Which programming language did you choose and why?

[gaussgunner]

Python is nothing special. It's basically PHP in drag.

It's fine if it's the most direct way to accomplish a small scripting task (just because it's everywhere) but it's just a badly designed language with a lot of clever ideas thrown in, but that doesn't make it good. I don't know why anyone would go out of their way to use it.

 

[PorkyThePaladin]

Until then, why should I spend my time on investigating all these other languages when Python is just fine for my needs...

You do whatever, we couldn't care less. You just made some sweeping generalisations about other languages, which we corrected, then you got butthurt, and made more sweeping generalisations and false claims about Python, which we corrected again.

The only one butthurt here is you, my friend. You've been twisting my statements from the start, because you are so insecure about your languages.

Most devs don't select what language to use. Technical founders, CTOs, etc do. And those guys are pretty smart, and are constantly selecting Python.

There are so many problems with these claims that I won't even start.

Yes, please don't start. You already lost.

Good thing that I am not some 3rd rate outsourced programmer then, so I won't have to interview with you people, eh?

No, you sound like some overconfident American talking out of his ass. Btw, it's an American/Australian company where I work, in Australia. But we have some good outsourced Ukrainian and Polish people too.

The only time Python gets used for anything is by the ops guys to hack together some scripts, and that's it.

Ok, I guess your Strayan outfit that likely runs a platform to track Kangaroo Defecation patterns knows better than Instagram, Pinterest, Reddit, Youtube, Quore, and all the other huge companies that used Python as their main programming language.

 

[gaussgunner]

What other languages do you like besides Python, PorkyThePaladin?

 

[PorkyThePaladin]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Other than that, I like Java more than C/C++ because it's higher level, but it's lower level than Python/Ruby. C was alright for my first language, it's kinda neat to do pointer arithmetic and memory allocation when you are starting out in school, but once you start dealing with real world problems, you want to just focus on business logic and not machine details. At least I do. The real test of software is how it helps the company grow, not how neat it is technically.

I don't have a theoretical problem with functional languages, but I've never seen them used much in real world settings, just remember university types praising them for hours. It seems to me that OOP fits real world problems better than FP in most cases (it's harder to match real world stuff to functions than to objects, so it takes more time), but mostly I am just going by what people use.

 

[Viata]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Sadly. After Smalltalk, Ruby was the only language that did OOP right. I really like it and I find it better to Python in every case. But Python won and there is nothing to change it.

 

[Not.AI]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Sadly. After Smalltalk, Ruby was the only language that did OOP right. I really like it and I find it better to Python in every case. But Python won and there is nothing to change it.

It could be said that Smalltalk won or at least survived to a degree too, in the form of Erlang.

 

[Rincewind]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Sadly. After Smalltalk, Ruby was the only language that did OOP right. I really like it and I find it better to Python in every case. But Python won and there is nothing to change it.

Ruby is one language that I really cannot stand.

 

[Raghar]

Pseudo-code is a cancer. Anyone that had to implement some complex algorithm and said algorithm was written in pseudo-code knows how horrible that shit is. Fuck whoever came up with the idea of writing in pseudo-code.

Better than math equations.

 

[kepler]

Can explain why your Python example is better than my Kotlin and Scala examples? Any would it take less time to write?

The reason why I would pick Python over Kotlin and Scala right now is because Python is one of the most popular programming languages (top 2 every year for a decade)

Then why not using the most popular programming language (JS)? You are dismantling your own argument here. Especially when you are so fixated on:

The real test of software is how it helps the company grow, not how neat it is technically.

JS is insane, super popular and you can develop anything in no time. You sound like someone who should enjoy JS.

 

[PorkyThePaladin]

Can explain why your Python example is better than my Kotlin and Scala examples? Any would it take less time to write?

The reason why I would pick Python over Kotlin and Scala right now is because Python is one of the most popular programming languages (top 2 every year for a decade)

Then why not using the most popular programming language (JS)? You are dismantling your own argument here. Especially when you are so fixated on:

The real test of software is how it helps the company grow, not how neat it is technically.

