We really need more tutorials from people creating stuff with Nim. Hopefully you can create something like a web app or some cli tool to see the language in action
There are tons of examples on GitHub and the documentation for the standard library is great. My resume website is all done with Nim. The discord is super active as well
Indeed. I am pretty new to Nim and am especially interested in Qt-GUIs. Examples are almost non-existant unfortunately and there are no really good in depth books. It will take a lot of trial and error I fear, especially when coming from a non-c-background.
I always confused your channel with CodeAesthetics. I was surprised to why there are so less like and views. Your channel is actually underrated. Need more subscribers. Hope you get them soon :)
Dude, you did this language justice. I've been writing Nim everyday for over 3 years now, never needed to touch another language since I started learning it
@@_tsu_ I recommend you try both, personally I enjoyed Nim a lot more. Also, since nim transpiles to C it works everywhere an ANSI C compiler exists, which is a lot more than the few platforms supported by LLVM. Nim also has the option of different levels of memory management, whereas rust forces a single style on you. You can write Nim the same way you write C and you'll get a filesize and performance that is pretty much identical to handwritten C, or you can use garbage collection, lots of heap allocated objects, etc. and you'll still compete with Rust and C++.
Thank you so much for the video! I started learning nim a few month ago and except some points, I clearly love the language and want everyone to hear about it! Especially since everyone say Mojo is the next Python nowing that Nim already exist with a powerfull compiler
I just installed nim and it's amazing. Such a shame that it isn't popular! The syntax is extremely readable and easy to learn and understand. And the performance is way better than python. Worth a try!
Nim is great! It's got tons of built-in features, it's fast, and its syntax is fantastic. Python is more than those features though, it's an entire ecosystem, and that's hard to compete with. But yeah, we should 100% be recommending Nim to college-level devs over Python.
While being a nice language at the first glance, Nim had some flaws too. 1. No package scope. You can declare a proc or a type or a variable as either private or public. These scopes are applicable for files. There is no concept of package. This is a big flaw when you write a gui library or similar packages. 2. No support for circular module dependency. Some times, we need to import module A in module B and vice versa. This is not allowed in Nim. In such situations, you will need to re-arrange your code and treat the entire source code as a single file. But then you can split the code by different files and combine them with include keyword. At first glance you will think it can solve the problem, but there is another problem. You don't get much support from intellisence. Theoretically module B comes after module A, so variables declared in A can be accessible in module B. But you will not see those variables in intellisence list. 3. It needs forward declaration. Although there is an experimental pragma to avoid this, I see this as an annoying problem. 4. Lack of proper documentation. I wrote more than 15K lines of code in nim and did several projects in it, but even today I don't know how to start a process asynchronously in Nim. Tried so many times, but without proper docs, I failed and did the same project in python easily. Despite all these things, Nim is a nice language.
I'd agree with 2 and 4. 2 has bitten me in the ass before and was annoying. 4 should be higher on the priority list, together with tooling. These things are essential. Other than that, it's a pleasure to work with.
@@godDIEmanLIVE Yeah, the best workaround for issue 2 is declaring ll your types in a single file and import in other module. Or you can use the 'include' feature.
I think it's fair to say 4 is the new(er) language experience though, right? Be that Nim, Zig, Crystal, etc. documentation tends to be thin early in life.
@@stretch8390 Yeah, I can agree with that. And we can depend discord channels for to overcome that issue. Nim's discord community is very helpful. But there is little problem. It's not a good idea to bother others for all your queries.
NIm is indeed a significantly better Python because it does not work like python it just has python syntax for redability . Mojo is however the next evolution of Safer and Fast python...but yeah If I had to choose Mojo or Nim I would always choose Nim.
Gotta be honest, it would've been nice to hear a bit more about why their package manager can do more then others, cause now I know it can use arbitrary packages from links, which most package managers can do, albeit with maybe a few commands, but what makes nimble more flexible/special? Can it also use c++ or js libraries in ním or something?
