Terminal Condition

I spend half my time, easily, in a command-line terminal running zsh. So a new one, even one on an OS I don't run, is interesting:

There are some modern, nice conveniences in this. It's a little ways behind Mac Terminal.app (based on the NeXTstep Terminal from 1990), and vastly far behind iTerm2, but it's more advanced than the usual Linux terminals like rxvt, urxvt, or cross-platform Alacritty or Hyper.

Between this and WSL2 being a full Linux, it's plausible that the best Linux dev environment now (well, this summer when it's released) is Windows. The Year of the Linux Desktop is 2019, and it is owned by Microsoft®. Can you hear the tiny, distant screams of the FSF cultists?

Comparison based on code, reviews, and reddit thread with MS devs involved:

  • Scrollback: The single most important thing a terminal can do. MS does this, but doesn't have logging.

    Surprisingly, a lot of them only support a few pages. I keep mine at 10,000 lines or so, which is probably wasteful but so handy; I don't bother logging since my .zhistory keeps everything I typed, and I have Terminal.app and iTerm2 set to not close tabs automatically.

    Alacritty only just added scrollback last year.

  • Prompt Marking: Nope.

    This is a feature it's hard to live without once you've had it, no more paging up trying to see prompt lines (I have a red ANSI-colored prompt and it's still hard to see). In Terminal.app, Edit, Marks, Automatically Mark Prompt Lines, and then ⌘↑ and ⌘↓ move between them. iTerm2 has it enabled by default, and ⇑⌘↑ ⇑⌘↓ are the keys, which took me some re-learning.

    Nothing else has this, as far as I've seen.

  • Fonts: MS has programming ligatures and displays emoji, finally. Does not support RTL languages.

    I use Fira Code in all my editors and shells, and it's enormously helpful, more readable, and catches bugs: I look for === as a fat-equals symbol in JS, etc.

    Hyper, urxvt, Alacritty support Unicode fonts. rxvt stopped development almost 20 years ago so it barely shows 8-bit fonts correctly.

  • Tabs: MS has tabs! They're currently invisible until you add a second tab, same shit Terminal.app does, which annoys the hell out of me; I don't like UI that reshapes itself, reminds me of T-1000 Terminators (also makes it hard to tile my windows up correctly when they get resized).

    It's not clear if you can drag Windows Terminal tabs around to different windows.

    In iTerm2, I normally keep: First window with tabs for home shell, ssh into my server (running screen, so that has many open shells). Second window with 2 tabs for REPL, editor/running/compiling tasks, and sometimes a third tab for reading local docs. If I need more shells, I usually open them on the first window. I rarely open a third window for monitoring some long-running task; I just drag a tab out to its own window. All terminal windows are stacked on the left side of my screen, because there's no icons under that side of the Desktop.

    urxvt has tabs, but they're kind of a hack, not fully usable.

    Hyper has tabs, but they replace the title bar. Which is cool but also awful like a lot of things it does.

    rxvt and Alacritty don't do tabs, because they insist you use screen or tmux. Which sucks if you want to move a process from one window to another.

  • Profiles: MS supports multiple profiles, so you can use different ones for each task.

    So does Terminal.app, iTerm2, urxvt (but it's buried in a text file config).

    Alacritty, rxvt, and Hyper have a single profile and no UI for changing anything, hope you like editing text files and reloading.

    As far as I can tell, nothing else does automatic profile switching like iTerm2; when I cd to my ~/Code/CodeScheme folder, iTerm2 switches to my dark red transparent profile, including Scheme-specific triggers and copy/paste filtering.

    You can probably do that in urxvt's Perl(!) scripting, but it's not normal or easy.

  • Copy/Paste Filtering: Nope.

    iTerm2 and urxvt both let you set a bunch of regexp to run over lines to get selections correctly matching boundaries, not just space-delimited.

  • URL Highlighting: Nope.

    iTerm2, Hyper, and urxvt notice URLs and filenames, and let you click on them. In iTerm2, hold down ⌘ and click on any URL or path (like in an ls or find result!) and it does some useful action: Opens the URL in your browser or file path in your editor, by default, but you can configure that in the profile.

  • Custom Keybindings: Sorta? Doesn't seem complete, no idea if there's UI for it, but it does exist in their config.

    Most terminals can do this, but most can only remap a few actions. I like iTerm2's, as usual, which lets you bind any action, menu, or run a program on any keybinding. I mostly just use it to launch different profiles with starting paths & scripts.