JS is insane, super popular and you can develop anything in no time. You sound like someone who should enjoy JS.

You keep trying to cherry pick silly 2nd grade level semantics because you have no actual argument. JavaScript is terrible at all the things Python is great at (ease of use, code elegance, programmer productivity, etc). In fact, these days most people use some kind of a framework built on top of JS, beause JS itself is so bad. Its popularity only owns to the fact that it has essentially a monopoly on browser development.

 

[kepler]

JavaScript is terrible at all the things Python is great at (ease of use, code elegance, programmer productivity, etc).

This is all subjective.

Its popularity only owns to the fact that it has essentially a monopoly on browser development.

Well, web dev is so popular and important now. This clearly means that JS > Python.

You keep trying to cherry pick silly 2nd grade level semantics because you have no actual argument.

I had an argument with you a couple of pages ago and you actually won. I just didn't know that we are arguing about who is more retarded.

Funny thing is how you just can't let go and even argue with people that agreed with you. Arbiter alone destroyed you. You had nothing to even leave a dent on my arguments (and I just pointed out your ignorance, nothing bout Python superiority or whatever). But you will keep replying adding nothing to the discussion. You are just a simple troll.

 

[PorkyThePaladin]

I am very glad that you, of all people, you who can't even follow this argument, have decided on who has won it. However, I, on the other hand, have decided to discard your claim as trash, and continue to make you and Arbiter seethe. Seethe along, boys, nothing to see here.

 

[Arbiter]

I am very glad that you, of all people, you who can't even follow this argument, has decided on who has won it. However, I, on the other hand, have decided to discard your claim as trash, and continue to make you and Arbiter seethe. Seethe along, boys, nothing to see here.

 

[Burning Bridges]

The real test of software is how it helps the company grow, not how neat it is technically.

Why not use VBA then? It is already preinstalled on every Windows computer, no compiler in the way, no need to declare variables. Users love it ("why are not all languages like that?!") and can be trained to make small changes themselves.

 

[Hag]

Why not use VBA then? It is already preinstalled on every Windows computer, no compiler in the way, no need to declare variables. Users love it ("why are not all languages like that?!") and can be trained to make small changes themselves.

You may be joking, be this is exactly the role of Excel. With its terrible macro language it may be the most used programming platform in the word, because it provides great UI (infinite I/O fields) and any user can learn in five minutes how to do basic linking and how to google more advanced formulas. Plus the ability to shit tables, graphs and pivot tables.
I have no data but I would say Excel sheets are the most commonly used internal tools in many companies.

 

[PorkyThePaladin]

The real test of software is how it helps the company grow, not how neat it is technically.

Why not use VBA then? It is already preinstalled on every Windows computer, no compiler in the way, no need to declare variables. Users love it ("why are not all languages like that?!") and can be trained to make small changes themselves.

Because you are conflating easiness with ease of use/productivity. VBA is very easy but doesn't give you that much power, so you are stuck coding simple shit. Python gives you the power to do anything, from billion daily hit websites to scientific tools to automation/scripts, while also being so high level that the programmer can churn out code many times faster than in most other languages. To do this, it trades performance and also gives enough freedom/flexibility that inexperienced programmers/large teams/etc can drown in it. But in the right hands, it can let you do ridiculous stuff.

 

[Rincewind]

Why not use VBA then? It is already preinstalled on every Windows computer, no compiler in the way, no need to declare variables. Users love it ("why are not all languages like that?!") and can be trained to make small changes themselves.

You may be joking, be this is exactly the role of Excel. With its terrible macro language it may be the most used programming platform in the word, because it provides great UI (infinite I/O fields) and any user can learn in five minutes how to do basic linking and how to google more advanced formulas. Plus the ability to shit tables, graphs and pivot tables.
I have no data but I would say Excel sheets are the most commonly used internal tools in many companies.

