Building my own Vi text editor in BASIC | Hacker News Reader

Building my own Vi text editor in BASIC

leetusman.com

87 pointszeech49 comments

In BASIC? Why? Why would you do that to yourself?

Don't get me wrong, if he enjoys writing code in BASIC, I'm not going to tell him to stop having fun. It would be silly to tell some random stranger "You're having fun wrong!" (Even if I totally think he's having fun wrong. Grin). If he's having fun with it, go for it.

But man, I looked at the code and got flashbacks. The bad kind. BASIC was the first language I learned when I was a kid, and it's what taught me programming (because after typing "LOAD WIZARD.BAS", I could type "LIST" instead of typing "RUN" and I could actually see what the program was doing. So I learned by reading other people's code. And The Wizard's Castle was pretty good for a BASIC program: it had subroutines, a multi-dimensional map stored in a single-dimensional array (and an actual function defined to convert X,Y,Z coordinates to an index in the array!), and so on. So I am grateful to BASIC for teaching me programming.

And I never, never, NEVER want to write another line of BASIC code again in my life.

But if he enjoys doing so, good for him. I'll just sit here muttering under my breath "But he's still having fun wrong"... :-)

I haven't written BASIC since I was a teenager! The more "modern" basics, without line numbers, were a definite improvement. I went from AppleSoft to AmigaBasic and QuickBasic.

> The more "modern" basics, without line numbers, were a definite improvement.

Never used them, and yet I can unquestioningly agree. The way you had to number each line in increments of 10 so that you could insert a line 15 later in between lines 10 and 20 was, looking back at it, insane.

I can understand why they did it. In the era of no full-screen text editors, where you just typed each line one at a time into a REPL (not that I knew the term REPL at the time), and it stored the lines in the order of their lines numbers... well, that was the only way to edit your code. Made a mistake on line 20? Type a brand-new line 20 and it will replace the old line 20. Want to insert a line between lines 10 and 20? Type in line 15 and it will go and insert. So you could actually load your program, edit your code at the REPL, and save it.

But man, using an actual visual editor is so, so, SO much better than that system. I'm glad it's on the dust heap of programming history where it belongs. It was a decent option for the time when computers had 4K of RAM, but once it was actually possible to edit code in a full-screen editor, line numbers were no longer useful.

The line numbers were also used for GOTO and GOSUB (although later on languages supported strings instead of numbers).

Yeah, forgot to mention that, but that's entirely correct. Modern BASIC variants (I assume) let you label lines so that GOTO and GOSUB can still be used. But with a few flow-control constructs (I don't know modern Basic but in the linked repo I saw a couple `while ... wend` blocks) the need for GOTO is much reduced, I'm sure. GOSUB, well, again I haven't checked. But if GOSUB is allowed to point to a string label rather than a line number then it just becomes a function call, and is still a useful construct. (GOTO only has utility in being able to do the equivalent of `break` or `continue` inside a loop).

A lot of 8 bit basics had a RENUM which would even out the line numbers and repoint all the GOTO statements.

I only use BASIC with line numbers, mostly GW-BASIC and pcbasic. Without numbers it just feels like Lua or python or any other scripting language, but worse? The line numbers BASICs come with their own almost-REPL (IDE?) that I find quite nice (or at least fun) to work in. Maybe mostly nostalgia, but it is the only reason for me to use BASIC at all. I have some basic-mode installed to edit BASIC code in emacs, but I only rarely edit the code outside of its natural built-in line-editor.

There were some advantages. The more advanced BASICs gave you functions / subroutines so you didn't have to mess around with using global variables to pass arguments to a "gosub".

Even with line numbers it's a hell of a lot easier once you get out of the 24 line X 80 character console window and into an editor you can scroll without needing the LIST command.

I had a little crisis at 12 when I went from BASIC to Pascal: how was that supposed to work without line numbers? The statements were just floating around without structure!

I spent two years on hn just to read this comment. So true!

After all these decades, I finally learned that I wasn't the only one who struggled with line numbers from BASIC to Pascal. Thanks, buddy!

Yeah, AmigaBasic especially was when I felt that all the rough edges were sanded off, and it felt like a real language.