Setting up Nim development environment is a nightmare, the Nim devs delegate formatting responsibility to 3rd party providers. Despite the 2.0 status, the Nim ecosystem is nowhere near mature. LSP support, linting, and formatting are all messy right now. Especially if you're using Neovim as the editor.
I agree with the devs about formatting, I don't think that is a responsibility of the compiler, especially in a flexible language lile Nim. But yes, the tooling is abysmal, completely unusable
@@PouriyaJamshidi Tell me that you know nothing about Nim without telling me you know nothing about Nim. Most stuff in Nim's stdlib is very performant.
Great language but once again the bottleneck is adoption. Why would you want to create a project in Nim / rewrite an entire codebase in Nim when the ecosystem is so small?
I made a few projects in nim, I wouldn't recommend building anything for production imo. It's too unstable. I usually rebuild my important projects in go and they work far far better in terms of stability. I don't know what makes nim unstable, but it just is for some reason.
I don't think there is going to bw next python as a language, and I don't think we need next python. Instead of reinventing the wheel let's keep things that have worked for a long time and will continue to work. There may be new interpreters or compilers, but the language itself will likely bot bw replaced
im generally not a fan of whitespace-based syntax, but nim seems interesting. the fact that its transpiled into C++ first instead of just being compiled directly into machine code seems a bit odd
I was also cautious of significant-whitespace syntax at first; I believe it came from my first time trying Python in pyCharm IDE some long years ago - I was trying it blindly, not really knowing language or what I'm doing, so I obviously made lots of syntactic and stylistic errors and just that experience of "popping up random PEP### warnings and fiddling with code until it works" left really bad first impression. But IMO Nim has better syntax than Python: It fixes some ugly parts and even allows you more freedom of formatting your code where python would throw syntax errors. Common pain points of indentation-based syntax are easy to fix with a good editor e.g. in neovim it is trivial to (de)indent code blocks and there are plugins for visible indentation levels. > the fact that its transpiled into C++ first instead of just being compiled directly into machine code seems a bit odd Main target is C, then C++ and JS. Nothing odd with *compiling* to C, many languages started as compilers to C, before switching to bootstrapping and native compilation. But staying C-dependant actually has a lot of benefits - easy interop with popular C libraries, virtually every platform is supported (C is usually the first compiler ported to literally any system), common C debugging tools (valgrind, gdb, etc.), decades of compiler optimizations in either gcc or clang.
The creator of the nim language believes that violence is justified to prevent a conspiracy of leftists and queer people: "Jesus will forgive us for what we have to do". No, this is not the next python, it is a showcase of a ton of great ideas in a language doomed to obscurity by mismanagement.
@@arjix8738 obviously it will not compile. you need to import it with a different name. If comming from C: var bar {.header: "foo.h", importc: "bar".}: cint var barb {.header: "foo.h", importc: "bAR".}: cint var Bar {.header: "foo.h", importc: "BAR".}: cint in nim bar and Bar are different, bar and bAr are the same variable. First letter must match capitalization.
No this wont be the next python, pythons popularity comes from the fact you can import everything to do all the work for you, leaving the simple scripting for you to do.
What's the point tho? If I need nim to compile to C++, why not just use C++? The same goes for JS and other languages. This cross compiling sounds good but won't do F in the real world since JS is much different from C and we would need two separate nim codebases for both C and JS. So, why not just write two separate codebases in an actual programming language rather than writing in Nim?
Writing in Nim over writing in C gives you the speed of C with the interop of existing C code but with the safety of a modern language and with features you would expect of a modern language, for example while you could implement your own resizable arrays in C or use someone else's implementation, they just exist and work in Nim, with type and memory safety. The purpose of cross compilation is not to reduce your number of code bases if that is what you're getting at. Personally my favorite use case of C compilation is that Nim can run on any architecture that has an ANSI C compiler (all of them btw), ARM microprocessors for example.
Because C++ is dog shit. Also Nim compiles to binary, not C++. C/C++ can be intermediate step in the compilation tool chain. Your argument is about as sensible as saying why do I need language X if it just compiles to assembly or IR? Just use assembly / IR.
