Does anyone use a general programming language?

[thornekin (Joakin Thorne)]

@Oreolek

so you can’t really expect to run any two-year old source code without major rewrites

Sorry, but no. Imagine if every code base broke every two years.

[Stefan_Hoffmann]

@Oreolek

Generic programming languages are painful to support. Javascript’s nodejs has two major releases per year and its community goes through six hype cycles in the meantime so you can’t really expect to run any two-year old source code without major rewrites.

Well, I would not say that for .NET and C#. The framework is definitely not perfect, but it is very consistent and stable. New components for WebApps and mobile apps were added recently, and I was amazed at how much code I was able to adopt without changes. Definitely not as ghastly as NodeJS. (hey, I’m not getting paid for this post. This is really my opinion)

[rileypb (Phil Riley)]

@thornekin

Sorry, but no. Imagine if every code base broke every two years.

You don’t use anything based on Node, do you? Granted, most code bases are fixable by updating dependencies, but they do tend to degrade quickly.

[Draconis (Daniel Stelzer)]

@Cruel

It reads naturally, sure, but it writes no different than an esoteric scripting language, making you learn all the correct verbage and idioms to use.

That’s the whole point, though! (I know I say this a lot, but that’s because, in my opinion, the documentation about I7 tends to sell it all wrong. So someone should be saying it.)

Code is read much more than it’s written. So I7 optimizes for code readability rather than writability. Since it looks like English, it’s easier to skim (using the part of your brain that’s already used to skimming English text) and get to the part that’s important.

@rileypb

multiple inheritance – why is it so damn hard to have a container-supporter?

Especially when I6 has multiple inheritance. The big issue is that I7 cares a lot about “specificity”—a rule about taking a container is more specific than a rule about taking a thing, so it should be given precedence—and once you have a container-supporter, is the rule about containers or the rule about supporters more specific?

It’s the diamond problem on a vast scale. So while I’d love to see multiple inheritance added to the language, I’m not sure how much of the specificity mechanism would have to be sacrificed for that.

[Draconis (Daniel Stelzer)]

Honestly, though, I’d say that’s a condemnation of Node more than anything else. I can generally run Python code from a decade ago or C code from before I was born, and it works just fine (assuming I can find the right libraries). I can’t imagine needing to update all my code that frequently just to keep up with changes to the language or the standard library.

[inventor200 (Joey "Jess" Tanden)]

@Stefan_Hoffmann

(hey, I’m not getting paid for this post. This is really my opinion)

No, no, C# is really cool. You wouldn’t need to pay me to say this stuff, either. I use Mono C#, though, as I’m on Linux.

[Dannii]

There’s a lot of churn in what’s the current hotness in JS, but it’s really not true that you can’t run old code. I can’t think of any times that’s happened to me.

[Cruel]

Haha, I thought I may have opened a can of worms with my Inform 7 comment. I apologize; I didn’t mean to sidetrack the topic.

@Oreolek

Javascript’s nodejs has two major releases per year and its community goes through six hype cycles in the meantime so you can’t really expect to run any two-year old source code without major rewrites.

Web development is the worst in this regard. It’s largely why I quit it. Two year from now, the industry standard will be whole new libraries, bulky frameworks, convoluted buildng environment, and maybe even new language. Not to mention all the meta shorthand languages that compile into other languages like coffeescript, sass, etc. And to compensate for compatibility, language environments like Ruby and Node have version managers to help you manage all the different versions you need to have installed. It’s triggering PTSD symptoms just talking about it.

@Eleas

For my part, I find it charming. I love how expressive you can get it, as long as you don’t pretend it’s a perfect subset of English.

I can completely agree with that. And thanks for the interesting source example.

@zarf

Don’t make assumptions about what other people like.

I just rarely encounter programmers that are like “yay, yet another language to learn!” And doubly so when it’s something so entirely different as Inform 7. I think that’s the main thing that urked me: that it took longer to grasp than if it were something like Lua, or any other traditional scripting language I’d never used, despite the clear efforts to make it easier to understand/learn. But I already know people love it, it’s a cool project.

