Language design push

[axch]

I reviewed the ticket tracker in May 2016, and collected all the nits I could find that are reasonable to classify as "language design". The main characteristic these things share is that more or less all of them incur emotional blockage, of the form "It seems like it would be better to do this as part of a coordinated language design effort rather than piecemeal." Perhaps we should have such a push?

Candidate additional procedure names:

• [ ] #547 (builtin filter and reduce, and zipWith)
• [ ] #477 (neq is missing)
• [ ] #437 (for_each/mapM problems) ? (adding mapM_ and forM ?)
• [ ] #374 (! for run), if we want that
• [ ] #274 (sleep)
• [ ] #170 (liftM/fmap)
• [ ] #169 (with for temporary directives)

Others:

• [ ] #569 (~ syntax) (Originally proposed by Vikash/Anthony)
• [ ] #563 (Is a model one estimate or many?)
• [x] #560 letrec syntactic sugar
• [x] #559 (dictionary literal syntax)
• [x] #558 (let_values and syntax)
• [ ] #557 (quote suppressing macro expansion)
• [ ] maybe #556 (language eval SP recognizes)
• [x] #553 ($x syntax)
• [x] #552 (lookup syntax)
• [ ] #551 (sequencing ignoring value syntax)
• [ ] #550 (skip define token at top level ?)
• [ ] #548 (first and rest on arrays)
• [ ] #524 (programmatic construction of names)
• [x] #485 (infix tagged)
• [ ] #481 (return type of bernoulli SP vs make_beta_bernoulli)
• [ ] maybe #473 (collapsed-choice categorical)
• [ ] #441 (each block keyword)
• [x] #438 (do-let syntax)
• [ ] #421 (unique scope keyword)
• [ ] #406 (selection of syntax to parse)
• [ ] #391 (strings print with quotes)
• [ ] #390 (generating labels)
• [ ] maybe #378 (callback and collect usability)
• [ ] whatever #369 (polish multiple model support) is
• [ ] #367 (NonNegativeType along with PositiveType)
• [ ] #352 (literal integers), #436 (problems with integers)
• [ ] #351 (atoms)
• [ ] #312 (what should evaluating a directive return?)
• [ ] maybe #311 (revisit the concept of instructions)
• [ ] #307 (global_log_likelihood ignores constraints that affect no random choices)
• [ ] #306 (ripl get_seed/set_seed are broken)
• [ ] #267 (uniformity in accepting labels)
• [ ] maybe #266 (native mutation in inf prog)
• [ ] #250 (list literals)
• [ ] #237 (keyword arguments)
• [ ] #231 (destructuring observe) (also syntax for it)
• [x] #230 (syntax for destructuring assume)
• [x] #222 (observe sugar)
• [ ] #221 (iteration syntax)?
• [ ] #186 (constrain and unconstrain)
• [ ] #168 (parens for top level define in church')
• [ ] #163 (cond syntactic sugar)
• [x] #162 (parsing x+1 (no whitespace) without confusing it for a literal +1)
• [ ] #132 (sequencing non-actions)
• [ ] #128 (explicit models)
• [ ] #124 (time out)
• [x] #122 (unary minus)
• [ ] #118 (optional arguments to compounds)
• [ ] #117 (variadic compounds)
• [x] #116 (thoughtful do concrete syntax)
• [ ] #111 (documentation of syntax)
• [x] #545 (document that foo = bar; baz parses as let (which is Scheme's let*))
• [ ] #105 (name of likelihood_weight)
• [ ] Are #75 (nested quasiquote) and #76 (macro expansion in quasiquote) part of this?
• [ ] #68 (serialization for the command line)

[fsaad]

How about https://github.com/probcomp/Venturecxx/issues/525

[axch]

Seems my search criteria missed 8 more candidates ^