While I find this comment section fascinating and will read it top to bottom, I can't help but make an observation that such articles often comply with:
+-------------------------------------------------+
| People really do love Python to death, do they? |
+-------------------------------------------------+
I find that extremely weird. As a bystander who never relied on Python for anything important, and as a person who regularly had to wrestle with it and tried to use it several times, the language is non-intuitive in terms of syntax, ecosystem, package management, different language version management, probably 10+ ways to install dependencies by now, subpar standard library and an absolute cosmic-wide Wild West state of things in general. Not to mention people keep making command-line tools with it, ignoring the fact that it often takes 0.3 seconds to even boot.
Why would a programmer that wants semi-predictable productivity choose Python today (or even 10 years ago) remains a mystery to me. (Example: I don't like Go that much but it seems to do everything that Python does, and better.)
Can somebody chime in and give me something better than "I got taught Python in university and never moved on since" or "it pays the bills and I don't want to learn more"?
And please don't give me the fabled "Python is good, you are just biased" crap. Python is, technically and factually and objectively, not that good at all. There are languages out there that do everything that it does much better, and some are pretty popular too (Go, Nim).
I suppose it's the well-trodden path on integrating with pandas and numpy?
Or is it a collective delusion and a self-feeding cycle of "we only ever hired for Python" from companies and "professors teach Python because it's all they know" from universities? Perhaps this is the most plausible explanation -- inertia. Maybe people just want to believe because they are scared they have to learn something else.
I am interested in what people think about why is Python popular regardless of a lot of objective evidence that as a tech it's not impressive at all.
I've started using micropython to interact with embedded arm chips, it's a revelation to interact with hardware through a REPL instead of compiling, transferring, resetting, and writing print statements to serial...
This talk by the creator of micropython [0] gives his reasoning for why to implement python on microcontrollers despite it being hundreds of times slower than C. Starts @ 3:00
- it has nice features like list comprehension, generators, and good exception handling
- it has a big, friendly, helpful community with lots of online learning resources
- it has a shallow but long learning curve. It's easy to get started as a beginner, but you never get bored of the language, there's always more advanced features to learn.
- it has native bitwise operations
- has good distinction between ints and floats, and floats are arbitrary precision, you're not restricted to doubles or even long longs. (I'll add that built in complex numbers is a plus)
- compiled language, so it can be optimized to improve performance
Ha, that actually looks pretty cool and good tech. I'd wonder if that can even be called Python anymore but still, it looks to be very useful. Thanks for the tidbit!
Completely agree that objectively, Python has really bad underlying tech.
Emotionally though (once you have the environment set up), it’s just such a breeze to write it. It’s like executable pseudo code with zero boilerplate. You can focus purely on the algorithms and business logic. Compared to many other languages the line count is often 50-80%, even if you include type annotations! This doesn’t only apply to plain imperative code, using the dynamic features you can also turn it into your own DSL where needed.
Then there is obviously the huge eco-system around it, there is not a single service, file format or database that doesn’t have a good python library for it. While go might have equally wide library choices, I wouldn’t be so sure about nim, go on the other hand has a lot of other wtfs even though it provides a lot of good fresh tech.
Would I use it for a big service with potentially lots of performance requirements? No. But there is no doubt why it’s so popular. For many applications where the the outcome of the program is more important than the performance or environment, like glue code, simple intranet applications or exploratory coding, it is still the perfect choice. You also have to consider what it is replacing, often the alternative would be even worse; bash-scripts, Excel or Matlab.
Another way to put it is that it’s a very good Swiss Army knife that is good at everything but not best at anything.
> Emotionally though (once you have the environment set up), it’s just such a breeze to write it.
That's no different than the JS fans saying "hey, if you have the exactly right versions combo of Node.JS, npm and webpack, everything works fine!". I almost never manage to install anything via `pip install`. I just cross my fingers and am not at all surprised when it doesn't work. Not for a lack of trying, mind you, I've fiddled with separating Python 2 and 3 environments, tried a few package managers etc.
If I can't install Python + a package manager and I can't then install any random tool that proudly says "we're just a `pip install` away" in their GitHub repo then to me that's a failed ecosystem relying on inertia and Stockholm Syndrome. ¯\_(ツ)_/¯
> It’s like executable pseudo code with zero boilerplate.
Is it? I've seen some pretty horrible constructor boilerplate that managed to look almost as alien as Perl. Also who decided that methods starting by underscores is a solid convention lol. Of course bad programmers will manage to butcher every language, think we all agree on that, but as I get older I appreciate languages that don't allow bad coding in the first place (or realistically, limit it as much as they can).
> You can focus purely on the algorithms and business logic.
Not my observation. I know a few data scientists and a few people who just learned Python out of desperation related to how mundane and repetitive their jobs are. They all had to wrestle with OS-specific quirks -- that a stdlib absolutely should abstract away -- the same package management woes as me, and libraries they need subtly breaking because apparently they only work well in Python 3.6 and not after (one random example).
I get the sentiment and I wanted to believe that such a language existed a while ago
but I am just not seeing it. Foot-guns and all, you know the old sayings.
> For many applications where the the outcome of the program is more important than the performance or environment, like glue code, simple intranet applications or exploratory coding, it is still the perfect choice.
I am still looking for such a tech for myself as well because my main choices of languages aren't a super good fit for tinkering. But so far I haven't seen Python filling that niche. Too much minutiae to handle.
> You also have to consider what it is replacing, often the alternative would be even worse; bash-scripts, Excel or Matlab.
Yeah that's a real problem, absolutely. For now I always got away by just learning an insane amount of CLI / TUI tools and always assembling them together just enough with bash/zsh scripting to get my task done... but that approach has deficiencies as well, and will not work forever.
> Another way to put it is that it’s a very good Swiss Army knife that is good at everything but not best at anything.
To add to the analogy: it also starts getting rusty and doesn't work as reliably as before but grandpa will let anyone replacing it only over his dead body.
(And I semi-mockingly call people "grandpas" as a 42 year old who is supposed to be conservative but I find it extremely amusing how many 28-year olds I've met that are more conservative than me and my 69-year old mother.)
How many of the better languages have equal or better readability than Python? IMO, that's the #1 reason for its continued popularity. Python is not full of parentheses, like a lisp, nor is it full of semicolons and brackets for bookkeeping, like most other C-style languages.
If that's a blocker to productivity to anyone I'd seriously argue their programming prowess. Especially nowadays with LSP auto-formatting, snippet management, and such.
Syntax is a subjective preference. I didn't like that fact for the longest time but it's a fact regardless.
Why would a programmer that wants semi-predictable productivity choose Python today (or even 10 years ago) remains a mystery to me. (Example: I don't like Go that much but it seems to do everything that Python does, and better.)
Can somebody chime in and give me something better than "I got taught Python in university and never moved on since" or "it pays the bills and I don't want to learn more"?
And please don't give me the fabled "Python is good, you are just biased" crap. Python is, technically and factually and objectively, not that good at all. There are languages out there that do everything that it does much better, and some are pretty popular too (Go, Nim).
I suppose it's the well-trodden path on integrating with pandas and numpy?
Or is it a collective delusion and a self-feeding cycle of "we only ever hired for Python" from companies and "professors teach Python because it's all they know" from universities? Perhaps this is the most plausible explanation -- inertia. Maybe people just want to believe because they are scared they have to learn something else.
I am interested in what people think about why is Python popular regardless of a lot of objective evidence that as a tech it's not impressive at all.