Меня забавляют люди, которые ненавидят компиляцию в C/C++. А с практической точки зрения это абсолютно верный выбор. Как минимум, потому что тогда вы получаете возможность компилировать свой код на все архитектуры и операционные системы. У тебя что, много лишнего времени и денег? Ты горишь желанием написать очередную тонну собственных компиляторов под десятки комбинаций ОС и архитектур? Ты безумец? + Такое решение позволяет легко импортировать целый океан библиотек C/C++, а не писать всё с нуля. Создатели Nim сильно сэкономили время и ресурсы, не создавая всё с нуля, а взяв уже готовую экосистему. Это позволило бросить все ресурсы на развитие самого языка, а не его экосистемы. Какая разница в какое ПРОМЕЖУТОЧНОЕ ПРЕДСТАВЛЕНИЕ компилируется исходный код? Байт-код, исходный код другого языка, или что-то ещё.... Какая разница? Просто потому что тебе приятнее компилироваться сразу в бинарный код? В случае с компилируемым языком код всё равно в конечном итоге компилируется в бинарный код. Так, кого ебёт? У тебя какое-то не практичное и фанатичное мышление. Удивительно, как в IT много фанатиков. Тут программная инженерия, а не церковь. Выбирать надо то что удобнее и практичнее работает, а не ориентироваться на какие-то слепые догмы и идеологию. Безумные фанатики.
0:45, this is an awful idea: the need for shift at each type works against productivity. 0:51, where's the apostrophe at 1'024? Don't you gonna say that ancient C++ is more modern than this 1, right? Plus, for a so called modern language, I expect to type less. So why force me to memorize let? I want to type const on both sides. Even better, const by default, so should not type them at all. PlusPlus, string could be deduced from " ", and numbers could have a default to platform size, and signal to gain low level performance. Joining this to my previous comment, I think a truly modern enjoyable language should has this syntax: name = "Code Persist" BIN_SIZE = 1'024
it does not change productivity at all, learn how to type. Const by default is useless outside of functional languages, nim is an imperative procedural language, if you don't like that, fine, just use Haskell, OCaml, F#, Scala, Elixir, etc... Typeless is a terrible idea, static typing + good union types is the tried and true method that builds useful and maintainable software And Nim does have deduction, literally just pay attention to the video
@@marcs9451 There's a hidden/silent loss of energy on the long run. Typing less saves some % from that, while helps to "follow the thoughts", reaching faster what the programmer thinks, while still "hot" inside human head. I think the "Const Correctness Principle" is universal, because it'll be reflected even for local variable. 1 won't put everything in classes, even local variables, right? And even for classes, it's good to keep them const whenever possible (less prone to errors). About deduction, I only saw 'let and type', 'const and type', 'proc and type'.
Beginners that never do anything with the languages just jump to the new language that’s trending but professionals just stick to what gets the job done, it’s not usually about the language it’s about what you want to do… to all beginners don’t get stuck in tutorial hell, choose a language or two and stick to it, go do something with it!
Такой проблемы не было бы, если бы python не был таким медленным как сейчас. Если бы он нативно был по скорости хотя бы на уровне javascript/java, то новички дольше бы писали не нём. Но приходится искать более быстрые нативные языки. Надеюсь, такие проекты как Nim и Mojo это исправят.
I like the concept of the language, but any language that uses significant whitespace is inherently flawed. It is bad enough that Python does this and actually gained mainstream popularity, but there is no excuse to design a new language that does the same. I apologize to anyone who disagrees and mistakenly believes otherwise without irony.
Забавно. Ты просто выставил своё субъективное мнение как неприложную истину. По моему субъективному опыту, люди которые так делают обычно очень глупы и ограничены. Отступы хорошо читаются. Особенно для новичков. И делают код более единообразным, что тоже улучшает чтение своего старого и чужого кодов. Если ты прав, давай, покажи нам аргументы ошибочности отступов. И какой формат самый лучший, и почему.
