Posting at a fiveish-week interval until this lines up with the first week of the month.
What books or pdfs have you been reading this month?
Neuromancer. I know, I know, it’s a classic and I should’ve read it years ago but better late than never.
I wonder why this book has become such a bible for our crowd. Sometimes I worry that we read it to the exclusion of a lot of other good literature. Sure, Neuromancer is good, but it shouldn’t be the last thing we ever read, nor should it be the thing we must all read, nor should SF and high fantasy be the only fiction genres we ever read.
I try to talk to people outside of our crowd because after a while it gets a little boring to be around people who all think Neuromancer is fantastic.
I still haven’t read it. It’s on my todo list for when things slow down more. What are some other ones you’d recommend?
Death’s End which is the third book in The Three Body Problem series.
Toward zero defect programming by Allan Stavely
Slightly disappointed it doesn’t talk more about the statistical methods used in the methodology.
I’m guessing that’s the textbook he referenced in the article I posted on Cleanroom. It gave me a “not found” error when I clicked on the link so I ignored it. Since you have it, how is it so far on explaining the methodology in a way that one can actually act on with modern tools? And do you expect little or lots of personal training from an expert before proficiency after reading that one?
The cleanroom methodology is based in two pillars:
The book mainly discussed the first one: it defines a precise language for writing specifications around box structures Around 80% of the book is defining this language that is verified in regular meetings (verification reviews) by the team.
For this there’s no need of special tools or software, as the cleanroom is more a process where the specs are shared & debated among the development team. This is probably today something not so controversial, and can be deployed applying some common sense.
For the second pillar, the book is much less useful, as the statistical process/methods are not explained, other than general guidance.
So, can it be used today with few changes to the tools: yes, with no doubt. Does it requires a mature and organised team? yes, completely.
Around 80% of the book is defining this language that is verified in regular meetings (verification reviews) by the team.
That’s good. That plus basic logic are what people need to know anyway to leverage the techs being made in fields of static analysis and lightweight, formal methods. Appreciate the feedback.
Calendrical Calculations deep niche knowledge but the presentation is very entertaining and fun to read.
I just finished The Magicians by Lev Grossman last night. Up next, I’m going to finally finish Clojure For the Brave and True.
I read a closed and common orbit as soon as it came out. Fluff (particularly as it’s not so much a sequel as a side story), but fun and engaging.
Gradually working through Hubris: the tragedy of war in the twentieth century; it’s interesting but slow going. I’d also started on The Ipcress File but think I’m about to give up.
Further reading delayed a bit at the moment as I’m waiting for a replacement credit card (don’t want the faff of changing cards on amazon twice).
i’m in the middle of “closed and common orbit” right now, and it’s great! good old-fashioned adventure sf with engaging characters.
My reading list is currently:
And a handful of others.
Every Day Is for the Thief by Teju Cole.
Oh, that looks interesting, a tale of what modern Nigeria looks like. I’ll look for this one.
Just finished Mark Tarver’s Logic, Proof, and Computation last weekend. It was really good, explained everything simply, and the references provided at the end of each chapter helped grow my reading list for the next few months :)
Also: Solaris by Stanislaw Lem, Godel, Escher, Bach by David Hofstadter.
I need a non-techie book to read next, maybe Walter Isaacson’s biography of Ben Franklin next (it’s been sitting on my shelf half-read for years)? Willing to take recommendations…
I’m going to try an intro book to formal methods at some point. Right now I’m collecting recommendations for any intro books that make it really easy for people that usually have hard time with formal logic or proofs. I don’t know what it is about them but many resources just didn’t work for me. Ideally, they’ll include examples of how the logics map to use with real-world problems even if they’re toy problems. I need have that mental connection for later application to proving software systems.
How’s Tarver’s fare against that criteria? And any others you know?
It took me a long time to learn formal methods too, and I’m still learning. Tarver’s books have a heavy computing and functional-programming bent to them, so maybe you’ll relate to that. Reading LPC, I feel like I’m having a discussion with a fellow hacker, not being lectured by a professor. Each chapter has a “Further reading” section with references to primary sources, which is nice. Also, LPC actually comes with a proof assistant written in Shen, you can download it for free here.
However, LPC only covers one strain of logic: that which has application to computing, and specifically the type of computing that Tarver likes to do (for example: Tarver really likes symbolic-based AI methods, he never mentions statistics-based AI or neural networks in either of his books). There are many more types of logic in the world; if that’s what you’re interested in then I recommend looking at Irving Copi’s intro book. According to my philosophy professor from college, this is a standard in academia, covers pretty much all the basics, and has been published since forever so you can find a cheap old edition. I have the second edition from 1961 on my shelf…
The SEP also has a ton of logic articles, more than you could probably ever read in a lifetime.
Tarver’s books have a heavy computing and functional-programming bent to them, so maybe you’ll relate to that.
Fortunately, I collected books to make functional programming easy, too. So, I guess I learn that then try Tarver’s books. Seems most proof-oriented books use LISP or ML’s. Guess I should learn their foundations thoroughly.
“ that which has application to computing, and specifically the type of computing that Tarver likes to do (for example: Tarver really likes symbolic-based AI methods, he never mentions statistics-based AI or neural networks in either of his books).”
I’m fine with that. It’s where basically all the results are happening in the formal methods community. I’m sure the others will get more useful over time but the symbolic stuff is both effective and traceable in how the system “thinks.”
I have A Brief History Of Seven Killings on my bedside table, glaring balefully at me. I need to get started on that; also, The Mandibles: A Family. I’m trying to read more fiction that isn’t simply a genre exercise.
Deathride by John Mosier. Great book about the German Eastern front from 1939-1945.
read a bunch of really good sf/f this month, most notably liu cixin’s “remembrance of earth’s past” trilogy, jo walton’s “necessity” and n k jemisin’s “fifth season” and sequel.
liu cixin is the closest i’ve found to “the new arthur c clarke”, albeit a fair bit bleaker. if you like clarke, baxter and brin you should definitely like him.
Finishing off Adrian Tchaikovsky’s Shadows of the Apt series, as well as N.K. Jemisin’s The Fifth Season and starting the Split Worlds series by Emma Newman.
These sound like SF titles. Are they?
Mix of flintlock fantasy and SF
West of Eden by Harry Harrison and OCaml from the Very Beginning by John Whitington.
My copy of the The DevOps Handbook arrived today so I’ll probably be reading that over the coming few weeks.
More Programming Pearls by Jon Bentley. I read a chapter already and I’m really liking the style of it.
Programming Pearls is amazing? So I expect a lot from “More”, but I haven’t read it yet, sadly,
I’m starting the second part of the book of the new sun. Then I want to read a few Ted Chiang stories, since I liked “The Story of your life” a lot.