I think of the big problems with Smalltalk is how bad the marketing is. There are few posts which get brought up over and over, and they claim the things that are clearly false ("It can be learned in its entirety within 15 minutes!") or misleading (the one productivity study which is by now decades old.. the claim that it has "simple syntax").
I'd like to encourage Smalltalk advocates to write more "from the trenches" posts. A Haskell is a great comparison: it looks like every few weeks there is a post on using Haskell in production, like [0].. Those posts work much than abstract claims of "tremendous benefits" and "magical language".
Sorry, all these articles are from a single person with the messianic syndrome who named himself "Mr Smalltalk". He absolutely rejects any complaints about his activities (he repeatedly told us that we do not understand marketing) and he will never be able to write something more solid because he actually is not a Smalltalk programmer and does not know it well. He is making a really bad service.
How well does one need to know Smalltalk before they can promote it? I'm a Smalltalk newbie; I've only been using Smalltalk on and off for the past few years. But I absolutely love it!
So I'm telling all my friends and colleagues to give Smalltalk a try. Does that invalidate my efforts?
It contains data for languages like Ceylon (2011), Dart (2011), Elixir (2011), F# (2005), Go (2009), Haxe (2005), and Julia (2012). So, no, not decades old.
Moreover, Haskell is an old language that has only recently come into the limelight. It's considered new-ish.
The study does contain data for languages going all the way back to the 1960s. This is a Good Thing. It means we have a statistically huge sample size which only further supports its validity.
In fact, this is the largest study of its kind. Why would anybody want a small sample size that only covers a few languages from the last 20 years???
If you do a sweeping Google search, you will find many Smalltalk posts "from the trenches" over the years. However, they've done little to elevate the language. I disagree that this is the best way to market a programming language.
Even Haskell hasn't done that well using this approach. With all these "from the trenches" posts, Haskell is still only #45 at TIOBE. It's #26 at PYPL and falling! In terms of pull requests at GitHub, Haskell is #28 and falling! See https://madnight.github.io/githut/#/pull_requests/2019/2.
The following tutorial can be finished within 15 minutes: https://amber-lang.net/learn.html. Thereafter, you will know the entire syntax of Smalltalk-80.
Is there a Python tutorial where you can learn all of Python's syntax within 15 minutes?
Indeed, you can. The first thing to learn here is that the hash (#) signifies a literal symbol. So, #a is a symbol, whereas 'a' is a string (of length 1 and containing the character $a).
The second thing to learn is that the dollar sign ($) signifies a character value, so $b is the value of the character "b" and $c is the value of the character "c".
#() signifies a literal array created at compile time.
Thus, #($a #a 'a' 1 1.0) is an array that contains the character $a, the symbol #a, the string 'a', the integer 1, and the floating value 1.0). Easy peasy.
I'd like to encourage Smalltalk advocates to write more "from the trenches" posts. A Haskell is a great comparison: it looks like every few weeks there is a post on using Haskell in production, like [0].. Those posts work much than abstract claims of "tremendous benefits" and "magical language".
[0] https://news.ycombinator.com/item?id=22308771