Deepness in the Sky by VernorVinge
It involves (among other characters) Pham Nuwen and the Qeng Ho.
It's set in the galaxy of the Zones of Thought - but no one in the book knows about the Zones. It's all in the Slow Zone, and all involves sublight travel. As a result, the sense of scope is, oddly, greater than in AFUTD - the people in it really are dealing with immense spans of time.
It contains one (1) race of absolutely wonderfully-done aliens, from whose point of view we see part of the story.
It's an extraordinarily well written clockwork thriller. I have noticed for a while that each successive book by VernorVinge is twice as well-written as the previous one. In 2009, I expect him to win the Nobel Prize for Literature.
If you like A FireUponTheDeep, check for an older book by Vernor named "True names and other stories". TrueNames is an award-winning short story (Hugo, I think or maybe Nebula) that defined Cyberpunk back before WilliamGibson ever thought about it. What's more, it's based on an idea that he should have patented, since it's now the "Greatest thing since sliced bread" (TM) - namely peer-to-peer computing and the ability to zip around a network grabbing processor power wherever you can get it.
-- KyleBrown
DeepnessInTheSky is one of my favorite science fiction books, not only for the good writing and engaging story, but for the ideas. Isn't that, after all, what distinguishes topnotch ScienceFiction from merely good writing?
To me, one of the deepest ideas is that of FailedDreams?, the idea that there are certain technological achievements which may seem plausible today (i.e., late 20th century), but will in fact forever stay out of our grasp. DeepnessInTheSky features two particular failed dreams prominently: artificial intelligence, and trustworthy computer systems.
The artificial intelligence debate is one that's been raging since Turing wrote has masterful paper, ComputingMachineryAndIntelligence 50 years ago. Since a great many science fiction stories feature talking robots and the like, it's interesting for Vinge to take the contrary position.
The other failed dream, trustworthy software, is probably a lot more immediately interesting to this community. Vinge proposes that the far future of software is built on layers of cruft that make existing systems like the x86 architecture seem coherently designed by comparison. Software of the future will have bugs and lots of them. Most people will work around the bugs, while the most clever will learn to exploit them. Actually, this doesn't sound very far-future, does it?
I personally believe that in the far future, software will be systematically designed and formally verified. To me, software is built from mathematical concepts, and the full power and trustworthiness of mathematics would be available to us if we only had the will power and determination to use it. Extrapolating linearly from trends today brings you a world much more like Vinge describes, but ultimately I can't imagine that people will be happy for much longer with the buggy and insecure state of software.
We'll just have to live a long time if we really want to know the answer, though.
-- RaphLevien
One thing of particular interest to computer geeks... the profession of one of the major characters is ProgrammerArcheologist?. The idea is that over thousands of years, virtually any bit of software that might need to be written has been written. Rather than writing new code, the ProgrammerArcheologist? learns to dig through the files for old stuff to rework. It feels vaguely familiar to any FreshMeat readers and other Unix programmer types.
-- DaveStagner