I think the whole C-syntax with its overuse of brackets of all kinds and convoluted syntax is inherently flawed. Different tastes and all that... Python, and to a lesser degree Nim, imho, is way more readable than anything in Javascrypt, C, C++ or, heaven forbid, Java.
Wow, great language... Bad syntax, no new ideas (except the GC edit), proc, no automatic returning type deduction. I gave thumbs up for the good video, though. 0:51, we don't have this: name = "Code Persist" BIN_SIZE = 1'024 So I'm forced to go to the catacombs, bring C++ back to life, to achieve something near that: const string name = "Code Persist"; //If char * is enough, could exchange const string by auto. const auto BIN_SIZE = 1'024; //Defaults to int. 0:56, please C++, guide these modern languages towards less typing: string n = "1st str", s = "2nd str"; //But I'd like const string by default: auto n = "1st str", s = "2nd str"; // ';' speeds compile time, so it stays. I want fast pace typing syntax, not Carbon-like bullshit. 3:44, is it serious that it'll put a crown on the head of a stupid language, that doesn't even offer pointers and memory management? I suggest the buffon hat.
@@agustinpizarrolet and const could be only 1 word. Lots of speed are earned by pointers. No wonder why C++ has all its standard algorithms expecting pointers. At least lambdas are "modern". Well, in the end, it's better f()s don't be automatic forever.
Nim just promotes the methods that work best for your situation. Most people who say they dislike Nim actually complain about its flexibility. Perhaps these are the same people who when faced with too many choices fail to make any...
IN ETHIOPIA WE LEARNED NIM AS FRESH MAN COURSE AT ANY UNIVERSTY FOR CS AND SOFTWARE ENGINEERING DEPARTEMENT
Cool
Cool
That's awesome!
did they forget to teach you how to disable Caps Lock in your university?
@@weeb3277😂
We really need more tutorials from people creating stuff with Nim. Hopefully you can create something like a web app or some cli tool to see the language in action
There are tons of examples on GitHub and the documentation for the standard library is great. My resume website is all done with Nim. The discord is super active as well
Indeed. I am pretty new to Nim and am especially interested in Qt-GUIs. Examples are almost non-existant unfortunately and there are no really good in depth books. It will take a lot of trial and error I fear, especially when coming from a non-c-background.
let's make nim more popular
I always confused your channel with CodeAesthetics. I was surprised to why there are so less like and views. Your channel is actually underrated. Need more subscribers. Hope you get them soon :)
Dude, you did this language justice. I've been writing Nim everyday for over 3 years now, never needed to touch another language since I started learning it
why not rust tho? has all the same features except the compile to JS thing.
@@_tsu_ Rust is much more complicated to do anything right with it, among other things
@@_tsu_ because it's not simple as nim
@@_tsu_ I recommend you try both, personally I enjoyed Nim a lot more. Also, since nim transpiles to C it works everywhere an ANSI C compiler exists, which is a lot more than the few platforms supported by LLVM. Nim also has the option of different levels of memory management, whereas rust forces a single style on you. You can write Nim the same way you write C and you'll get a filesize and performance that is pretty much identical to handwritten C, or you can use garbage collection, lots of heap allocated objects, etc. and you'll still compete with Rust and C++.
@@_tsu_nim is more productive and fun to write
I am head over heels for nim. I hope it gains great momentum, I just love it.
Thank you so much for the video! I started learning nim a few month ago and except some points, I clearly love the language and want everyone to hear about it! Especially since everyone say Mojo is the next Python nowing that Nim already exist with a powerfull compiler
I just installed nim and it's amazing. Such a shame that it isn't popular! The syntax is extremely readable and easy to learn and understand. And the performance is way better than python. Worth a try!
Nim is great! It's got tons of built-in features, it's fast, and its syntax is fantastic. Python is more than those features though, it's an entire ecosystem, and that's hard to compete with. But yeah, we should 100% be recommending Nim to college-level devs over Python.
some nice nim packages (modules installed with nimble) that you should check: sugar (nice macros), nimpy (superb python interop), zero_functional (nice FP), cppstl (C++STL), nimx (GUI Apps) karax(safe HTML), weave (faaaast multithread/async), interfaced (OOP), ormin (ORM, safe SQL), NimTorch (PyTorch without python)
sugar is in the standard library (std/sugar), not in the nimble packages.
