How to fix "This profile has no local account"

I have various users set up in our Indico (2.2.8) installation. I created some of these users manually, for example so that I could give them “reviewer” status in the “call for abstracts” section of the site.

However, if any of these users attempt to log in, eg using the ‘forgot password’ pathway, then they get an error message saying “This profile has not local account”. Given that I manually created these users, there seems to be no way that those people can fix the problem.

The only way I can see to fix such broken users is to go to Administration→User Management→Users then click Accounts and type in a username and password to create an account for that user.

This seems like a broken process to me – why can I not create a users and assign them with a role, using only their email address, and then leave it up to them to select a user name and password, and so on?

I imagine that the expected approach is that new users should register themselves, however that is not the typical interaction process that will occur for a conference. Normally some reviewers will not have been involved as users of the conference system until such time as they are invited to give a review.

(I note that I am seeking reviews of abstracts here, and not full papers… our conference has only a single-step process with just an extended abstract… perhaps I should be doing things some other way…?)

During manual account creation, there is an option to set account details. This would have created the local account, just use some random password there.

And many conferences can live happily with reviewers self-registering there account (at least in the physics/chemistry/natural sciences where indico evolved).

We found the option to set account details, but it seemed a bit cumbersome. I feel that I should be able to just give an email address for a proposed reviewer, and they should get the chance to type in all their relevant details upon receipt of an email. However, I can’t assign a reviewer without typing all these details myself, as far as I can tell.

In some cases, the ‘person’ record is already in the system because they are a coauthor, and I can assign them as reviewer, but they may never know about it, because there is no ‘user’ associated with the ‘person’ record. I might be misunderstanding here, but I think I’m not; I think there is still something a little bit cumbersome about the ‘onboarding’ process for reviewers, at least during the ‘call for abstracts’ stage.

When their emails have been entered as ‘people’ but with no user record created, those people, later attempting to log in, start creating new user records with other (personal) email addresses, and it all becomes a bit messy.