    Terminal.app only lets you send specific text for a key.

  • Images: Sorta? Only if they're embedded in fonts.

    This is a neat trick in iTerm2: images. I use imgls all the time to see a thumbnail of every file with details (protip: I changed ls -ld in the script to ls -1Fskd for a more concise listing), and then ⌘-click to open what I want in Acorn; it's better than opening Finder and trying to read a long filename under a thumbnail.

    I'm unaware of anyone else being able to do this.

Tower of Babble

Programmers almost compulsively make new languages; within just a few years of there being computers, multiple competing languages appeared:

It proliferated from there into millions; probably half of all programmers with 10+ years of experience have written one or more.

I've written several, as scripting systems or toys. I really liked my Minimal script in Hephaestus 1.0, which was like BASIC+LISP, but implemented as it was in Java the performance was shitty and I had better options to replace it. My XML game schemas in GameScroll and Aiee! were half programmer humor, but very usable if you had a good XML editor. Multiple apps have shipped with my tiny lisp interpreter Aspic, despite the fruit company's ban on such things at the time. A Brainfuck/FORTH-like Stream, working-but-incomplete tbasic, and a couple PILOT variants (I think PILOT is hilariously on the border of "almost useful").

Almost every new language is invented as marketing bullshit based on a few Ur-languages:

  • C++: Swift
  • Java: Javascript (sorta), C#, Go
  • Awk: Perl, Python, PHP, Julia
  • C: Rust
  • Smalltalk: Objective-C
  • Prolog: Erlang, Elixir
  • ALGOL: C, Pascal, PL/1, Simula, Smalltalk, Java
  • LISP: Scheme, ML, Haskell, Clojure, Racket
  • BASIC: None, other than more dialects of BASIC.
  • FORTRAN: None in decades, but is the direct ancestor of ALGOL & BASIC.
  • COBOL: None in decades.

A few of these improve on their ancestors in some useful way, often performance is better, but most do nothing new; it's plausible that ALGOL 68 is a better language than any of its descendants, it just has mediocre compiler support these days.

Certainly I've made it clear I think Swift is a major regression, less capable, stable, fast, or even readable than C++, a feat I would've called impossible except as a practical joke a decade ago. When Marzipan comes out, I'll be able to rebuild all my 15 years of Objective-C code and it'll work on 2 platforms. The Swift 1.0 app I wrote and painfully ported to 2.0 is dead as a doornail, and current Swift apps will be uncompilable in 1-2 years; and be lost when Apple abandons Swift.

When I want to move my Scheme code to a new version or any other Scheme, it's pretty simple, I made only a handful of changes other than library importing from MIT Scheme to Chez to Chicken 4 to Chicken 5. When I tested it in Racket (which I won't be using) I had to make a handful of aliases. Probably even CLISP (which is the Swift of LISPs, except it fossilized in 1994) would be 20 or 30 aliases; their broken do iterator would be hard but the rest is just naming.

Javascript is a pernicious Herpes-virus-like infection of browsers and desktops, and nothing can ever kill it, so where it fits the problem, there's no reason not to use it. But there's a lot it doesn't do well.

I was leery of using FreePascal because it has a single implementation (technically Delphi still exists, but it's $X,000 per seat on Windows) and minimal libraries, and in fact when it broke on OS X Mojave, I was disappointed but I-told-you-so.

I'm not saying we should quit making new Brainfuck and LOLCODE things, I don't think it's possible for programmers to stop without radical brain surgery. But when you're evaluating a language for a real-world problem, try moving backwards until you find the oldest and most stable thing that works and will continue to work, not piling more crap into a rickety new framework.

The Biblical reference in the title amuses me, because we know now that it requires no malevolent genocidal war deity scared of us invading Heaven to magically confuse our languages and make us work at cross purposes; anyone who can write and think splinters their thought into a unique language and then argues about it.

Scheme Record to RecordType

So, in "classic" Scheme (up to R5RS), there were no structs/records/classes. You could fake them with a Vector and writing all the accessor methods by hand, but it sucked.

SRFI-9 added a very minimalist record type with a lot of repetition, and no inheritance, though at least SRFI-17 fixed setters.

R6RS (as implemented in Chez Scheme) added a much more powerful system with inheritance and constructor methods, but half the Scheme community hates nice things and voted it down. There's a half-assed reimplementation of R6RS in SRFI-99 and sequels, but it still doesn't have read/write mechanisms. R7RS still only ships with SRFI-9 built in. Unbelievable mess.

