus was mesmerizing. It's well-accepted that our brains do amazing things under the covers, and the idea of bending our brains to a single focus speaks volumes about human intelligence and genius. Very evocative.

It's kind of interesting that the review (and the other comments here so far) doesn't even mention the book's political themes - the pro-market forces vs the evil socialistic Emergency. Is that view just so ingrained that it's not even worth mentioning?

the book's political themes - the pro-market forces vs the evil socialistic Emergency

I enjoyed that in the books, but I'll point out that his Realtime books are much more explicitly anarcho-capitalist.

As an aside, I think you're wrong to characterize the economics as politics. It's very common to assume that such as "minimum wages increase unemployment at the margin" as a direct condemnation of minimum wage policies. But that presupposes a given set of values. It's quite quite possible to make a political statement in favor of minimum wages (for example), while acknowledging the economics of it.

Economics and politics are two separate things. They are related, they inform each other, but there's always a case to be made for things like "I accept that doing X would have a negative effect on total economic efficiency, but we have a moral obligation to pay that price."

For some evidence that your point "it's quite possible" isn't merely some idle hypothetical, search for "Webb" and "Seager" in

http://www.princeton.edu/~tleonard/papers/retrospectives.pdf

The same author (Leonard) also published a paper in _History of Political Economy_, "More Merciful and Not Less Effective." I remember it as similar, but I'm not finding an ungated version online.

The Emergents were socialists? I've always seen them much more as fascists. For me, the sensor network plus 'focused' monitoring was a pretty terrifying exploration of a technologicially perfect fascist state.

The author has a sequel to _A Fire Upon the Deep_ coming out in the fall -- Children of the Sky. Set in the same universe.

I've read all of Vinge's work and it really is phenomenal. It's been years since I read it and can only remember vague concepts now... the singularities, the 'focus' drug that makes his programmers super-human.

I really must read them again. Quite excited to hear a new book is coming out.

I found it really really hard to gather with it and never finished it. Same was with "A Fire Upon Te Deep", but after the first quater it was okay for me, and I finished and liked it.

What I really found hard was, that this takes all place in it's own Universe with it's own Rules. And no where is described/explained what the these things or rules are. Just like you would dive into e.g. Star Trek TV Show for the first time and see someone beam. The person would disapear and repear somewhere. Due to the fact that you see it, you understand what this beaming thing is. Not so in Vinges Universe. Same applies to Gibson and the Sprawl. It's hard to mess around with it, when no where things are explained. Mabye this is due to the translation of these books (I am a german reader). Any others experienced the same/similar ?

Daniel, you've really fleshed out the site since last time I looked at it -- I think when you first posted it here. It's looking really good. Any update as to how the project as a whole is going for you?

F- yeah! This (don't forget to read the other book first - a fire upon the deep) is my third favorite sci fi novel series of all time. Right after foundation series and dune collection.

A must read for all hn readers.

Blabber is a short story, Vinge's first foray into the Zones of Thought Universe, and features several cameos from A Fire Upon the Deep:

http://books.google.com/books?id=txBNZiQw2GUC&pg=PA78...

Ah, the Kindle version is not available in Europe.

> Your search "A Deepness In The Sky" did not match any products in: Kindle Store

Oh come on. I'm thinking about selling my kindle. Not one of the books that really interest me are available for it :/

?? It's there, and also on the B&N nook:

http://www.amazon.com/Deepness-Sky-ebook/dp/B002H8ORKM/

http://search.barnesandnoble.com/Deepness-in-the-Sky/Vernor-...

Oh, yes. You're right.

But when I click on your like I get:

> This title is not available for customers from your location in: > Europe

Reply on news.ycombinator.com