Using AmigaBasic or other polished Basics doesn't seem very different than using something like Python.

Too bad Amiga Basic stopped being supported after Workbench 1.3! Supposedly it was Microsoft's only Amiga product.

Laughing agreement. I’m very sentimental for BASIC.

By the standard of getting kids started it’s an amazing and wildly successful language.

By any other standard, it is absolutely terrible! It commits every famous programming sin! Stay away! Do not learn from this language! LOL

    10 PRINT "THIS IS FINE"
    20 GOTO 10

Classic Basic (line numbers, GOTO) resembles 1960s/1970s FORTRAN before it adopted structured programming. The idea when BASIC was invented in Dartmouth in the 1960s that it would be a good introduction to programming for students who might have to learn FORTRAN. The thing is, when BASIC was at its most popular (1980s home computers) FORTRAN itself had moved on from "spaghetti code".

Beginners' All-Purpose Symbolic Instruction Code, if I remember correctly.

I also started programming in BASIC, and moved on to Pascal when I discovered Turbo Pascal around 1983, which also happened to be the language most of my CS classes used a year later when I switched majors from foundering in EE to excelling in CS.

I still have a bunch of my BASIC programs from the day.

"It is practically impossible to teach good programming to students that have had a prior exposure to BASIC: as potential programmers they are mentally mutilated beyond hope of regeneration." -- Edsger Dijkstra

even smart people say dumb things sometimes.

like "goto considered harmful" when all assembly language depends on it

In context though, it was trying to move standards forward.

Look at how acceptable C has changed over the years.

In ye old days it was acceptable to just read in any old text without checking bounds, now that is considered harmful.

There are still use cases for hairier parts of C, but that doesn't mean they should be used all the time. Same with goto.

It's bizarre that someone would think that it's dumb to say "goto considered harmful" because "all assembly language depends on it" ... Dunning-Kruger effect, anyone? People only write in assembler when it's strictly necessary for many good reasons, its lack of structure being one of them. When Dijkstra wrote that, FORTRAN--the major language for scientific programming in the U.S.--had no structured constructs ... `if` statements had a value and 3 numbered branch labels, for value < 0, value = 0, and value > 0. Dijkstra's paper resulted in an explosion of experiments in structured programming, and a huge improvement in programming languages and programmer tools.

Likewise out of context are the complaints about Dijkstra's statement about BASIC ... while it was hyperbole, it wasn't all that far off at the time, when BASIC variables were a single letter, all global, all statements were numbered, there was no structure, not even named functions.

So goto and basic aren't bad, it's just how you use them?

I heard Basic is pretty good as far as embedded script interpreter

> And I never, never, NEVER want to write another line of BASIC code again in my life.

Awww. I downloaded the ECMA-55 standard from 1978 about a year ago and wrote an interpreter in C to be compliant with its "Minimal BASIC". I then had fun typing in code from old computer magazines. So much nostalgic fun.

> Why?

https://justforfunnoreally.dev/

One who has a true mastery of programming should be able to write any program in any language, or at least see how to do so, because one thinks in terms more abstract than language-specific constructs yet is able to map them to any language.

Relatedly, here's TLS 1.3 in VB6: https://news.ycombinator.com/item?id=35882985

What's interesting about it precisely that despite the equivalence (Turing machines and all that) the difficulty to do one thing is a language versus another is what make it prohibitively hard.

If one is "just" translating something that is idiomatic to a language to something that is not in another, it might take 100x for lines for codes in a way that is terribly hard to understand.

I've had the unfortunate exercise of copying an API from pure android java to typescript because "they are similar". It goes against best practices and creates all kinds of weirdness resulting in a finished product which is almost the same but far enough away you could just as well have designed a better API in a few hours from scratch (the real complexity is in the backend of the code, not the interface fortunately). But requirements must be met.

Funnily the conclusion is precisely why the whole idea of "just" giving specs and let <insert here trendiest marketing slang which tries to replace actual experts, developers or not> do the work is risky. If nobody challenges requirements from a manager or client who "just to get it done" you end up with actual work done... that is totally inefficient, potentially dangerous or just irrelevant.

