Lobster!

[SiliconWizard]

https://strlen.com/lobster/

[Bud]

Not reading not watching clickbites 

[coppercone2]

so its about as useful as zoidberg as a doctor?

[brucehoult]

I don't hate it.

The worst part is the Python-inspired syntax, but lots of people seem to like that and will find it familiar.

It doesn't look to be better feature-wose than Dylan, but Dylan made the mistake of going from an s-expression syntax to a Wirth-style syntax, when what everyone seems to want is a C-style syntax even on wildly different languages such as JavaScript. Maybe Python-style is the current equivalent.

I tend to prefer hygienic macros for syntax extension over raw lambdas (though they often will expand to something containing lambdas), but I don't hate it. I'd still prefer syntax more like Ruby though.

I didn't get far enough to understand whether the "static dispatch if it can" falling back to "dynamic if it has to" method dispatch extends to multiple arguments, so like C++ template functions in the static case, and Dylan / CLOS multi-methods in the dynamic case.

In theory, Julia seems like it should be the modern programming language I would like the best for large-scale programming, but I haven't taken the time to look at it closely.

15-25 years ago I was doing almost all my programming in Dylan but the last few years I've regressed to mostly using assembly language, or pseudo-assembly language in C (not even C++ usually, or at least not recent C++).

[westfw]

In theory, Julia seems like it should be the modern programming language I would like the best for large-scale programming ...15-25 years ago I was doing almost all my programming in Dylan but the last few years I've regressed to mostly using assembly language, or pseudo-assembly language in C

That's an interesting observation.  Sometimes I feel like "modern languages" keep implementing increasingly obfuscated methods of doing things that could have been done a lot more obviously in a simpler language (like assembly or C.)  Sure, it might be slightly more portable (for some task that isn't portable anyway) and safer (or perhaps closer to "provably safer?), but ...


For example, I'm sure it's swell that the C++ STL library is fully polymorphic and throws all the proper exceptions and ... all sorts of other things, but I find its actual implementation code generally less intelligible than an unfamiliar assembly language.

[DiTBho]

OMG

No-one of these new programming languages goes in my direction.
Nobody believes that C can do what I do with my-C.

='(

[brucehoult]

In theory, Julia seems like it should be the modern programming language I would like the best for large-scale programming ...15-25 years ago I was doing almost all my programming in Dylan but the last few years I've regressed to mostly using assembly language, or pseudo-assembly language in C

That's an interesting observation.  Sometimes I feel like "modern languages" keep implementing increasingly obfuscated methods of doing things that could have been done a lot more obviously in a simpler language (like assembly or C.)

It's not entirely about the programming languages so much as the kind of work and kind of projects I'm doing now.

15-25 years ago I was a one-man consultant writing applications of various sorts for local businesses and organizations, from scratch. I was free to choose the most effective tools, and for me that is a dynamic language with multimethods, strong gradual typing, garbage collection, ability to create nice domain-specific-languages and embed them in other code.

More recently I'm working with people designing CPU cores, I'm helping to design new machine code instructions and adding them to cores, emulators, assemblers and compilers. I'm contributing to large existing projects and need to use whatever language they are already written in. Code I write often has to run on everything from microcontrollers to resource-constrained Linux SBCs to servers -- in some cases running relatively old compilers.

[SiliconWizard]

I don't hate it.

Yeah. Doesn't look to bring anything much compared to what already exists though.

The worst part is the Python-inspired syntax, but lots of people seem to like that and will find it familiar.

That could be one reason to hate it.

I tend to prefer hygienic macros for syntax extension over raw lambdas

Ditto. I don't like lambdas anyway.

I didn't get far enough to understand whether the "static dispatch if it can" falling back to "dynamic if it has to" method dispatch extends to multiple arguments, so like C++ template functions in the static case, and Dylan / CLOS multi-methods in the dynamic case.

I didn't find any more info on that in their docs, which are pretty sparse at this point.

In theory, Julia seems like it should be the modern programming language I would like the best for large-scale programming, but I haven't taken the time to look at it closely.

I use Julia for most of the tasks for which I used Matlab before (and similar tools.)

Not sure I see it as a good fit for large-scale programming though.