cosmir / openmic-annotator

Annotation framework for annotating data for OpenMIC
MIT License
56 stars 1 forks source link

Authenticated user sessions w/backend #9

Closed ejhumphrey closed 7 years ago

ejhumphrey commented 8 years ago
alastair commented 8 years ago

How do you want to manage authentication? Do it ourselves, or pass off to oauth (facebook,google,github,twitter?)

Can any authenticated user submit audio and annotations or are groups/permissions required?

ejhumphrey commented 8 years ago

per this, I'm planning to use OAuth via other services. Google makes it wicked easy, so that's probably a fine place to start.

Can anyone submit audio? No, we'll restrict that to at least the dev group, and from there maybe even a few select individuals. This might be easiest by actually not having an audio/ingest endpoint, instead having machinery that must be properly credentialed locally.

Can anyone submit annotations? Yes, once logged in. If it comes to it, we can scrub data from bad actors, or lock them out of the system if / when identified.

julian-urbano commented 8 years ago

If we plan on implementing the credit system by which annotations give you points and system runs consume them, we should be able to link several user accounts together as part of an org account or something like that, so they can share credit.

ejhumphrey commented 8 years ago

A "credit system" is a good thing to be aware of / have on the roadmap, though I might argue against it for the first iteration.

Regardless, it'd be great if someone could document a proposal about how such a system would work, what problems it solves, and other such thoughts ... 😉

dmcennis commented 8 years ago

Love the credit system idea. It will help keep analysis costs in reach for PhD students and still keep the door open for those that would rather donate cash than provide free labor.

On Wed, Aug 31, 2016 at 11:57 AM Eric J. Humphrey notifications@github.com wrote:

A "credit system" is a good thing to be aware of / have on the roadmap, though I might argue against it for the first iteration.

Regardless, it'd be great if someone could document a proposal about how such a system would work, what problems it solves, and other such thoughts ... 😉

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/omec/bedevere/issues/9#issuecomment-243811103, or mute the thread https://github.com/notifications/unsubscribe-auth/AFItP15d5eaT5gYg57GlAgAexpkSp0JTks5qlaRbgaJpZM4Jvlr4 .

ejhumphrey commented 7 years ago

in the course of progress elsewhere, there are a few options to consider:

These bullets are ordered in increasing complexity, so it probably makes the most sense to work top-down until we find a solution that works.

ejhumphrey commented 7 years ago

0a8ef65 takes care of a big chunk of this, not far off from a PR

ejhumphrey commented 7 years ago

this is done as of #48