To be clear: I’m not suggesting something like Lua would be an improvement. Heaven forbid lol. When dealing with big bodies of interactive text, having a proper interpreted shorthand is inevitably a good investment, even if it requires ditching traditional languages.

[thornekin (Joakin Thorne)]

Actually, I do. Besides you’re contradicting yourself. First, they said applications needed major rewrites, now you are saying applications just need to update dependencies. In any case, ECMAScript rarely breaks backward compatibility. So I don’t know where you’re getting this notion from unless you’re thinking about specific libraries or frameworks, but then that’s a different game.

Anyway, that is off topic I suppose.

[Cruel]

@Draconis

I can generally run Python code from a decade ago

To be fair (though Node hardly deserves fairness), to do this, you will inevitably need both Python 2 and 3 installed, and respective duplicates of libraries needed. That is still somewhat of a pain.

[thornekin (Joakin Thorne)]

I don’t know why I’m playing the advocate here, I’m not a fan by any means, but I think you guys are giving Node too much of a bad time.

[Draconis (Daniel Stelzer)]

@Cruel

I just rarely encounter programmers that are like “yay, yet another language to learn!” And doubly so when it’s something so entirely different as Inform 7.

Around here, I think you’ll find a decent number. Remember that most of us are writing IF just for fun; I learned I7, Dialog, Scottkit, a little bit of ZIL, and some fragments of others just because they’re cool and different.

[rileypb (Phil Riley)]

@thornekin

I don’t know why I’m playing the advocate here, I’m not a fan by any means, but I think you guys are giving Node too much of a bad time.

Perhaps. It’s probably more of a problem of immature web frameworks (which just happen to require node/npm) that change entire paradigms between releases. See AngularJS. Anyway, yes, way off-topic.

And I’ve got to admit there have been a few times when a well-known extension to Inform 7 has turned out to be out-of-date and unusable.

[Eleas (Björn Paulsen)]

This. For the most part, I suspect my ADHD is more of a liability than a benefit to me as a programmer, but I find it helpful when learning esolangs. There’s something about learning new programming paradigms that appeals to me on a visceral level.

Doesn’t hurt that it often yields new approaches to old problems, either.

[blindHunter]

The personification of a TADS library was not something I knew I needed. Adv3 is such a crewel mistress, so I myself have been discovering the joys of adv3Lite.

[johnnywz00 (John Ziegler)]

Curious in what ways you view adv3 as cruel, and so improved in Lite…

[Piergiorgio_d_errico]

concur and second your motion. Blindhunter, why you consider “cruel” adv3 and enjoy adv3Lite ?

Best regards from Italy,
dott. Piergiorgio.

[Piergiorgio_d_errico]

On the point of portability, perhaps I have reached the “event horizon” of C, having successfully compiled with the then-current GCC the oldest surviving C program (cp(1) from 1972 UNIX…), OUT OF BOX albeit I wisely don’t tested the resulting a.out…

it’s a shame that gfortran dropped f77 support, but there’s plenty of F77 and even Fortran IV compiler for emulated environment, and I have around a sizeable set of Fortran II/IV/77 source code in form of .pdf listings from scientific/technical reports from at least late 1950s, and I have successfully “borrowed” many lines from those listings, and even some full sources, some ported to F77 or gfortran…
(side note, is interesting & extremely gratifying watching this machine doing in real-time calculations & computation whose taken hours on the 1960s big irons, even the supercomputers…)

Best regards from Italy,
dott. Piergiorgio.

[borg323]

@Piergiorgio_d_errico

it’s a shame that gfortran dropped f77 support

We are obviously way out of topic here, but do you have a reference for this?

[Piergiorgio_d_errico]

Borg: my mistake. mixed 66 (IV) with 77…

Best regards from Italy,
dott. Piergiorgio.