From a25cb2b0bf26033c9bc778d816618a752d015d99 Mon Sep 17 00:00:00 2001 From: Marshall Lochbaum Date: Tue, 5 Jul 2022 16:46:42 -0400 Subject: Somehow, all the docs have now been edited --- docs/commentary/why.html | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'docs/commentary/why.html') diff --git a/docs/commentary/why.html b/docs/commentary/why.html index 9ab7e0a0..1fe07371 100644 --- a/docs/commentary/why.html +++ b/docs/commentary/why.html @@ -11,9 +11,9 @@

BQN has no intention of being the last word in programming, but could be a practical and elegant tool in your kit—even if only used to inform your use of another language. Give it a try!

Versus APL and J

Here are some more specific comparisons against the two most similar languages to BQN. I'll try to bring up the areas where BQN can be considered worse, but my focus here is definitely on BQN's strong points—I'm not trying to offer an unbiased account.

-

BQN is more like APL, but adopts some of the developments made by J as well. However, it's much simpler than both, with fewer and less overloaded primitives as well as less special syntax (J has fewer syntactic rules, but more special cases handled during execution that I think should have been implemented with syntax).

+

BQN is more like APL, but adopts some of the developments made by J as well. However, it's much simpler than both, with fewer and less overloaded primitives as well as less special syntax (J has fewer syntactic rules, but more special cases handled during execution that I think should have been designed as syntax).

The major differences are listed on the front page ("But it's redesigned…"): based arrays, list notation, context-free grammar and first-class functions, reworked primitives, and dedicated namespace syntax.

-

In addition to these, BQN's block system extends APL dfns with headers, adding some very useful functionality: the header specifies block type and argument names, and also allows for simple pattern matching when used with multiple block bodies.

+

In addition to these, BQN's block system extends APL dfns with headers, adding some very useful functionality: the header specifies block type and argument names, and also allows for some pattern matching when used with multiple block bodies.

Since this section gets into the details, it's worth highlighting stranding, a feature I think of as an obvious improvement but that many BQN newcomers see as an obvious sign that I don't know what I'm doing! My full argument for this decision is here; the two key points are that stranding is a source of ambiguity that can strike at any time, requiring a correction with ⊢ or ], and that typing ‿ is really not hard I promise.

BQN's heavier-weight ⟨⟩ syntax for lists also has its own advantages, because it can be formatted nicely across multiple lines, and also allows functions and modifiers to be used easily as elements. Being able to easily map over a list of functions is surprisingly useful!

BQN has no built-in control structures, which can be quite an adjustment coming from certain styles of APL or J. The control structures page gives some ways to write in a more imperative style, but it's definitely not the same.

@@ -21,7 +21,7 @@

Some factors specific to APL or J are given in the sections below.

APL

@@ -29,10 +29,10 @@

BQN cleans up some awkward syntax left over from when each APL operator was special: the outer product is written Fn⌜ rather than ∘.fn, and reduction Fn´ arr is separated from compress b/arr instead of overloading.

BQN adopts leading axis theory as developed in SHARP APL and applied in A+ and J. With this it can collapse APL pairs such as ⌽⊖ and /⌿ to one primitive each, and remove APL's complicated function axis (such as ⌽[2]) mechanism. The Rank modifier ⎉ then applies these primitives to non-leading axes. While this method is required in J and also favored by many users of Dyalog APL, it definitely doesn't enjoy universal support—it can be harder to learn, and less convenient for some common cases. Summing rows with +/ in APL is quite convenient, and BQN's +˝⎉1, or +˝˘ for matrices, just aren't as nice.

Arguably BQN cuts down the set of primitives too much. Base conversion ⊥⊤, partitioning ⊂⊆, and matrix division ⌹ are commonly asked-for primitives, but they don't match my conception of a primitive. And while each can be implemented (with short snippets, other than ⌹ which requires a library), there's definitely a convenience loss. But there's always ReBQN…

-

BQN's Power modifier ⍟ allows an array operand to specify multiple results, for example Fn⍟(↕4) to get 0 up to 3 iterations. Intermediate results are saved, so the number of calls only depends on the highest iteration number present. On the other hand, BQN has no direct equivalent of Power Limit ⍣≡, requiring it to be implemented manually.