@@llIllIlI search for khchen/xl
@@llIllIlI khchen/xl
@@llIllIlI my responses get deleted instantly. khchen/xl
@@llIllIlI khchen-xl
This was a great video, I had heard about nim but never looked into it until now
Oh man I want your channel to grow bigger.
Great work my man, nim's looking pretty hot!
While being a nice language at the first glance, Nim had some flaws too.
1. No package scope. You can declare a proc or a type or a variable as either private or public. These scopes are applicable for files. There is no concept of package. This is a big flaw when you write a gui library or similar packages.
2. No support for circular module dependency. Some times, we need to import module A in module B and vice versa. This is not allowed in Nim. In such situations, you will need to re-arrange your code and treat the entire source code as a single file. But then you can split the code by different files and combine them with include keyword. At first glance you will think it can solve the problem, but there is another problem. You don't get much support from intellisence. Theoretically module B comes after module A, so variables declared in A can be accessible in module B. But you will not see those variables in intellisence list.
3. It needs forward declaration. Although there is an experimental pragma to avoid this, I see this as an annoying problem.
4. Lack of proper documentation. I wrote more than 15K lines of code in nim and did several projects in it, but even today I don't know how to start a process asynchronously in Nim. Tried so many times, but without proper docs, I failed and did the same project in python easily.
Despite all these things, Nim is a nice language.
I'd agree with 2 and 4. 2 has bitten me in the ass before and was annoying. 4 should be higher on the priority list, together with tooling. These things are essential. Other than that, it's a pleasure to work with.
@@godDIEmanLIVE Yeah, the best workaround for issue 2 is declaring ll your types in a single file and import in other module. Or you can use the 'include' feature.
@@kcvinu yeah I've resorted to a "type" file or you could call it an interface file with just the types for other files to interact with etc.
I think it's fair to say 4 is the new(er) language experience though, right? Be that Nim, Zig, Crystal, etc. documentation tends to be thin early in life.
@@stretch8390 Yeah, I can agree with that. And we can depend discord channels for to overcome that issue. Nim's discord community is very helpful. But there is little problem. It's not a good idea to bother others for all your queries.
I thought you were gonna say Mojo
yeah me too, but I was impressed
god i cant wait for mojo as well
Me too 😂
Mojo is also pretty similar maybe I'll make a video on it later!
"Mojo is also fast, but it's literally the newest programming language, which means it's unstable."
Nim is great and deserves more attention but it is definitely not batteries included.
Also, the drama of Rust
Pure gold! Going to give it a try
I love how this has the python zen and it looks fairly easy to work with unlike Rust
Curious what you use for your animations?
NIm is indeed a significantly better Python because it does not work like python it just has python syntax for redability . Mojo is however the next evolution of Safer and Fast python...but yeah If I had to choose Mojo or Nim I would always choose Nim.
My response to your opening statement is I'd cry
What did you use to make those transitions? They look super cool
looks like it's power point or something similar. You can do quite a bit with text morph
From the intro I thought you're gonna say Julia; Nim is ok 😅
Oh dang! I thought you were going for Mojo but I guess Iam picking up nim now.
Gotta be honest, it would've been nice to hear a bit more about why their package manager can do more then others, cause now I know it can use arbitrary packages from links, which most package managers can do, albeit with maybe a few commands, but what makes nimble more flexible/special? Can it also use c++ or js libraries in ním or something?
Static typing honestly makes it much more appealing to me.
i wish there was a good opencv wrapper for it :(
Thanks for this incredible video!!
Awesome video
Setting up Nim development environment is a nightmare, the Nim devs delegate formatting responsibility to 3rd party providers. Despite the 2.0 status, the Nim ecosystem is nowhere near mature. LSP support, linting, and formatting are all messy right now. Especially if you're using Neovim as the editor.