Actually, Linus Torvalds cited VBA as one his favourite languages like 10 or more years ago in an interview, I distinctly remember that. You can look it up. (Not that it matters much what he likes, but I always found that interesting.)

The Excel scripting language (and probably VBA) hits a local optimum; it's the #1 tool I reach for when I need to experiment with some formulas, tabular data, etc. (well, I don't specifically use Excel, but either LibreOffice or Google Spreadsheets, but still the same idea). It's pretty much the same deal you get with stringing together Unix commands in a bash script; quick'n'dirty for sure, but gets relatively simple jobs done faster and with less effort than any other solution (BUT never let your script go over 100 lines or so...)

Same deal with Python; I actually do use it from time to time to write some web scraper or some shit like that. There's just tons of useful libraries out there for Python to do that kind of stuff, so you can hack together something working quickly in a few hundred lines.

 

[Rincewind]

[Python] let you do ridiculous stuff.

Finally something that we agree on!

 

[Not.AI]

Speaking of high level languages ... Pharo anybody?

: >

 

[PorkyThePaladin]

[Python] let you do ridiculous stuff.

Finally something that we agree on!

It's funny how praise for Python triggers you so easily. Must be hard to live in a world where Python is always #1 or #2, and keeps on growing...

 

[gaussgunner]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Wasn't there a big split between 1.8, 1.9, and some alternative implementation? I remember having to choose between slow, unstable, and vaporware. Documentation was a mess, there were too many compatibility problems taking up our time, so we just dropped it at work.

Ruby is a difficult language to optimize, that's the price of being so dynamic and flexible. That was a big problem around 2010 when growing tech companies were trying to scale up Rails backends, so they switched to performance languages like Scala and Go.

Other than that, I like Java more than C/C++ because it's higher level, but it's lower level than Python/Ruby. C was alright for my first language, it's kinda neat to do pointer arithmetic and memory allocation when you are starting out in school, but once you start dealing with real world problems, you want to just focus on business logic and not machine details. At least I do. The real test of software is how it helps the company grow, not how neat it is technically.

C was your first language? That explains a lot. It's for pros not beginners. The only worse first languages are C++ and Java, the standard ones taught in universities since 1995.

So besides C and C++ (which you haven't used for anything serious) I guess you know Java, Ruby, Python, and maybe some PHP or Javascript. These are dirty high-level languages geared toward business productivity. Not high performance, high reliability, realtime, lightweight, microcontrollers, flight controls, medical devices, etc. You've never even made a video game right? At least not without Unity or something?

I don't have a theoretical problem with functional languages, but I've never seen them used much in real world settings, just remember university types praising them for hours. It seems to me that OOP fits real world problems better than FP in most cases (it's harder to match real world stuff to functions than to objects, so it takes more time), but mostly I am just going by what people use.

FP and OOP are just programming styles you can use together or separately in any general purpose language. Professors and "gurus" who religiously evangelize OO or FP languages are just idea guys with a lack of real world experience.

 

[PorkyThePaladin]

I liked Ruby quite a bit, cause it's very similar to Python in terms of usability, productivity, etc, but it lost out for whatever reasons and is now almost gone.

Wasn't there a big split between 1.8, 1.9, and some alternative implementation? I remember having to choose between slow, unstable, and vaporware. Documentation was a mess, there were too many compatibility problems taking up our time, so we just dropped it at work.

Ruby is a difficult language to optimize, that's the price of being so dynamic and flexible. That was a big problem around 2010 when growing tech companies were trying to scale up Rails backends, so they switched to performance languages like Scala and Go.

There are always trade-offs in the real world. And yet many start-ups still go with languages like Python and Ruby to develop their brilliant platforms, and then, once those platforms are huge, they might in some cases switch to more higher performing languages, but of course by that point, the ones coding are route programmers. So brilliant founders -> Python/Ruby, code monkey -> Java/Go/Scala/C#, etc.