Chicken has a convenient define-record macro, and read/write methods, but by default uses SRFI-9, and hides SRFI-99 in an egg; so chicken-install -s srfi-99 and then (import srfi-99) everywhere, and then write a ton of boilerplate for every type. So I just automated it with Python (doing string parsing in Scheme is more annoying):

Documentation is in the module help (or just read the source, Luke). I use it by writing the Chicken macro (define-record Point x y), then at shell:

% pbpaste|schemeRecordToRecordType.py|pbcopy

And paste back:

;; (define-record Point x y)
(define-record-type Point  
    (x)
    (y)
)
(define-reader-ctor 'Point make-Point)
(define-record-printer (Point p out)
    (format out "#,(Point ~S ~S)"
        (Point-x p) (Point-y p) 
))
; module exports:
; make-Point Point? Point-x Point-y

Note that none of this really gets me an "object-oriented" system, but records are sufficient for most programs, and inheritance works with define-record-type. There are OOP systems but I don't especially like any of them so I'm passing on them for now.

Writing Objective-C with Mulle-Objc

mkdir CLICalc
cd CLICalc
mulle-sde init -m foundation/objc-developer executable

This takes more or less forever.

… Still going. OK, finally done. I hate to think it's gonna do that every new project? Or whenever it updates?

Anyway, bbedit . (fuck Xcode), and add at the bottom of import.h and import-private.h:

#import <Foundation/Foundation.h>

Make src/main.m useful:

// main.m
 "import-private.h"
 "CLICalc.h"

int main(int argc, char *argv[]) {
    NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];

    CLICalc *calc = [[CLICalc alloc] init];
    [calc push:42.0];
    [calc push:69.0];
    double a = [calc pop];
    double b = [calc pop];
    NSLog(@"a=%f, b=%f", a, b);

    [pool release];
    return 0;
}

Create an Objective-C class, src/CLICalc.h:

// CLICalc.h
 "import-private.h"

@interface CLICalc : NSObject

@property (retain) NSMutableArray *stack;

- (void)push:(double)n;
- (double)pop;

@end

and src/CLICalc.m:

// CLICalc.m
 "CLICalc.h"

@implementation CLICalc

@synthesize stack = _stack;

- (id)init {
    self = [super init];
    _stack = [[NSMutableArray alloc] init];
    return self;
}

- (void)dealloc {
    NSLog(@"CLICalc dealloc");
    [_stack release];
    [super dealloc];
}

- (void)push:(double)n {
    [_stack addObject:@(n)];
}

- (double)pop {
    if ( ! [_stack count]) {
        // ERROR: stack underflow
        return 0.0;
    }
    double n = [[_stack lastObject] doubleValue];
    [_stack removeLastObject];
    return n;
}

@end

Doing that without a template was a little hard on the old memory, and I had to use DDG to look up some method names without autocompletion. But I'm pretty sure that's fine.

In mulle-ide, type update to add the new class to cmake: If you look in cmake/_Sources.cmake you should now see CLICalc.m listed.

Now craft to compile. You'll get a spew of crap, but hopefully no errors.

I am getting this, which I can't resolve:

/Users/mdh/Code/CodeMac/CLICalc/src/main.m:21:55: warning: 'NSAutoreleasePool'
      may not respond to 'init'
        NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
                                   ~~~~~~~~~~~~~~~~~~~~~~~~~ ^
1 warning generated.

But NSAutoreleasePool certainly has init, and it seems to not die?

% ./build/Debug/CLICalc
a=69.000000, b=42.000000

Hooray!

Yeah, this isn't amazing. Except: It's supposedly portable now. I can maybe rebuild this on Linux, or Windows? I dunno.

This is almost classic Objective-C, slightly enhanced from 1.0: We didn't have property/synthesize, or nice object wrappers like @() when I were a lad. I typed so many [NSNumber numberWithInteger:n]. So get used to the retain/release/autorelease dance. There's no dot-syntax for property access, type them [] like old-school. But hey, it's a proper compiled language with a nice object system and no GC pausing.

I tried importing Cocoa and got a ludicrous spew of errors, so Mac GUI is gonna be a challenge. But I could import SDL and use that for portable UI, since Objective-C is just C.

