Editing Fossil Specimen Data

[Ksepka]

Similar to locality problem: I botched one specimen number entry by making its collection "PIN" and specimen something like "PIN 999" (not the real number). This results in the number being displayed as "PIN PIN 999". I was not immediately sure how to correct that without spawning a separate fossil (PIN 999 and PIN PIN 999, both with the same information), but that seems like it just creates a mess for later.

Proposed solution: Is it possible to add a "Manage Fossils" option for Admins, similar to the "Manage Calibrations"? To be clear this would be to allow access to specimens, rather than fossil species.

[kcranston]

This is certainly possible. It will require design + feedback + implementation + testing + bug fixes (i.e. more than a few calendar days, even if the code changes are only measured in hours).

[jimallman]

The fossil editor are ready for review and testing. Click Manage Fossils on the site's Administrative Dashboard to see the "core" fossil records, what calibrations are linked to each, and tools to edit or delete them.

I've added hints about which fields can be edited here vs. in the calibration editor, as well as safeguards against deleting a fossil that's still linked to one or more calibrations. I've tested new, modified, and deleted fossils using my usual test calibration and fossils in the bogus 'FOO' collection.

EDIT: fyi, i also have a snapshot of the database from earlier this afternoon, just in case we experience data loss.

[jimallman]

UPDATE: I've added an editor for fossil collections as well, since this addresses the main issue of incorrectly entered data. This is a dirt-simple editor that allows updates to the collection's acronym and full name, with the usual "management" page; see Manage Collections in the Admin Dashboard.

This includes reporting of related fossils, with these consequences:

• Attempts to delete a collection that's assigned to 1 or more fossils will be blocked until the admin deletes those fossils or (more likely) re-assigns them to another collection.
• When a collection's acronym is changed, the new acronym will be applied to any fossil that was using the old one.

[jparham]

If someone mis-enters an acronym that matches an old acronym (possible) this could lead to problems where correct acronyms are being affected. Is there a way to escape this pitfall? Is it necessary to have the "new acronym ... be applied to any fossil that was using the old one."

[jimallman]

Is it necessary to have the "new acronym ... be applied to any fossil that was using the old one."

I believe so; otherwise we'll end up with fossils that are clinging to an old / unused acronym.

In the new collection editor, a proposed acronym change is checked against the existing list before submission and rejected if it's already found. I'll check to make sure the calibration editor is doing the same.

[jimallman]

I'll check to make sure the calibration editor is doing the same.

This was not checking for duplicate acronyms, but now it blocks a new fossil submission if a duplicate is entered for a "new" collection.

[jparham]

Checking duplicate acronyms should fix it. My query was because many of these acronyms are VERY similar so I could imagine someone entering a wrong one and the while fixing it changing other entries unintentionally. Ancillary to this is the issue that some acronyms are shared among institutions (e.g., MNHN). But I think this won't be a big issue and if database checks for duplicates we can annotate identical acronyms upon entry.

[jimallman]

EDIT: Here I'm talking about acronyms shared across institutions, as described by @jparham above.

Ah, too bad about duplicate acronyms! Yes, with the new test for duplication, you'll be forced to modify all but one of the matching acronyms.

[Ksepka]

I've figured out a (probably dumb) way to fix the duplicate acronyms and applied it to the problem of having BMNH and two separate NMHUK entries. All of these should be merged (because BMNH was officially changed to NMHUK and we had one NMHUK for "Natural HIstory Museum" and one for "The Natural History Museum".

What I did:

1. Attach each of the 5 fossils that were attached to BOTH NHMUK collections to a dummy collection ("XXXX")
2. Delete one NMHUK.
3. Re-assign the 5 fossils to the remaining NHMUK.
4. Delete the dummy collection.
5. Re-assign the BMNH fossils to NHMUK.
6. Delete BMNH.

Now everything is squared away for this set. I can do this for the other cases of duplicate collections. Its not much work and we should never have to do it again now that duplicate abbreviations are banned.

So, I think Manage Collections is good - now will check the Manage Fossils part of this.

[jimallman]

I've figured out a (probably dumb) way to fix the duplicate acronyms and applied it to the problem of having BMNH and two separate NMHUK entries.

Actually, that's a great solution given the current constraints. Of course there are still a few duplicates in the system, here are the acronyms and number of records for each:

• {EMPTY STRING} (5)
• DPC (2)
• FOOZ (2)
• SDSNH (2)
• SMF (2)
• UCMP (2)

Rather than do the "Ksepka Shuffle" (above), I can clobber these dupes in a one-time operation through the MySQL command line. Just let me know which IDs to keep (or discard) for each acronym above. You can see these IDs on the Manage Collections page. Again, please be clear about whether you're providing the IDs to keep, or the ones to discard.

[Ksepka]

Thanks Jim.

EMPTY STRING: delete all 5 DPC: delete 87 Duke University
FOOZ: delete both (these are your Test Foo fossils, which can also be deleted unless you need them) SDSNM: delete 43 San Diego State Natural History SMF: delete 16 Forschungsinstitut Senckenberg, Frankfurt, Germany UCMP: delete 2 University of California Museum of Paleontology

[jimallman]

EMPTY STRING: delete all 5

Done. Note that the single fossil using this "collection" had no useful data, but it was the only fossil linked to Bovinae-Antilopinae.

DPC: delete 87 Duke University SDSNM: delete 43 San Diego State Natural History SMF: delete 16 Forschungsinstitut Senckenberg, Frankfurt, Germany UCMP: delete 2 University of California Museum of Paleontology

Done.

FOOZ: delete both (these are your Test Foo fossils, which can also be deleted unless you need them)

I'd like to keep these test records around until we've closed all related issues.

[Ksepka]

David / Jim, I am satisfied with this and ready to close pending approval.