PS: the <> used here could be the next HN thread I opened, https://news.ycombinator.com/item?id=48047826

But you still need to know the language specific constructs.

I can't magically speak German because I know a load of abstract language theory.

And this can work the opposite way. If I know small talk, and then read that C++ can do OO code, I could then think about writing X in C++ and OO being the best model. But I then hit problems when the limitations of OO in C++ become apparent.

In my experience, programming in a language is finding what works in that language, not trying to make the language fit what is in your head.

That's true, but the point the GP was getting at is, if you want to make a program that does xyz, you can do that in any complete language provided you have the mathematical reasoning to figure out how to make it happen. Whether you do that in an object-oriented manner because the language you're using makes that easy, or with some other paradigm.

Programming in a language is finding what works in that language, you're right, but programming at all is trying to get what's in your head into the computer.

I suppose it depends how we define 'true mastery'

Yes you have to understand the problem and state the solution, and that is a prerequisite to 'mastery' but I don't think that's everything that's required.

And this is just talking about the algorithm level. Is Ed equivalent to Emacs? They both do the same thing. Is a complicated solution that demonstrates your mastery better than a simple solution?

Just because you can doesn't mean you should.

awesome project and cv.

speaking of which, I was pleased to see FORTH in there. not that I've ever used it but I was introduced to it in the early 90s and it's cool to see that it's still useful

really enjoying your site content

Not very pro or anti BASIC but very pro your rationale, and love that you do what you do because it is fun and stimulating. I love reinventing wheels also, and, in fact, believe it is under appreciated just how useful it actually is. I'd encourage people to reinvent things for their own benefit, maybe you might end up building a better version of something, but more importantly you will possibly encourage, energise or inspire another to explore and learn.

Great Basic interpreter: https://www.moria.de/~michael/bas/ (it will almost all of the 101 Basic Computer Games if formatted).

On another non-compatible but really small Basic, check NMH BASIC 3:

https://www.t3x.org/nmhbasic/index.html

Compile it under Unix (or windows with w64devkit):

       cc -o tcvm tcvm.c

Run primes.bas:

       tcvm basic primes.bas
       NMH BASIC III - 7599 BYTES FREE
       #3 = primes.bas
       OK
       load #3

Also, as it's T3X0 code, you might be able to port and run under DOS and CP/M. Yes, you read it right. By default I didn't compile it to Unix native 32/64 bit with T3X/0 because it NMH Basic requires a 16 bit machine/interpreter. But if it's cross-compiled to a 16 bit DOS or CP/M, it will run native.

Show up what you can do. Port scoundrel, for instance:

https://codeberg.org/luxferre/scoundrel-ports

I already doing that to JimTCL and it's a piece of cake.

This takes me back. When I was about 12 years old I used Tandy BASIC to build a rudimentary programming editor that had a Multiplan-like interface. It's all part of the fun and challenge of programming: how can you construct something interesting or useful under these constraints?

There's a little BASIC resurgance now. I've been having fun with this https://github.com/ReuvenSwirsky/erlbasic "Erlang BASIC" which implements a mini-computer style time sharing BASIC system in Erlang. There's a beta of it online now, running a more advanced version than the branch on GitHub I’ve been using.

I'll see if I can port this "Vi" to this flavor of BASIC.

Me and the Apple IIc on my desk are ready for it

It’s a great learning project, and there’s nothing like building your own tools.

Never occurred to me how much ms dos batch syntax must have been inspired by basic when I saw the comments (aka remarks) start with “rem”

SEEK HELP.

HELP is not a correct argument to SEEK. You must not have a lot of BASIC experience.

He was trying to exit from Vi.

    ?Illegal function call
    Ok

cant tell if vibecoded or not, function comments seem slightly redundant, either way its a cool idea, cant wait to see the language server written in basic also :D

I vibe coded my own SQL query tui client mostly for myself, when I wanted vim editing mode, I just asked for it, and it just worked for a huge amount of common vim editing motions/actions. I did have to prompt it to add a few more when things when I noticed they were missing. Sort of makes this kind of stuff trivial. However, coding it yourself is a great mental exercise

obv my sarcasm was not detected

Reply on news.ycombinator.com