Sweet. I'll finish up the calculator's parser in a bit, but then see about doing something useful in it.

Interactive Python IDLE

When using Python from a shell, the REPL is fairly awful and doesn't let you copy-paste or save, except through the shell itself. I especially find that copy-pasting functions in is error-prone. There's a nice interactive environment for Python called IDLE (after Eric ): It's probably an application in your /Applications/Python 3.7 folder, or whatever other OS's do, or can be run with IDLE3 from the shell. Other than using ^N/^P for next/previous history line, it works pretty much exactly as you'd expect a GUI REPL to work, and lets you save your session as a text file, easy to extract some functions from later or use as a doctest.

Trouble is, IDLE doesn't automatically pick up the ~/.pystartup script; I had to remember to call it with IDLE3 -s, and there's no easy way to do that from the desktop, where I'm often lazily clicking. This has been frustrating me very slightly for years.

So: open Automator, new Document, add Utilities/Run Shell Script, and paste in:

IDLE3 -s

Save as an Application, name it IdleStart. The icon's ugly and generic, so I made a half-assed icon, copy it from an image editor or Preview, and paste into the icon in Get Info on the application.

python-foot

Now I have a nice stupid foot to click on, and get a proper REPL. The running program has a hideous 16x16 or 32x32 icon scaled up, which I don't think I can easily solve; I looked at the idlelib source and while I could patch it, it's not easily configured. Maybe later. There's also no way to specify the window location, which I'd like to have mid-left of my screen, but again, maybe later.

While I'm at it, the themes are garish, and the customizer is unpleasant. So I just edited this in ~/.idlerc/config-highlight.cfg, then picked Mark's Dark theme:

[Mark's Dark]
normal-foreground = 
normal-background = 
keyword-foreground = 
keyword-background = 
builtin-foreground = 
builtin-background = 
comment-foreground = 
comment-background = 
string-foreground = 
string-background = 
definition-foreground = 
definition-background = 
hilite-foreground = 
hilite-background = 
break-foreground = 
break-background = 
hit-foreground = 
hit-background = 
error-foreground = 
error-background = 
cursor-foreground = 
stdout-foreground = 
stdout-background = 
stderr-foreground = 
stderr-background = 
console-foreground = 
console-background = 
context-foreground = 
context-background = 

My current ~/.pystartup is short (in python2 it was much longer, since nothing worked right), just some common imports and functions I use enough to miss:

import os, sys
import math
import random as R
import re
import time
import turtle as T

def dice(n, s):
    t = 0
    for i in range(n):
        t += R.randint(1, s)
    return t

def roundup(n):
    return math.floor(n+0.5)

def sign(i):
    if i > 0: return 1
    elif i == 0: return 0
    return -1

print("READY")

Now if I click the foot and see "READY" above >>>, it's all working.

Debugging in C

I've spent a horribly long time tonight staring at nested stack manipulation code now to get algebraic expression parsing (mostly?) working. I hate unary minus; life is pretty good except for that ugly little weiner with its binary operator twin, and then BAM weird compromises in your code.

For most errors, I rely on testing (even if just firing a test script through the language, as I'm doing with tbasic) and debug mode with verbose stderr logging. But this is C, where the slightest mistake can be EXC_BAD_ACCESS with no clue where. So then I need a debugger…

% make && lldb -o run -- tbasic -d test1.bas

[update: Forgot the -- before the program, which prevents lldb from reading those params. Command lines without parens are easy to get wrong!]

I don't really do much serious with lldb, I just need to see where an error occurred, backtrace (bt), and sometimes print some variables, to usually be able to solve a crash. It's a little frustrating that the lldb environment is so primitive, though, doesn't even have stdout, stderr (weirdly \<stdio.h> is callable), so how do I call utility functions? Had to rewrite some functions to take default NULL values.

Anyway, let and print (and error, my idiosyncratic stderr print) work, there's not a lot left in BASIC then I can get back to more serious things.

Programming in C

On tbasic, I've been doing all my C coding in BBEdit, not fucking Xcode, and it is fantastic. Doesn't crash. Syntax highlighting works, and by "works" I mean doesn't replace my text with Cyrillic as Xcode is wont to do (I do not like the new BBEdit color theme editor, but it's a far cry from stabbing me in the face like Xcode does). BBedit's window stays where I fucking put it, and sidebar shows clearly which files are open and modified. Running make from iTerm2 works fine, if you aren't an idiot and each compile produces less than a handful of errors. I can't really use BBEdit for JavaScript which needs more tool support, but for simpler languages, it's fine.