I agree with the devs about formatting, I don't think that is a responsibility of the compiler, especially in a flexible language lile Nim.
But yes, the tooling is abysmal, completely unusable
apt install gcc nim
done
even worse, much of its stdlib is not aimed for writing high-performance programs. Basically you need to write almost everything from scratch
@@agustinpizarroYou missed their point. I've been using Nim happily for a long time but nimsuggest is top tier garbage.
@@PouriyaJamshidi Tell me that you know nothing about Nim without telling me you know nothing about Nim. Most stuff in Nim's stdlib is very performant.
great vid. what font?
Great video! What font is being used?
I haven't checked it but looks like JetBrains Mono to me.
JetBrains Mono for the code and Brandon Grotesque for everything else!
Great language but once again the bottleneck is adoption. Why would you want to create a project in Nim / rewrite an entire codebase in Nim when the ecosystem is so small?
Can you easily import python libs on pip to it? If so then that would be insanely useful
yes, with nimble, and you can mix and match with python with nimpy package
👍👍!!
I really want to like Nim, it’s got some really cool features. But I don’t like Python syntax…
Elixir
I made a few projects in nim, I wouldn't recommend building anything for production imo. It's too unstable. I usually rebuild my important projects in go and they work far far better in terms of stability. I don't know what makes nim unstable, but it just is for some reason.
I think I would use it if it had curly braces
I don't think there is going to bw next python as a language, and I don't think we need next python. Instead of reinventing the wheel let's keep things that have worked for a long time and will continue to work.
There may be new interpreters or compilers, but the language itself will likely bot bw replaced
im generally not a fan of whitespace-based syntax, but nim seems interesting. the fact that its transpiled into C++ first instead of just being compiled directly into machine code seems a bit odd
I was also cautious of significant-whitespace syntax at first; I believe it came from my first time trying Python in pyCharm IDE some long years ago - I was trying it blindly, not really knowing language or what I'm doing, so I obviously made lots of syntactic and stylistic errors and just that experience of "popping up random PEP### warnings and fiddling with code until it works" left really bad first impression.
But IMO Nim has better syntax than Python: It fixes some ugly parts and even allows you more freedom of formatting your code where python would throw syntax errors.
Common pain points of indentation-based syntax are easy to fix with a good editor e.g. in neovim it is trivial to (de)indent code blocks and there are plugins for visible indentation levels.
> the fact that its transpiled into C++ first instead of just being compiled directly into machine code seems a bit odd
Main target is C, then C++ and JS. Nothing odd with *compiling* to C, many languages started as compilers to C, before switching to bootstrapping and native compilation. But staying C-dependant actually has a lot of benefits - easy interop with popular C libraries, virtually every platform is supported (C is usually the first compiler ported to literally any system), common C debugging tools (valgrind, gdb, etc.), decades of compiler optimizations in either gcc or clang.
The creator of the nim language believes that violence is justified to prevent a conspiracy of leftists and queer people: "Jesus will forgive us for what we have to do". No, this is not the next python, it is a showcase of a ton of great ideas in a language doomed to obscurity by mismanagement.
this is how cursed nim is:
var hi_there = 5;
hiThere = 6;
echo hi_there;
that is a feature to easily interop with C without having to stick with C naming format.
@@agustinpizarro what happens when there is a collision?
@@arjix8738 obviously it will not compile. you need to import it with a different name. If comming from C:
var bar {.header: "foo.h", importc: "bar".}: cint
var barb {.header: "foo.h", importc: "bAR".}: cint
var Bar {.header: "foo.h", importc: "BAR".}: cint
in nim bar and Bar are different, bar and bAr are the same variable. First letter must match capitalization.
Literally a non issue.
@@godDIEmanLIVE since when are "cursed" things an issue?
Like, yeah, one can use nim w/o issue, that doesn't make it less cursed tho
No this wont be the next python, pythons popularity comes from the fact you can import everything to do all the work for you, leaving the simple scripting for you to do.
