Learn C and build your own Lisp

A lot of times TiB is spelled 1024 * 1024 * 1024 * 1024 or 1 << 40, relying on compile-time constant folding to compute the final value. I usually find that a lot more readable than remembering the actual number.

On the compiler course taught by Brucee (Bruce Ellis - one half of Mark V. Shaney - the other half being Rob Pike) he got you to do two projects : write a Lisp implementation in C, and a C compiler in Lisp. Most instructive.

If you have an HP printer with a Postscript renderer and you can get an image of it, you can find a digitised photograph of him too :)

I've just begun starting a group to create a Scheme implementation from scratch in London (UK). If you're already somewhat proficient in Lisp/Scheme and interested in participating, see [1] for further info and send me a message through meetup.com or through the link on my contact page.

[1] http://www.meetup.com/London-SICP-Study-Group/messages/board...

In fact, this one (implementing Lispy thing, I mean) seems the best to me so far. Why? Paradoxically, because it's so obviously incomplete, so it makes you search something by yourself while you read it and helps to actually understand a little bit more.

Both K&R and "Modern Approach" lack that word "Modern" very much, yet are written with undertone "it's everything you have to know about C", when it isn't. Person who doesn't know C very likely doesn't know how OS works, what architecture layers are behind the software he uses every day. He just knows somehow he wants to learn C, but doesn't really understand what C is, yet he knows that virtually everything is written in C, and "everything" usually doesn't run in terminal, but processes and produces sound, images, video, can have GUI, use some external devices, run in parallel, run on GPU. Also, it's pretty obvious (especially if you've used some language like Python already) that nobody writes everything from scratch these days, but there're many libraries that proved to be useful.

And stuff like branching and cycles seems to be pretty obvious even for somebody without coding experience, as far as I can judge from what I've seen so far. Yet we have plenty books that spends 20 pages to explain "if" keyword and mentions every function in standard library (why?! it's 2014, people, we have cplusplus.com now!)and covers nothing in sense of what useful libraries are out there, what are these domains where you still should use C today (because feet to meter converter shouldn't be written in C today and you probably should choose Python if there's no specific reason to use C), what tools to use for testing and debugging and such.

And while it can be justified for K&R (jeez, how old is that book!) it's just ridiculous that book that has "Modern" (well, it was 2008, but still…) in it's title covers almost nothing of what person who wants to program in C should know today. Worse, you'll see that only after reading these 800+ pages.

"C the Hard Way" is a little bit better, but still not as good as this one. This one also isn't perfect, but is the best of what I've seen so far.

> it really jumps fairly quickly over the basic C stuff at the beginning

In fact, for me it seems to be the best thing about this book. It mentions everything you need to hear once to be able to use search engine and clarify anything you don't understand. So help yourself! You don't know syntax of "switch" statement? Google it (or duckduckgo, or whatever)! You don't know how to use printf? You've been shown the way to cplusplus.com already, take your time and find out everything you want to know about that function. Then come back for a new piece of information to think about.

The only two things I guess are missing is "make" (I'd mention it from the start to help everybody save some time, with gdb and valgrind) and a little bit more information about pointers from the very beginning, because searching for help about "free" function will be no good if you don't know about stack/heap allocations.