Talk:Extended Brainfuck
Storage
One suggestion I have to the author of this programming language is to change "storage" to "register" throughout the documentation. This is the terminology used by most microcontrollers, and in my opinion such a change would make the language documentation easier to understand. Other than that, great language. Javawizard 07:31, 23 November 2009 (UTC)
Type 3???
What happened to hello world in type 3?--TehZ 10:11, 16 March 2010 (UTC)
- Someone removed it, after modifying the language, so presumably it no longer worked in the current version. --Ørjan 00:24, 17 March 2010 (UTC)
Impossible to compile?
Is it actualy impossible? I'm assuming it's harder because of the source-modifying capability but that problem has been solved in languages like Befunge so why wouldn't it be possible for brainfuck aswell? --(this comment by 46.44.180.22 at 08:34, 14 March 2013 UTC; please sign your comments with ~~~~)
- There are no 'true' Befunge compilers: the existing ones all cheat in some way. Compiling languages with self-modifying source in the general case requires a level of static analysis that ends up running afoul of the halting problem, which is why it's referred to as impossible. Phantom Hoover (talk) 12:46, 14 March 2013 (UTC)
The page should not be mentioning it helps improving speed of execution
While some of these instructions are actually helpful to make faster programs (I'm thinking about the storage most notably), some are ridiculously slow on most imaginable implementations (locking cells and shifting memory is heavy) and some design choices make it nearly impossible to work with VMs / JITs (in such cases, it's a nightmare, as the language goals part says - in those cases, it either has to drop self-modification either needs to embed a (partial) recompiler or interpreter) or other optimized interpreters especially to support the self-modifying ability. Keeping retro-compatibility with brainfuck on such an interpreter is also very, very tedious, to keep a good level of performance, and this applies between Extended Types themselves. AsuMagic (talk) 21:40, 23 July 2016 (UTC)
Why use a new symbol @ to separate program from input?
In my BF interpreter at https://gist.github.com/tromp/86b3184f852f65bfb814e3ab0987d861#brainfuck I use the existing ] symbol for that, which is distinguishable from all others in that it lacks a matching [.