I was one of the coauthors of the PHP Property Hook (accessor) and asymmetric visibility features. I'd be happy to talk with you about it on the show some time, if you're interested!
Changes and discussions are super active right now, kinda hard to provide a single compiled summary. Roc's zulip is the best place to keep up with everything.
Looks like youtube didn't like my longer comment on this but basically: C++26 will probably include basic versions of reflection, contracts and executors but they're starting to cut it close. C++26 will also almost definitely include a basic version of a highly controversial and possibly unimplementable in its current state feature called "safety profiles" that is currently getting pushed through the standardisation committee at very high speeds at the cost of a lot of other things in general leading to a lot of chaos in the C++26 space at the moment.
Look up Izzy Muerte's blog post about this (On "safe" C++) if you want some more info on a few of the issues that have crept up in the c++ committee within the last few years (though as a warning it is very highly opinionated and very long)
@@RustIsWinning Dude, how does it feel to know that Zig will replace Rust as the language of the future and you have to cry in every comment? Zig is the language of high optimization. Things that Rust can't achieve even with unsafe are trivially easy in Zig. Did you see what happened with Nvidia? All that money is burned when people don't focus on maximum performance, that levels of optimization that companies need in the AI race will never come from an virgin memory management system like borrow chequer. Go, Cry! 😎
7:15 I believe std::execution (a.k.a senders) is on track for C++26. Reflection and Contracts are also aiming C++26, but I think both are “MVP” in the sense that both might seem rudimentary in C++26 and expected to grow on C++29. Most notably I don’t know if “token streams” will make it for C++26.
Yeah, the way C++ standardization works now is that if a feature is ready in time for publication then it'll be included and if it isn't then it'll have to wait three years for the next version of the standard. Reflection and contracts seem quite likely to be included (especially contracts which had previously been planned for C++23) but there's always a chance one or both will be nixed at the last minute.
My biggest highlight about Nim in 2024 is how much they improved tooling. It's still not perfect, but a lot of work was done on LSP (nimsuggest), VS-code plugin and package manager (nimble). And for 2025, there's a roadmap with new intermediate language that will get us easier interop with C, LLVM, Wasm and incremental compilation. I'm beyond excited for more Nim in new year!
Until incremental compilation gets to be a thing I don't see Nim as bringing enough of a value proposition compared to other low level languages in terms of developer productivity.
@@encapsulatio Nim compiler is already fast, IC will be good for repl though. Also I mainly use Nim as a high-level language. Think of Python, but typed, fast and with small executables.
Nothing about Raku? I can't really point out anything in particular, but it does have monthly-ish updates :) I find it to be a really fun language to code in when speed isn't of the essence. It can express so many crazy things.
Why don't you popularize it if you like it so much? People like to praise Raku yet they don't ACTUALLY do anything about popularizing it on youtube, showing comparisons in improved developer productivity compared to Python or Typescript/Javascript for different language features or mini features unique to that language alone.
It's such a shame that Nim has not been used for any big project. It is the perfect general purpose lang without memory management. Regarding the C alternatives, I think Odin is simpler to learn and write than Zig, even when the latter has an amazing performance.
But you can still find F# on Languish. Here it is with some nearby languages. tjpalmer.github.io/languish/#y=mean&weights=issues%3D0%26pulls%3D0%26stars%3D1%26soQuestions%3D1&names=f%23%2Ctypst%2Cwebassembly
Do you think it’s fair to separate typescript in JavaScript when typescript is just a superset of JavaScript? It’s not like your average type script programmer is very different from your average JavaScript programmer.
I like seeing both for now, and it's fairly easy to imagine the combo. Maybe someday I'll just combine them. If I keep track of Stack Overflow tag overlaps, doing custom combos might work better, though.
@ Same answer as @contextfree. But we can go that route if you want to. I believe that AI work isn't the same as "programming", I see it as something more related to math than software development. There is also the fact that most of all AI heavy work is done by C/C++ that are just exposed as APIs inside python, because python is an easy language, so one could argue that AI stuff in Python doesn't actually count as software development.
2025 is the year I match my Odin and Swift knowledge with C. If C++ comes to my mind, I will likely use it only when I know what I am doing lol. And that probably means learning an older subset. Some things about the data, most languages are pretty stable, so I think that implies to me the language market is pretty competitive, no? More languages seem to be kind of just keeping uo rather then rising/falling. The ones that are rising are usually consequential to something else in the ecosystem like how Javascript is being eaten by Typescript. My only thing is how many systems languages can the market sustain? As many as godbolt can host, apparently lol 😂. Also ones that have out of the box C/C++ support, it seems.
On stats, Languish even focuses just on quarterly activity, not sum totals across time. Over the past 15 years or so, there have been some new big languages (such as Dart, Go, Kotlin, Rust, Swift, & TypeScript), but it does take a while, and I don't know how to predict the future.