Other than that, I like Java more than C/C++ because it's higher level, but it's lower level than Python/Ruby. C was alright for my first language, it's kinda neat to do pointer arithmetic and memory allocation when you are starting out in school, but once you start dealing with real world problems, you want to just focus on business logic and not machine details. At least I do. The real test of software is how it helps the company grow, not how neat it is technically.

C was your first language? That explains a lot. It's for pros not beginners. The only worse first languages are C++ and Java, the standard ones taught in universities since 1995.

So besides C and C++ (which you haven't used for anything serious) I guess you know Java, Ruby, Python, and maybe some PHP or Javascript. These are dirty high-level languages geared toward business productivity. Not high performance, high reliability, realtime, lightweight, microcontrollers, flight controls, medical devices, etc. You've never even made a video game right? At least not without Unity or something?

What you are saying is a myth. There is nothing inherently more difficult about C versus Python. Any person that can successfully program in one of them, can also do so in the other. It's just a question of allocation of time. In C, you will spend much of your time thinking about memory addresses and allocation and other low level machine crap, in Python, you will spend that time thinking about problem domain logic.

And who on earth would want to work on flight controls and medical devices? I mean somebody's gotta do it, but god that sounds awfully boring.

I don't have a theoretical problem with functional languages, but I've never seen them used much in real world settings, just remember university types praising them for hours. It seems to me that OOP fits real world problems better than FP in most cases (it's harder to match real world stuff to functions than to objects, so it takes more time), but mostly I am just going by what people use.

FP and OOP are just programming styles you can use together or separately in any general purpose language. Professors and "gurus" who religiously evangelize OO or FP languages are just idea guys with a lack of real world experience.

Converting real world logic into functions takes more time than into objects, imho.

 

[Not.AI]

I guess I'll larp as the rabbi in this discussion and sort of agree and sort of disagree with all sides at the same time.

So brilliant founders -> Python

Re: Startups.

Basically most startups begin with limited resources and time and it is easiest and fastest to find Python devs to hire than in the majority of other languages. That has produced somewhat of a positive feedback and the tech sector lives and breathes path dependence like that.

Most founders are pragmatically indifferent to the language used. Most projects involve a stack with several languages, as timely and as needed.

Hiring is the largest early cost.

The firm, whose hierarchy is still relatively flat, requires higher level team members than usual to spend time interviewing candidates because the firm is not large enough to delegate that. Time they could have spent getting their critical work done.

Professors and "gurus" who religiously evangelize OO or FP languages are just idea guys with a lack of real world experience.

Also not quite the case. Sometimes it is, yes.

But often they are significantly better developers, but they just happen not to have, for historical reasons, ownership of any large firm that can hire so many people that it overcomes trends created by the past.

If a language is somehow 10X more productive in every sense, but there are few tools in it already and few people trained in it, onboarding is too expensive. Either too slow or too unreliable. Risk is a cost. And that establishes or reinforces path dependence.

Spoiler

Path dependence in the software space (postindustrial space) has led to much slower technology growth compared to typical, industrial style growth. (Notice how fast, for example, Tesla or any purely industrial style firm can expand in physical output. Because that sector has much more limited path dependence besides project survival at the start.)

 

[gaussgunner]

There are always trade-offs in the real world. And yet many start-ups still go with languages like Python and Ruby to develop their brilliant platforms, and then, once those platforms are huge, they might in some cases switch to more higher performing languages, but of course by that point, the ones coding are route programmers. So brilliant founders -> Python/Ruby, code monkey -> Java/Go/Scala/C#, etc.

What makes you think that? Projection?


Just because you're a code monkey in a typical failing startup run by a marketing grifter doesn't mean highly successful startups are like that. They start with good programmers and replace them with monkeys as the company goes to shit. Good programmers don't work for inferior programmers if they can help it.

 

[Rincewind]

TIL: higher performing languages are used by code monkeys.

Checks out.