Comments by "raianmr" (@raianmr2843) on "Lex Clips" channel.

I don't think Python will ever manage to fix a problem so fundamental. It's like asking for natively supported green threads in C - never going to happen. I'm glad languages like Julia exist and I hope they get the mindshare they deserve.
12
JS is objectively the worst mainstream language. It began as an underfunded toy project, got stuck between several corporate wars, and is now left with permanent unfixable bugs due to the insane backward compatibility requirements of the web. If you can get all those things done with JS, it's probably because you're a really capable engineer and not because the language is any good. An actual decent engineer would, however, use the best tool for the job.
8
Just start with Python. This sounds like an unresearched opinion but trust me, once you've actually explored all other options, you'll come right back at Python. This is not because Python is the best language ever; it's not even the simplest mainstream language (that would be Go). Python just happens sit nicely on almost all categories. I would generally advise against starting with scratch. Scratch is a fun educational platform for children, that much is true. But if your end goal is just to learn a real language and if your kid is already smart enough (which they are), then Scratch is basically just a fun but ultimately wasteful use of their time. I personally started programming with Python as a 15 year old (and moved to the C and C++ world soon after), and this was in the last decade. Kids today are much smarter, and have a mature internet at their disposal. I strongly advise against starting with anything low-level-ish like C and assembly (even though that knowledge will be of much help later on). Your kid will simply miss the point of working at this level of complexity with zero feedback and you'll only be gatekeeping them from a large chunk of the fun and fruitful world of programming. I also advice against two major languages: Java and JavaScript. Both of these are excellent for getting a job in the industry but have major red flags for your use case. The industry is detoxing from Java at a pretty fast rate. By the time your kid steps into the industry, very few new shops and projects depend on Java and Java-based technologies. Investing in the Web is a much, much better option. However, JS is not a 'normal' language. It has a long and rough legacy that has left it with counterintuitive language features and what are essentially implementation bugs that later got merged into the language as features for backward compatibility reasons. You'll basically spend most of your time dealing with JS's peculiarities that don't translate to the other languages and are of zero significance to the greater programming landscape. A good point to note and keep in mind from early on is the importance of not tying oneself to any particular language. By the time your kid turns 20, they can very easily achieve proficiency in 5+ languages. Some people scoff at the idea of being multi-language, because all that time and effort can be spent learning tech stacks and investing in being a good engineer. This is good advice, but only if you want to get a job asap. The tech world moves very, very fast. Learning tools that are trendy today is of very little value to someone who won't step into the industry at least for 5 or more years. Learning languages (preferably of different paradigms), however, makes you a better programmer in every possible way and translates to timeless skills and knowledge that quite literally changes the way you approach every new problem. I'll finish with what I regard as the ideal programming journey: python -> go -> rust -> clojure -> haskell
7
Extremely unlikely, other than that, Python isn't a concurrent language. You can get pretty far using the tools Python gives you, but Go still remains the best mainstream language for your concurrency needs.
7
There are more reasons. Navigating around any Java codebase, reading through stacktraces, fixing null pointer exceptions, understanding inheritance hierarchies - all these require a substantial amount of coffee.
6
They picked Python because of the ecosystem. Chris being experienced in lisps and functional langs I really doubt he would've picked Python had it all boiled down to purely elegance and expressivity.
5
You can get both in Go.
3
They decided not to go in this route. I forgot the reasoning presented but this is pretty much a closed case.
2
lmao
2
try clojure
2
@dingding4898 Yup. You can use Pydantic for all kinds of validation. But it's nothing like a real type system if that's what you're looking for.
1
He didn't say those things in the context of legacy codebases. The discussion was about new people. There's absolutely no reason for new people to 'inherit' legacy unless they really want to.
1
they bought the white nerd voice pack
1
That's not necessarily true for languages with well-designed simpler syntaxes. Languages like Go and Kotlin don't have these issues, for example. Semicolons are only meant for compilers, and can easily be left out of the language if you stay careful about this from the very first stages of design (patching the language with automatic semicolon insertion retroactively may cause issues e.g., JS or might not be possible at all e.g., Dart). Another thing about the C and Python worlds that makes this annoying is the issue of occasional unhelpful error messages. In my experience, Go and Rust have the best errors compared to others and can more often than not pinpoint the exact place where you've committed a syntactical mistake.
1
Everyone who knows more than 3 langs, people contributing to python's scientific stack, companies realizing they need to rewrite services to save money, Guido van Rossum himself: Yeah, CPython is slow Some dude in the comment section of every YouTube video about Python:
1
They are probably going to introduce you to something higher level later on. There are no pure C++ programs any more afaik.
1
You'll love Go.
1
same lmao
1
Most blockchain developers are just app devs (mobile and web) who use blockchain APIs on their apps. Any mainstream language will do fine.
1
F# should've been the main spotlight in .NET. C# is a language stuck in perpetually playing catch-up with F# and running from it's Java induced past.
1
This lack of overlap probably has more to do with being failed by the US's education policies than being men of culture or whatever that means.
1
Safety is actually more important in scientific software, because, you know, people literally die if they mess up. Science-y people recovering from the two-language problem have great options these days. Julia, Go, Rust are all great options, and have seen good adoption in recent years. You should probably hop on the next Rocket that lands on the Moon because you've been out of the world for so long.
1
The JVM is barely under Oracle's thumb any more. Their only source of profit these days are enterprises growing too old to carry their own baggage. Languages like Clojure and Scala are absolute gems and you're the only loser here if you're ignoring them just because the JVM stinks.
1
I really doubt vanilla CPython with be of any relevance once Mojo gets its open-source release.
1