OS X Mojave no longer has C man pages visible anywhere I can find, so Dash is the only way to look anything up:

To use it from the shell, create dashman: (hashtag command-line integration, I couldn't find this in any search, and Dash has no AppleScript which is my usual solution to o'erweening GUIs)

#!/bin/zsh
open "dash://$*"

Hm. So, I've worked with people who don't learn their languages, they just rely on autocomplete in an IDE, snippets, and StackOverflow. If this is you, if you can't code without an Internet connection, you can't code. Please stop programming, go away, and read a book until you know the syntax and fundamental APIs, because right now you do more harm than good.

That said, while I studied K&R (and Stephen Kochan's Programming in C, my introduction back in the '80s) with the intensity of a snake-handler reading his Bible, I certainly can't remember every strcspn, strcoll, strstrn or whatever random series of 7-letter identifiers they had to use back in the '70s (even in the late '80s, I was still using C compilers which only distinguished 7-letter identifiers). C's libraries are often gibberish and searchable man pages are all we have.

Chicken Soup

(a bunch of stuff in a pot)

REPL

The Chicken csi REPL is appalling after using some nice REPLs, it doesn't even have history by default. I couldn't reliably get non-GNU readline-likes to work, so:

% chicken-install -sudo readline
% cat >~/.csirc
(use readline)
(current-input-port (make-readline-port))
(install-history-file #f "/.csi_history")
^D

So at least now it has the usual up/down/emacs-like keys.

Long fucking ways from the old Symbolics LISP Machines. Why don't we have environments like that anymore? Why is everyone content to just use fucking emacs (I've never been emacsulated) or other editor, and a boring REPL? DrRacket is just a REPL that destroys its memory every time you edit code, and it's the most graphically advanced LISP-type environment. And this is why I still just use Atom with Symbols Tree View (even though it thinks variable definitions are functions), and copy-paste into iTerm if I want to test something.

Value Unpacking

Not having nice R6RS macros for this, and unwilling to fight with classic macros, I've been using values to unpack lists into variables, and because I can never remember the exact syntax, I made this cheat-sheet:

(define a '(1 2 3))
(define b '(4 5 6))
;; then one of these:
(define-values (x y z) (apply values a)) (printf "~s,~s,~s\n" x y z)
(set!-values (x y z) (apply values b)) (printf "~s,~s,~s\n" x y z)
(let-values [[(x y z) (apply values a)] [(q r s) (apply values b)]] (printf "qrs:~s,~s,~s xyz:~s,~s,~s\n" q r s x y z))

Probably not efficient, but better than car, cadr, caddr, etc. Maybe I should move all my list-structures into vectors, but then I'd still have to convert them to lists half the time. Here's where Python is the programmer's best friend, even if it is 10,000x slower:

a = (1, 2, 3)
x, y, z = a
print(f"{x},{y},{z}")

Why Did LISP Fail?

How did a more advanced language with better tools just die off commercially, and now if you want to work in it, you have to cobble together a bunch of half-broken shit?

I think there's 3 reasons:

  1. It's hard and ugly. It may be logically compelling, but when you see a page of parens your brain panics and looks for a place to hide.
  2. Companies value the fake productivity of thousands of lines of C, Java, or Swift (aka C++2020) code more than having safety, security, and correct reasoning. Who cares if millions of people will suffer and possibly die from your code, as long as you can ship TODAY?
  3. A lot of LISP "hackers" are insufferable douchebags, both old beardy fuckers who've been doing it for 50 years and mewling children who learned it last week. Every new variant makes the older contingent more angry at even seeing a mention of it, and the sneering fetuses think whatever variant they learned is Divine Wisdom, rather than just an engineering tool that may need to be improved.

Text Filter

  1. I need a utility to filter some text, like awk but more modern regexp.
  2. Start writing utility.
  3. Discover I wrote this exact utility in 2006, and forgot about it. Apparently I did something with it in 2009. I may have blogged about it on my old site, too lazy to go searching.
  4. Clean it up for Python 3.7 and release.

Utility/Filter 1.1 can be downloaded, put it in your ~/bin or /usr/local/bin folder and call it with Filter.py -h.

BSD license, so do what thou wilt with it, but don't be a dick (uploading my stuff to github as if it's yours is dickish), OK?