Free Programming Language(s)?

[whartung]

I tried to understand VAX assembler once. It didn't go well. :-/ It makes 68000 assembler, with its memory paging and all, look simple.

68K didn't have any real paging issues, it wasn't segmented like the x86. 68K is a direct descendant of the PDP-11, in that they're both essentially orthogonal in how they use general purpose registers and memory. The VAX was the same way, just with large word sizes.

I get a laugh any time someone claims I'm a [system] fan.

I'm a Mac fan for all the reasons Mac fans say they're Mac fans, because, even in their imperfection, the day to day existence and experience they offer are head and shoulders above messing with a Windows XXX machine, and the vile, petri dish online environment they live in.

You know what's scary? I saw an ad for Visual Object COBOL. This implies that someone, somewhere, is writing new programs in COBOL, not just maintaining legacy stuff. And that there are COBOL programmers. COBOL programmers who want brand-new development environments. With COBOL.

That's no more scary than modern FORTRAN, most recently updated in 2010, and scheduled for a minor update in 2018.

There's are zillions of line of code written in COBOL, and rewriting working code for the sake of rewriting it, despite being popular with developers "because they can", is the highest of folly.

What is anyone doing in COBOL that they can't do in something ... well, that isn't COBOL? Square peg, round hole, wham, wham, wham!

They're re-using their long history of working COBOL code, is what they're doing.

Would someone doing a greenfield application today use COBOL? Unlikely, but that's a different issue.

Ah, just thought of another language I crashed head-on into, though I never tried to write a random anything generator in it (possibly the only language this is true of): LISP. I think I did once write a LISP program that could output "Hello, World" but the mentality you have to have to be a LISP programmer and the mentality of a FORTRAN programmer (or procedural languages in general) are completely opposite. I'm just not one of the people who's smart enough to doublethink both at once. Or LISP at all.

You can write perfectly acceptable FORTRAN in Lisp. You can write perfectly fine procedural style code in Lisp. You can write functional style code in Lisp, and you can write OOP style code in Lisp (though the default object system is quite different from the class based systems most are used to, but, you know, you can always write your own object system).

I have the HG combat system written up in Common Lisp right now (I need to add damage allocation, the rest works). It doesn't have one lambda function or even any macros, and only one function that does any particularly extensive list processing.

You know what's depressing? Windows doesn't come with any flavor of programming language, and hasn't for years. That's just sad. If you're going to write something, you have to take the additional step of going off and looking for a suitable interpreter or compiler, possibly buying one, installing it, and so on ... not just popping up your BASIC interpreter and writing a bit of code to do whatever you need done. I write all my utility stuff (ex., something to convert old table files to account for the 21st century changes in TBL) in Delphi now, because it's what I have, but you have to be the kind of person who has a compiler already, not just any random computer user who wants something done. I think we've lost a lot there.

Windows comes with Powershell, which is crazy capable. You also have Javascript in all of the browsers. Notepad and a browser is very powerful today.

Downloading a development environment today is straightforward if you're so motivated.

 

[aramis]

There is little that can be done in any language that cannot be done in a given other language by coders competent in that language. It's often easier to use the correct language optimized for the task, but...

One big issue, however, is that it's not easy (and often not practical) to rewrite from scratch to maintain the combination of function, look, and feel. And that combination is often what people want when the developer recodes. And that's the very thing that recoding does poorly.

Hence, lots of legacy COBOL, SNOBOL, ForTran, and Basic, recompiled, tweaked a bit, but still running...

 

[JimMarn]

My senior year I had to write an assembler, input pseudo assembler in VAX Pascal, and take the output and run it through the assembler... we groaned and moaned over that. Then next semester we wrote a compiler. Each of those was around 3,000 lines of code. At least we only did literal and integers. The professor said we didn't have time to do reals.

Oh, thanks for reviving those memories... :-(

And my relatives wondered why I changed to computer repair.