DEV Community

Discussion on: Python For Web Development

Collapse
 
mburszley profile image
Maximilian Burszley

Out of curiosity, why recommend Django over Flask?

Collapse
 
minchulkim87 profile image
Min

Just a quick thought on learning web dev vs actual web dev:

Maybe Bottle is the best to start off with as it is the easiest. It inspired Flask and similar enough that learning Bottle will transfer very easily to learning Flask.

Then, perhaps, FastAPI is the python framework to learn next. Again, very Flask-esque, but just a more modern implementation.

Collapse
 
steelwolf180 profile image
Max Ong Zong Bao • Edited

Hmm... It depends I'm a person that is inclined to depend on the use case and the types of jobs that is available in your country. So to learn that particular technology and practices.

So far for jobs wise the popular choice I see is either Flask or Django as a must if you are looking for anything related web development in python.

Therefore I'll pass on bottle but definitely FastAPI is in my radar as I had heard praises of it. So I'm thinking of learning it as well in the future and write a article on it.

Given that, I'll give a higher priority on other topics like learning a frontend web framework, CSS grid or Flexbox, GraphQL, typescript or ES6, cloud providers, TDD, clean code, design patterns.

Lastly with soft skills related topics to improve one's ability to be a better developer.

Thread Thread
 
minchulkim87 profile image
Min

Definitely. For jobs, you'd need to go with what is popular in the field/industry/area.

But "learning" though... Bottle wouldn't take more than a day to "learn". Not even an hour, even for a beginner. Sure having the mastery to make something deployable takes a lot of practice in any framework or tools.

But I am talking about the pedagogical value here. Bottle won't require the learner install much or do much project set up or configuration at all.

Within minutes of starting, even to an entirely new beginner, you can get to an "ah that's what you mean by a backend", and "ah that is what an api is" and so on.

Much in the same way that Svelte would be a more gentler intro to the front end than React or Angular are for a beginner. Even though, eventually you'll need to learn the tool you need to get a job or do a job.

Anyways, Flask "hello world" is simple enough.

Thread Thread
 
steelwolf180 profile image
Max Ong Zong Bao

Yup Flask's hello world is simple enough and it won't take less than 1 hour to do it.

In terms of learning method wise, the best Flask or Django tutorials that I was watching uses bootstrap to cover the front-end portion using Django or Flask's templating engine.

Instead of diving deep in using a front-end web framework like Svelte, React, Angular or Vue. Since there are tons of beautiful and well-maded templates Creative Tim.

Which you can modify those pre-made templates to get what you want that any Flask or Django developer with a good foundation in CSS, HTML, Javascript can do it.

Collapse
 
janmpeterka profile image
Jan Peterka

(disclamer: I work in Flask, never wrote anything in Django, but have some idea about how it works - also work with Rails, which I guess is a similar beast)

I would guess it's because of simplicity of making basic app. If you main goal is to provide web interface for data manipulation (dashboards, APIs), there's little need to understand how it works and write everything yourself (as you will probably do when starting in Flask until you find all the right plugins and/or write abstractions you need).

Collapse
 
steelwolf180 profile image
Max Ong Zong Bao

Yup it's actually a replica of Ruby for Django. Which is why data scientists will work with Flask more than Django.

Due to its huge learning curve that follows to just to understand it and use it fully because you are really going past the microservice and API aspects of it.

Thread Thread
 
mburszley profile image
Maximilian Burszley

Are you saying flask has a huge learning curve while also advocating Django? Or am I misunderstanding? I've done professional work in both; flask was significantly easier to get people productive in fastest but I did lay solid foundations following good patterns and recommended extensions like flask-talisman and co.

Thread Thread
 
steelwolf180 profile image
Max Ong Zong Bao

Ahh.. its my bad I wrote wrongly. I was referring to Django for steep learning curve.

To me it is more like of a overkill to use Django. Which Data Scientists just prefer to get it done with a quick & dirty solution.

Depending on who you talk to they might not even know of the concept of clean code or TDD.

When they are building a data product as it is not part of their job to do it.

Collapse
 
steelwolf180 profile image
Max Ong Zong Bao • Edited

Django has security built-in by default.

You do more with less, without the need to code everything by yourself.

Because of their batteries included philosophy like Object Relationship Mapping (ORM), customisable Admin Panel, Account Management with a very wide 3rd party ecosystem.

Which if you want to do it yourself, you need to explore multiple types of python packages.

Don't get me started in projects structures for Flask. Which the last good option was to use cookiescutter to help you set it up because there is literally 2 or 3 way to build Flask that differs greatly.

Which you need to have a consensus to structure your project.

In a way so that it is easy to maintain but it requires your understanding of concepts to abstract parts of your code and manage the project structure correctly.