Re: [Treebase-devel] harvesting

[William P. <wil...@ya...>]

On Feb 1, 2012, at 11:01 AM, Hilmar Lapp wrote:

> Yes, there'd have to be admin privileges that override all other access permissions. Are you saying that such a role does currently not exist in TreeBASE? Introducing it would then likely mean a quite significant change, I would imagine. Maybe this could be accomplished in the short term by having a "super-password" that allows logging in as any user.

Admin users can edit any existing submission, even if it does not belong to them. So that's not a problem.

The only "missing" functionality is when the admin person needs to create a new submission for someone else -- in some cases because they can't seem to do *anything* for themselves. In other cases because I discover that a submission is corrupt in some way, and the most efficient solution is for me to create a new submission for them, re-upoad non-corrupting data, and then delete the old one. It's in these cases that I need to login as the user.  A super password, like the one you mention, would accomplish this -- but another feature would also accomplish it while providing additional benefits: the ability for the admin person to move a submission from one user to a different user. 

I don't think it would be significant work -- at least the SQL is pretty minor -- just two updates within the same transaction: 

UPDATE submission SET user_id = <new user> WHERE study_id = 10; 
UPDATE study SET user_id = <new user> WHERE study_id = 10; 

So, instead of spoofing another user, the Admin person can create a new submission in his/her own account, and then transfer it to the other user. Also, this would solve other issues, like the request from Susanne Renner (below). She wanted to edit/update some citations, but these and been submitted by prior students who have since moved on. So she wanted to "repatriate" these submissions under her own account username -- something I could not do for her. 

Another example is with the student worker that Karen has just advertised. This student will be creating new submissions for data belonging to other scientists, so therefore parking them under this students own account. Which is fine, but maybe in the future these other scientists will (finally) be creating their own user accounts on TreeBASE, and they may (like Susanne Renner) want to repatriate their own submissions from the student account. 

On Feb 1, 2012, at 11:01 AM, Hilmar Lapp wrote:

> The assignment to Vladimir, if it is still there, was from the time of the public release and post-release. Where we're currently at with grant funding I don't see much sense in him investing the time in digging through the code, acquiring enough understanding of it, and then implementing it. So consider this task unassigned. Is Harry in a position to take it on?

Harry isn't funded for TreeBASE, so while minor tweaks might be okay (out of the kindness of his heart), major thinks like redoing the authentication system need to be budgeted.  But perhaps this is the ideal use for the emergency funds (the one ones that Karen is now dipping into regarding the student worker) ?

bp


> From: "Susanne Renner via RT" <he...@tr...>
> Date: January 5, 2012 6:26:17 AM EST
> To: undisclosed-recipients:;
> Subject: [treebase-help #12703] Up-dating citations 
> Reply-To: he...@tr...
> 
> 
> Thu Jan 05 06:26:17 2012: Request 12703 was acted upon.
> Transaction: Ticket created by re...@lr...
>       Queue: Treebase-help
>     Subject: Up-dating citations
>       Owner: Nobody
>  Requestors: re...@lr...
>      Status: new
> Ticket <URL: https://help.nescent.org/Ticket/Display.html?id=12703 >
> 
> 
> Dear Bill,
> 
>  All but one of my "in progress" submissions are now solved.
> 
> Could I get permission to update the citations of all the matrices associated with my paper? There are 19 or 20, many not up-dated.  I understand that only the original submitter can make changes, but often these were graduate students, who have long moved on.
> 
> Best regards,
> 
> Susanne