I am Naveen, a final year student pursuing Computer Science and Engineering. I have been working with Django for past 3-4 years now. I have worked over 50+ projects in Django and DRM. I have built a complete course too at geeksforgeeks for django-basics. I want to participate in GSOC with a good proposal. Thinking all over, i came over this -
Proposal - Integrate Django with popular Production servers such as heroku, pythonanywhere plus the paid ones like AWS, Azure, etc.
I have been facing this problem for a while now so as to push something from local to production takes a 15-20 minute process. I wish to automate it using manage.py layer so that any fellow developer could push a local Django web app at one command to any of popular servers.
Kindly review and provide guidelines for the same. I have huge interest with Django now and would love to contribute to the community. I have been helping fellowmates through the Django telegram channel as well.
Thanks and Regards
This kind of thing has been talked about a few times, and there is a gap here. (It’s a long way from the Django Tutorial to a deployed app.)
Previously it’s been worried that it’s simply too big a topic, with too many options. So I guess the first step is to show how you’d narrow it down.
There are some common steps: migrate the DB, get the static files in place, get the production settings into the environment, run. Do you have an abstraction in mind here that would encapsulate these?
This is an ambitious project. It’s going to be about the quality of the proposal, showing that what you’re trying to achieve is realistic in the time frame.
Happy to discuss some if you can flesh it out.
First things First, this can be thought of a big topic for once but other side says if we write down the major changes to be made and the entire process that an app covers from local to production. Lots of things are common, as specified - migrating the database, getting static files in place, separating local and production settings, generating procfile, etc.
We can encapsulate this into a single command or somewhat, once we are ready with production ready local part.
and i wanted to check what more can be done to obtain the overall motive of easing out the process of deployement.
and at last i wanted to ask if this can be considered as an valid project for GSOC or i should think something better. Whether this one is worth it ?
If you start from the command end of the story, I worry that it’s more appropriate to tools such as Fabric, or Ansible, or…, rather than Django itself.
Beginning from the existing docs and working upwards though it is a valid idea at least. It’s an area folks really struggle with.
It’s going to depend on the proposal. Explain what you think you can achieve. It needs to be more than just a how-to guide for deploying on a particular cloud provider.
Without concrete suggestions it’s hard to say more.
After messing around, i figured out one more idea for GSOC’2020. This time as i be very clear what i want to achieve from the project and how.
So here it begins -
Problem Statement - Out of 100 developers starting a Django project, 90 of them have to create a Authentication portal, their own USER model for authenticating users. After my recent contributions to Django Users, i found out most of people mess up with defining their user model either way.
What to achieve - Create a Model mapper that can directly be used to create USER model directly, specifying all the required attributes fields, default user model, etc inside the wrapper only. Also, create support for this in LoginView, and create a SignUpView too. Overall, easing the process of creating “accounts” app for every project.
What constitutes success - Creation of a custom USER model, including required forms, etc. via single model interface.
I will plan out how to achieve this after concrete suggestion whether i am in right direction, and if not can you direct me to the right direction on how can i find a innvoative project for GSOC’2020. I have looked into ideas list, also on Django users mailing list. Also, can anyone help me how can i help in resolving tickets, with some basics.
BTW, i am loving Django codebase
@carltongibson Can you help me out here ?
HI @naveenkrnl. I didn’t see this.
This kind of project has long been considered out of scope for the core. There are lots of third-party options, and it’s been considered that they’re best off that way, since there are so many different ways you might go.
As such, I can’t say it would make a good GSoC proposal.
Hope that makes sense.
Thank you for the valuable response. I will try out more ideas for sure