What's the point tho? If I need nim to compile to C++, why not just use C++? The same goes for JS and other languages. This cross compiling sounds good but won't do F in the real world since JS is much different from C and we would need two separate nim codebases for both C and JS. So, why not just write two separate codebases in an actual programming language rather than writing in Nim?
Maybe for fast Python-like code writing speed and C++-like performance after transpilation
It is a cool proof but Not useful Nor used by companies.
Writing in Nim over writing in C gives you the speed of C with the interop of existing C code but with the safety of a modern language and with features you would expect of a modern language, for example while you could implement your own resizable arrays in C or use someone else's implementation, they just exist and work in Nim, with type and memory safety. The purpose of cross compilation is not to reduce your number of code bases if that is what you're getting at. Personally my favorite use case of C compilation is that Nim can run on any architecture that has an ANSI C compiler (all of them btw), ARM microprocessors for example.
Because C++ is dog shit. Also Nim compiles to binary, not C++. C/C++ can be intermediate step in the compilation tool chain. Your argument is about as sensible as saying why do I need language X if it just compiles to assembly or IR? Just use assembly / IR.
Меня забавляют люди, которые ненавидят компиляцию в C/C++. А с практической точки зрения это абсолютно верный выбор. Как минимум, потому что тогда вы получаете возможность компилировать свой код на все архитектуры и операционные системы. У тебя что, много лишнего времени и денег? Ты горишь желанием написать очередную тонну собственных компиляторов под десятки комбинаций ОС и архитектур? Ты безумец?
+ Такое решение позволяет легко импортировать целый океан библиотек C/C++, а не писать всё с нуля.
Создатели Nim сильно сэкономили время и ресурсы, не создавая всё с нуля, а взяв уже готовую экосистему. Это позволило бросить все ресурсы на развитие самого языка, а не его экосистемы.
Какая разница в какое ПРОМЕЖУТОЧНОЕ ПРЕДСТАВЛЕНИЕ компилируется исходный код? Байт-код, исходный код другого языка, или что-то ещё.... Какая разница? Просто потому что тебе приятнее компилироваться сразу в бинарный код? В случае с компилируемым языком код всё равно в конечном итоге компилируется в бинарный код. Так, кого ебёт?
У тебя какое-то не практичное и фанатичное мышление. Удивительно, как в IT много фанатиков. Тут программная инженерия, а не церковь. Выбирать надо то что удобнее и практичнее работает, а не ориентироваться на какие-то слепые догмы и идеологию.
Безумные фанатики.
mojo lang
I dont like the syntax of python in the slightest
0:45, this is an awful idea: the need for shift at each type works against productivity.
0:51, where's the apostrophe at 1'024? Don't you gonna say that ancient C++ is more modern than this 1, right? Plus, for a so called modern language, I expect to type less. So why force me to memorize let? I want to type const on both sides. Even better, const by default, so should not type them at all. PlusPlus, string could be deduced from " ", and numbers could have a default to platform size, and signal to gain low level performance. Joining this to my previous comment, I think a truly modern enjoyable language should has this syntax:
name = "Code Persist"
BIN_SIZE = 1'024
it does not change productivity at all, learn how to type. Const by default is useless outside of functional languages, nim is an imperative procedural language, if you don't like that, fine, just use Haskell, OCaml, F#, Scala, Elixir, etc...
Typeless is a terrible idea, static typing + good union types is the tried and true method that builds useful and maintainable software
And Nim does have deduction, literally just pay attention to the video
@@marcs9451 There's a hidden/silent loss of energy on the long run. Typing less saves some % from that, while helps to "follow the thoughts", reaching faster what the programmer thinks, while still "hot" inside human head.
I think the "Const Correctness Principle" is universal, because it'll be reflected even for local variable. 1 won't put everything in classes, even local variables, right? And even for classes, it's good to keep them const whenever possible (less prone to errors).
About deduction, I only saw 'let and type', 'const and type', 'proc and type'.
@@MrAbrazildo "let" defines a C++ "const", "var" defines a non-const. No need to type more chars. "const" in nim is compile time checked C++ #define
there is unfortunately no momentum behind this language. and it's all because of the 2space indent.
