GSOC 19 PROPOSAL IDEA

[Chinmay Relkar]

Hi,

       My name is Chinmay Relkar. I'm a final year undergrad from India. This is about my GSOC proposal idea. 

       While playing with django, I've always come across the absence of bidirectional ManyToMany feature ticketed at

code.djangoproject.com/ticket/897. 

I would like to work on this feature as a part of the GSOC program. 

        

       To me this feature feels a bit small for a 12 week program. So here's something I would like to add on:

       

        While adding a Model item in the admin interface with a ManyToMany field, we get a MultiSelectWindow to choose multiple items of the related model. This UI isn't very intuitive.

        

Taking inspiration from the Chips Component defined in Material Design Guidelines at https://material.io/design/components/chips.html I would like to implement FilterChip for the same multi select window. The example of which is attached below. 

This is not a final proposal. I hope I'm not too late to share this here. 

I'm eagerly waiting for a response of feedback and suggestions. 

About my experience with Django, I've been working with it for past 2 years. Completed a total of 3 projects, one of which is live at acsinet.net with a heavily customised admin panel.

Talking about my level of expertise, I'm not a master to be true. But I can customise any component of Django according to the stated requirements. 

This is a very rough sketch of the proposal. Any suggestions, remarks or comments will be of great help. 

Hoping to get a feedback soon. 

Regards,

Chinmay Relkar

[Carlton Gibson]

Hi Chinmay. 

Thanks for your interest. A 13 year old ticket would certainly be a nice-to-close. 🙂

Some quick thoughts: 

1. We're close to the deadline here, so can you turn this into a final proposal? (Look at the example for a previous year: https://gist.github.com/chrismedrela/82cbda8d2a78a280a129 — we need something like this.) 
2. You've begun well telling us who you are. But tell us more in your proposal. We don't know you already and need to be confident that you'd be likely to succeed in your project. 
3. We already use select2 for the autocomplete in the admin. https://select2.org. This would integrate well with that by the looks of it. 
   
4. Would this take 12 weeks? If not there are lots of Admin tickets — I'm sure we can help find some, but any related target ticket you identify are all for the good. 

Time is short here. :)

Kind Regards,

Carlton