-

An APL selective assignment arr[2 3]+←1 should usually be written with Under in BQN: 1⊸+⌾(2‿3⊸⊏)arr (but the correspondence might not always be so direct). You can think of this as a very fancy At (@) operator, that lets you pull out an arbitrary part of an array.

+

BQN's version of the Power modifier ⍟ allows an array operand to specify multiple results, for example Fn⍟(↕4) to get 0 up to 3 iterations. Intermediate results are saved, so the number of calls only depends on the highest iteration number present. On the other hand, BQN has no direct equivalent of Power Limit ⍣≡, requiring it to be implemented manually.

+

An APL selective assignment arr[2 3]+←1 should usually be written with Under in BQN: 1⊸+⌾(2‿3⊸⊏)arr (but the correspondence might not always be so direct). You can think of this as a very fancy At (@) operator, that lets you pull out an arbitrary part of an array.

Dfns are adjusted in a few ways that make them more useful for general-purpose programming. A BQN block always runs to the last statement, so a block like {Update 𝕩 ⋄ 1+x} won't return early. Writing modification with ↩ makes it clearer which variable's which. Dfns also do a weird shadowing thing where a←1⋄a←2 makes two different variables; in BQN this is an error because the second should use ↩. Tradfns are removed entirely, along with control structures.

-

BQN doesn't have an exact replacement for dfn guards, although the predicate ? can look similar: {2|⍵ : 1+3×⍵ ⋄ ⍵÷2} is equivalent to {2|𝕩 ? 1+3×𝕩 ; 𝕩÷2}. But note that where APL uses the statement separator ⋄, BQN uses the body separator ;. This means that the if-true branch in BQN can consist of multiple statements (including additional predicates), but also that the if-false branch can't access variables defined in or before the condition. In both cases the "better" behavior can be obtained with an extra set of braces and possibly assigning names to arguments ⍵/𝕩. I think guards end up being cleaner when they work, and predicates are more versatile.

+

BQN doesn't have an exact replacement for dfn guards, although the predicate ? can look similar: {2|⍵ : 1+3×⍵ ⋄ ⍵÷2} is equivalent to {2|𝕩 ? 1+3×𝕩 ; 𝕩÷2}. But note that where APL uses the statement separator ⋄, BQN uses the body separator ;. This means that the if-true branch in BQN can consist of multiple statements (including additional predicates), but also that the if-false branch can't access variables defined in or before the condition. In both cases the "better" behavior can be obtained with an extra set of braces and possibly assigning names to arguments ⍵/𝕩. I think guards end up being cleaner when they work, and predicates are more versatile.

BQN's namespaces have a dedicated syntax, are much easier to create than Dyalog namespaces, and have better performance. I use them all the time, and they feel like a natural part of the language.

J

See also the BQN-J dictionary. J is under development again and a moving target. I stopped using it completely shortly after starting work on BQN in 2020, and while I try to keep up to date on language changes, some remarks here might not fit with the experience you'd get starting with J today.

@@ -40,15 +40,15 @@

The biggest difference could be in file loading. If you write a script that depends on other files, and want it to work regardless of the directory it's called from, you need to deal with this. In J, >{:4!:3 '' gives the name of the most recently loaded script (the current one, if you put it before any imports), but to make it into a utility you need this glob of what's-going-on:

cur_script =: {{(4!:3$0) {::~ 4!:4<'y'}}
-

In BQN it's •path. And usually you don't need it because •Import resolves paths relative to the file containing it.

+

In BQN it's •path. And usually you don't need it because •Import resolves paths relative to the file containing it (if you want to use the shell's current directory, you have to use •wdpath explicitly).

J uses numeric codes; BQN uses mostly names. So J's 1&o. is BQN's •math.Sin, and 6!:9 corresponds to BQN's •MonoTime.

J uses bytestrings by default, making Unicode handling a significant difficulty (see u:). BQN strings are lists of codepoints, so you don't have to worry about how they're encoded or fight to avoid splitting up UTF-8 bytes that need to go together.

But J has its type advantages as well. I miss complex number support in BQN, as it's an optional extension that we haven't yet implemented. And BQN has a hard rule that only one numeric type is exposed to the programmer, which means high-precision integers and rationals aren't allowed at all for a float-based implementation. I think this rule is worth it because J's implicit type conversion is hard to predict and an unexpected numeric type can cause sporadic or subtle program errors.