? Отступы могут быть любыми. Можно и 4.
Beginners that never do anything with the languages just jump to the new language that’s trending but professionals just stick to what gets the job done, it’s not usually about the language it’s about what you want to do… to all beginners don’t get stuck in tutorial hell, choose a language or two and stick to it, go do something with it!
Такой проблемы не было бы, если бы python не был таким медленным как сейчас. Если бы он нативно был по скорости хотя бы на уровне javascript/java, то новички дольше бы писали не нём. Но приходится искать более быстрые нативные языки.
Надеюсь, такие проекты как Nim и Mojo это исправят.
Nim us no match to Mojo
Why? It would be easy to implement mojo's functionality (like autotune) with macros.
I like the concept of the language, but any language that uses significant whitespace is inherently flawed. It is bad enough that Python does this and actually gained mainstream popularity, but there is no excuse to design a new language that does the same. I apologize to anyone who disagrees and mistakenly believes otherwise without irony.
How is it "inherently flawed"? What problem do you have with it? Other than you not liking how it looks but that's just a personal preference.
Literally a non-issue, completely uninteresting.
Забавно. Ты просто выставил своё субъективное мнение как неприложную истину. По моему субъективному опыту, люди которые так делают обычно очень глупы и ограничены. Отступы хорошо читаются. Особенно для новичков. И делают код более единообразным, что тоже улучшает чтение своего старого и чужого кодов.
Если ты прав, давай, покажи нам аргументы ошибочности отступов. И какой формат самый лучший, и почему.
I think the whole C-syntax with its overuse of brackets of all kinds and convoluted syntax is inherently flawed. Different tastes and all that... Python, and to a lesser degree Nim, imho, is way more readable than anything in Javascrypt, C, C++ or, heaven forbid, Java.
Wow, great language... Bad syntax, no new ideas (except the GC edit), proc, no automatic returning type deduction. I gave thumbs up for the good video, though.
0:51, we don't have this:
name = "Code Persist"
BIN_SIZE = 1'024
So I'm forced to go to the catacombs, bring C++ back to life, to achieve something near that:
const string name = "Code Persist"; //If char * is enough, could exchange const string by auto.
const auto BIN_SIZE = 1'024; //Defaults to int.
0:56, please C++, guide these modern languages towards less typing:
string n = "1st str", s = "2nd str"; //But I'd like const string by default:
auto n = "1st str", s = "2nd str"; // ';' speeds compile time, so it stays.
I want fast pace typing syntax, not Carbon-like bullshit.
3:44, is it serious that it'll put a crown on the head of a stupid language, that doesn't even offer pointers and memory management? I suggest the buffon hat.
has automatic return type with lambdas. Check "sugar" module
of course it has pointers... serach for "ptr" and manual memory management.... but 99.99% of the times you DO NOT need that.
let a = "a is not mutable"
learn the basics of nim before writing nonsense, otherwise your arguments look stupid.
@@agustinpizarrolet and const could be only 1 word.
Lots of speed are earned by pointers. No wonder why C++ has all its standard algorithms expecting pointers.
At least lambdas are "modern". Well, in the end, it's better f()s don't be automatic forever.
As long as THIS_VAR contain the same value as THISVAR I don't think so ;)
😂
There is special ring in hell for people who use both THIS_VAR and THISVAR in the same project.
I don't see the problem, other than you consciously choosing to mix styles.
Do we really need another language...
its not a new language, it's been around for years
1. Not that new
2. Languages are tools, sometimes we need new tools
There's a language for just about anything 😅
Nim's existed since 2006.
This honestly just seems like worse Rust
Rust's syntax is subjectively an awful subset of C++'s syntax, which is already a messy subset of C's near-perfect syntax.
Rust syntax IS garbage
Nim just promotes the methods that work best for your situation. Most people who say they dislike Nim actually complain about its flexibility. Perhaps these are the same people who when faced with too many choices fail to make any...
An even better python
PYTHON5.0 aka: Haskell