This covers several languages I'm interested in, as well as many I haven't even heard of. 🙂 One language that I'm interested in which you didn't dive into in depth is Mojo, although I do notice it in the chart. Very nice chart, btw. It's fun to play around with that web page.
@@contextfree hehe I'm also excited about it :) Feb 2025 is the scheduled date for the next zig update which will probably contain the first implementation of the incremental compilation. Exciting stuff.
Take this advice after watch the video: "Languages in 20XX" is like "Ice cream flavors in 20XX". You can do what you want to, in any language. You can make your own back-end using Pistache and C++, creating HTML reponse by creating strings with tags. Will it be insane harder instead of using Node, Bun, etc? Yep, but you can. Can you build a big-large modularized CRM for Windows, using Dart instead of WinForms? Yep, but you can. Languages are designed since the begining to bring solution but languages have a first-vision or mission, so that why his put focus for specific fields or problems because are inherently better suited for these problems. If "your" language are trying to do 'all', is a red flag (Don't bite off more than you can chew). Many languages like ECMAScript (JavaScript for kids) are constantly evolving and introducing new tools and feature to make easier what just are easier. However, if you are looking for a new language based on fashiontards tredings from reddit, maybe you need to looking for a new profesion.
Such a great yearly appointment. Thanks for the video!
I was one of the coauthors of the PHP Property Hook (accessor) and asymmetric visibility features. I'd be happy to talk with you about it on the show some time, if you're interested!
There will be some BIG things coming to Roc soon. Hope someone like you would like to cover that when it happens.
Got some links to discussions? I haven't kept up well nor been in touch with Richard for quite some while.
Changes and discussions are super active right now, kinda hard to provide a single compiled summary. Roc's zulip is the best place to keep up with everything.
I thought I was following Roc pretty well but I hadn't any news
Looks like youtube didn't like my longer comment on this but basically:
C++26 will probably include basic versions of reflection, contracts and executors but they're starting to cut it close.
C++26 will also almost definitely include a basic version of a highly controversial and possibly unimplementable in its current state feature called "safety profiles" that is currently getting pushed through the standardisation committee at very high speeds at the cost of a lot of other things in general leading to a lot of chaos in the C++26 space at the moment.
Look up Izzy Muerte's blog post about this (On "safe" C++) if you want some more info on a few of the issues that have crept up in the c++ committee within the last few years (though as a warning it is very highly opinionated and very long)
Honored to see Plangs! featured! Cheers to an incredible 2025 for programming languages!
Don't sleep on Zig!!
Zig could never 😂
@@RustIsWinning
Dude, how does it feel to know that Zig will replace Rust as the language of the future and you have to cry in every comment? Zig is the language of high optimization. Things that Rust can't achieve even with unsafe are trivially easy in Zig. Did you see what happened with Nvidia? All that money is burned when people don't focus on maximum performance, that levels of optimization that companies need in the AI race will never come from an virgin memory management system like borrow chequer. Go, Cry! 😎
Awesome video! Thanks for share!
7:15 I believe std::execution (a.k.a senders) is on track for C++26. Reflection and Contracts are also aiming C++26, but I think both are “MVP” in the sense that both might seem rudimentary in C++26 and expected to grow on C++29. Most notably I don’t know if “token streams” will make it for C++26.
Yeah, the way C++ standardization works now is that if a feature is ready in time for publication then it'll be included and if it isn't then it'll have to wait three years for the next version of the standard. Reflection and contracts seem quite likely to be included (especially contracts which had previously been planned for C++23) but there's always a chance one or both will be nixed at the last minute.
Great reviee ❤
using gleam and wanting to use it more
Nice to see GDScript being a player amongst the myriad of programming languages.
Let's go Tom upload!
My biggest highlight about Nim in 2024 is how much they improved tooling. It's still not perfect, but a lot of work was done on LSP (nimsuggest), VS-code plugin and package manager (nimble).
And for 2025, there's a roadmap with new intermediate language that will get us easier interop with C, LLVM, Wasm and incremental compilation. I'm beyond excited for more Nim in new year!
Until incremental compilation gets to be a thing I don't see Nim as bringing enough of a value proposition compared to other low level languages in terms of developer productivity.
@@encapsulatio Nim compiler is already fast, IC will be good for repl though. Also I mainly use Nim as a high-level language. Think of Python, but typed, fast and with small executables.
@@dolorsitametblue You can't use right now Nim to do exploratory programming like you can in Python with equal responsiveness..
"where >>this year
Gleam mentioned! ❤
Gleam as a language looks good, but the community is hopelessly stupid.
Nothing about Raku? I can't really point out anything in particular, but it does have monthly-ish updates :) I find it to be a really fun language to code in when speed isn't of the essence. It can express so many crazy things.
Why don't you popularize it if you like it so much? People like to praise Raku yet they don't ACTUALLY do anything about popularizing it on youtube, showing comparisons in improved developer productivity compared to Python or Typescript/Javascript for different language features or mini features unique to that language alone.
