What I think Gru needs for adoption

Users
1

I had been having various thoughts about what Gru needs. So, I want to jot them down here and look for more ideas.

Gru Server

  • GUI: Allow modifying the questions.
  • GUI: Allow inviting candidates, with an email id.
  • GUI: Show report for a candidate on click. This shows the total score, score division by tags, %-ile of this candidate compared to others, rank out of total candidates, and a list of all questions in the order shown with their respective scores etc. This I envision could be a single pane interface, converting to two pane interface. So, on the left one, one can just click, and the right pane would automatically update, without a full page refresh.
  • GUI: In general, the GUI would use Angular or something and be a single page application.
  • Back: Automatically send reminder emails on a frequent basis to candidates who haven’t taken the test.
  • Back: Automatically cancel a candidate’s quiz token after X days, where X is some number defined by a backend config. This config, btw, doesn’t need to be editable by GUI. It’s something one can pass in when running the server.

Gru Client

  • GUI: The entire UI

I think we could use the help of the design team for how the UI should look.

What do people think? @pawan

2

This covers pretty much everything.

  • It also needs some sort of authentication for accessing the admin panel. Maybe we can have just one account that anyone can login with.

With regards to the JS library, heard a lot of positive feedback about React. Angular provides you things out of the box but doesn’t scale well from what I know. It would also depend on the person working on it.

The challenge I feel here is to find the right person. It’s a very interesting project.

1 Like
3

Maybe we could use Github login. Or, maybe something like auth0.

In total, Github, Twitter, and Facebook should cover all the logins. The config file for the server could again hold all the valid login ids.

Also, we could hire 2 interns, if that helps.

1 Like
4

Sure if we find 2 good people. If not, then maybe we just hire one good person. Would require less coordination, communication and monitoring.

1 Like
5

You could create a quiz for hiring a frontend person :smile: . And then just advertise in whatever medium Indians are using for finding talent.

1 Like
6

I think a lot of people (read app developers) would be more comfortable to have and easier to use CRUD backend.

  • Support for RDBMS as storage backend alongside YAML. Not everyone would find it intuitive to manage their question repository on flat files.
  • Some folks have requested for web frontend.
  • Admin to manage the questionnaire corpus / performance reporting. They are important features to quiz @ scale.
7

Right, we plan to shift all that to a SQL database @mohitranka. Performing Update/Delete operations programmatically on a file would be an unnecessary complication.

8

Gathering ideas for the server. Now that we are considering a webapp should we move away from gRPC to using a web server over HTTP?

We could also use something like GitHub - grpc-ecosystem/grpc-gateway: gRPC to JSON proxy generator following the gRPC HTTP spec but that might just complicate things. Whats your advise @mrjn?

9

HTTP + JSON. Yeah, sounds like it.

Seems like a waste of our previous effort, but a different direction definitely requires different choices.

1 Like
10

Sounds good.

Would we still want to have a demo or do we only have the test now? One of the reasons for the demo was that people should have an idea about how the command line client works. Now that we are moving to the web it should be fairly intuitive.

11

I think demo still serves a good purpose. We can hand out a link and let people try it. Would help with adoption.

1 Like
12

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.