Get users to publish their dandisets

[bendichter]

In the meeting today, we discussed our desire to have more of the draft datasets transition to published datasets. We generally believe that many contributed datasets are open and left in draft mode simply because the user does not realize that they are supposed to hit publish. They do not understand the mechanism, the fact that it is best practice, and the advantages to themselves and others (making it persistent, versioned, and citable with a DOI). How do we get people to hit publish? We already have a page about the advantages here. Some ideas were proposed:

1. Inform - improve the docs page to make it more clear that publishing a dataset is the proper way to share data on DANDI
2. Stink - put a banner in the page warning a user that the dataset has not been published
3. Nudge - send an email to the contact person recommending that they publish the dataset
4. Stick - threaten to delete the dataset on a reasonable schedule

Any other ideas?

[bendichter]

This came up again in the scientific core meeting today. Let's move forward with Nudge. Here's a draft of an email which could go out to the contact person of every dandiset that has been stable for X+ days (I would suggest 30, but I'm flexible): https://docs.google.com/document/d/1A9NcrkqqUlJkfUCum5d6YQ0yJ2H7zUW423BjDgW4ias/edit?tab=t.0

[bendichter]

Full res image:

[bendichter]

@satra I would estimate this email, if timed appropriately, would reduce our forever-draft Dandisets by 50-60%. @waxlamp is this something you think would be straightforward to pull off? Are there hidden complications here?