BQN uses a modifier ⟜ for J's hook, adding ⊸ for a reversed version (which I use nearly twice as often). This frees up the 2-train, which is made equivalent to Atop (∘). It's the system Roger Hui came to advocate, since he argued in favor of a hook conjunction here and made 2-train an Atop when he brought it to Dyalog APL. As an example, the J hook (#~0&<:) to remove negative numbers becomes 0⊸≤⊸/ in BQN. Hooks are also the topic of Array Cast episode 14, where the panel points out that in J, adding a verb at the far left of a dyadic train changes the rest of the train from dyadic to monadic or vice-versa, an effect that doesn't happen in BQN.

-

J locales are not first-class values, and BQN namespaces are. I think BQN's namespaces are a lot more convenient to construct, although it is lacking an inheritance mechanism (but J's path system can become confusing quickly). More importantly, BQN namespaces (and closures) are garbage collected. J locales leak unless manually freed by the programmer. More generally, J has no mutable data at all, and to simulate it properly you'd have to write your own tracing garbage collection as the J interpreter doesn't have any. I discussed this issue some in this J forum thread.

+

J locales are not first-class values, and BQN namespaces are. I think BQN's namespaces are a lot more convenient to construct, although it is lacking an inheritance mechanism (but J's path system can become confusing quickly). More importantly, BQN namespaces (and closures) are garbage collected. J locales leak unless manually freed by the programmer. J has no mutable data at all; to simulate it properly you'd have to write your own tracing garbage collector, as the J interpreter doesn't have one.

In J, each function has a built-in rank attribute: for example the ranks of + are 0 0 0. This rank is accessed by the "close" compositions @, &, and &.. Choosing the shorter form for the close compositions—for example @ rather than @:—is often considered a mistake within the J community. And function ranks are unreliable: consider that the ranks of ]@:+, a function that always has the same result as +, are _ _ _. In BQN there aren't any close compositions at all, and no function ranks. J's &.> is simply ¨, and other close compositions, in my opinion, just aren't needed.

-

J has several adverbs (key, prefix, infix, outfix…) to slice up an argument in various ways and apply a verb to those parts. In BQN, I rejected this approach: there are 1-modifiers for basic iteration patterns, and functions such as Group (⊔) that do the slicing but don't apply anything. So </.~a is ⊐⊸⊔a, but fn/.~a is >Fn¨⊐⊸⊔a (I also reject J's implicit merge except for the Rank modifier, as I don't think function results should be homogeneous by default). BQN's approach composes better, and is more predictable from a performance perspective.

-

Gerunds are J's answer to BQN's first-class functions. For example J's (%&2)`(1+3*])@.(2&|) would be written 2⊸|◶⟨÷⟜2,1+3×⊢⟩ with a list of functions. I think lists of functions are a big improvement, since there's no need to convert between gerund and function, and no worries about arrays that just happen to be valid gerunds (worried about losing the ability to construct gerunds? Constructing tacit functions in BQN is much easier). The usability gap widens because passing J functions around either as values or gerunds has presents some highly idiosyncratic challenges, discussed below.

+

J has several adverbs (key, prefix, infix, outfix…) to slice up an argument in various ways and apply a verb to those parts. In BQN, I rejected this approach: there are modifiers for basic iteration patterns, and functions such as Group (⊔) that do the slicing but don't apply anything. So </.~a is ⊐⊸⊔a, but fn/.~a is >Fn¨⊐⊸⊔a (I also reject J's implicit merge except for the Rank modifier, as I don't think function results should be assumed homogeneous by default). BQN's approach composes better, and is more predictable from a performance perspective.

+

Gerunds are J's answer to BQN's first-class functions. For example J's (%&2)`(1+3*])@.(2&|) would be written 2⊸|◶⟨÷⟜2,1+3×⊢⟩ with a list of functions. I think lists of functions are a big improvement, since there's no need to convert between gerund and function, and no worries about arrays that just happen to be valid gerunds (worried about losing the ability to construct gerunds? Constructing tacit functions in BQN is much easier). Unrelated to these fundamental issues, passing J functions around either as values or gerunds presents some idiosyncratic challenges, discussed below.

Named functions

Its impact on the programmer is smaller than a lot of the issues above, but this section describes a behavior that I find pretty hard to justify. What does the identifier fn indicate in a J expression? The value of fn in the current scope, one might suppose. Nope—only if the value is a noun. Let's make it a function.

   fn =: -
-- 
cgit v1.2.3