Can you do an episode on Seed7? It's most niche language I know, but it's very impressive for what one person can achieve in a span of 20+! years.
It's such a shame that Nim has not been used for any big project. It is the perfect general purpose lang without memory management. Regarding the C alternatives, I think Odin is simpler to learn and write than Zig, even when the latter has an amazing performance.
3:26 I assume a lot is missing from this list. I do note that F# is missing for sure.
Yeah, showing all languages would be noisy, so I handpicked some for good or ill.
But you can still find F# on Languish. Here it is with some nearby languages. tjpalmer.github.io/languish/#y=mean&weights=issues%3D0%26pulls%3D0%26stars%3D1%26soQuestions%3D1&names=f%23%2Ctypst%2Cwebassembly
Lua and J ?
Do you think it’s fair to separate typescript in JavaScript when typescript is just a superset of JavaScript? It’s not like your average type script programmer is very different from your average JavaScript programmer.
I like seeing both for now, and it's fairly easy to imagine the combo. Maybe someday I'll just combine them. If I keep track of Stack Overflow tag overlaps, doing custom combos might work better, though.
Where did Zig land?
Zigfault landed on last place 😂
I would love to see Python's numbers without including AI related projects
I agree.
Why? Would you want to see ts/js numbers without webdev projects?
@ I personally don't mean that they shouldn't count toward the Python total. I just mean I'd be curious to see the fraction.
@ Same answer as @contextfree. But we can go that route if you want to. I believe that AI work isn't the same as "programming", I see it as something more related to math than software development. There is also the fact that most of all AI heavy work is done by C/C++ that are just exposed as APIs inside python, because python is an easy language, so one could argue that AI stuff in Python doesn't actually count as software development.
Primeagen ditched Rust for Zig in 2025 so expect Zig to replace it on the charts by next year
Primer is a golangooner. Nice copium LOL Zigfault could never 🤣
D mentioned
i think in ever list they should just combine js and ts
Hype on Rust is over and only the people who see its real value on long and large projects stay.
and now hype zig 😂
Wdym we are just getting started!! 🦀
3:40 The GIL is sleeping.
jq best programming language
It's definitely awesome!
2025 is the year I match my Odin and Swift knowledge with C. If C++ comes to my mind, I will likely use it only when I know what I am doing lol. And that probably means learning an older subset.
Some things about the data, most languages are pretty stable, so I think that implies to me the language market is pretty competitive, no? More languages seem to be kind of just keeping uo rather then rising/falling. The ones that are rising are usually consequential to something else in the ecosystem like how Javascript is being eaten by Typescript.
My only thing is how many systems languages can the market sustain? As many as godbolt can host, apparently lol 😂.
Also ones that have out of the box C/C++ support, it seems.
On stats, Languish even focuses just on quarterly activity, not sum totals across time. Over the past 15 years or so, there have been some new big languages (such as Dart, Go, Kotlin, Rust, Swift, & TypeScript), but it does take a while, and I don't know how to predict the future.
JavaScript is dying, and we can be all happy about it. It was about time!
This covers several languages I'm interested in, as well as many I haven't even heard of. 🙂 One language that I'm interested in which you didn't dive into in depth is Mojo, although I do notice it in the chart. Very nice chart, btw. It's fun to play around with that web page.
I wonder when will Mojo will be released? I know this is Chris Lattner's baby. I am very curious to learn more about its release date.
Wake me up when they release it under open source license. ZzzzzzZzzz
MoJoMama not even open LOL
No Lua because lua is the OG
The zig toolchain can not be stopped.
You are not prepared!
It's a great toolchain. I'm always hoping for Zig 1.0.
@@contextfree hehe I'm also excited about it :)
Feb 2025 is the scheduled date for the next zig update which will probably contain the first implementation of the incremental compilation. Exciting stuff.
@@10e999 Exciting is an understatement! Incremental compilation is one-of-a-kind never-before-seen leave-everyone-else-in-the-dust feature
Zigooners think they are winning. How cute 😂
@ All the languages are great!
Take this advice after watch the video:
"Languages in 20XX" is like "Ice cream flavors in 20XX". You can do what you want to, in any language. You can make your own back-end using Pistache and C++, creating HTML reponse by creating strings with tags.
Will it be insane harder instead of using Node, Bun, etc? Yep, but you can.
Can you build a big-large modularized CRM for Windows, using Dart instead of WinForms? Yep, but you can.
Languages are designed since the begining to bring solution but languages have a first-vision or mission, so that why his put focus for specific fields or problems because are inherently better suited for these problems.
If "your" language are trying to do 'all', is a red flag (Don't bite off more than you can chew).
Many languages like ECMAScript (JavaScript for kids) are constantly evolving and introducing new tools and feature to make easier what just are easier. However, if you are looking for a new language based on fashiontards tredings from reddit, maybe you need to looking for a new profesion.