scrook / neuroml-db Goto Github PK

NML CL 73 - needs channels
On Tue Apr 14, J&Sharon decided to wait till automated model imports from OSB is available

What is not quite clear to me in the documentation is that I think we want to be sure that all the database and curation tool code is on github. Ideally, one would pull, make changes, test, and push? Then the production site would pull as well, right? Please clarify in documentation.

The visual schema does not show a primary-foreign key relationship between the refers table and the new resources table that you created. <<

along with changing the name of the key of resources to Resource_ID to be the same format as the rest of the tables (you only called it ID)? This will necessitate a change to the Web code for the join.<<

NMLSYN 79 - needs metadata and references (ie. modeldb or osb)

As discussed with Sharon and Justas on Tue March 31st

I will send an email with an update of text for the documentation page. For the video, I think you should cover the simple and advanced search but by this I mean cover how you do the search using the query box and hitting search and also using the advanced search page. The script for the "simple search" is fine but I would actually click on the pubmed link, the resource links (highlighting ModelDB and OSB), as well as what you are already doing. I wouldn't call any of what you show where you use the ontology assisted search "advanced search".

NeuronDB models like Ch 49 (http://neuroml-db.org/model_info?model_id=NMLCH000049) have links to broken URLs to neuronDB website.

There are some old ones on dendrite. I'll figure out which ones are safe to delete.

http://neuroml-db.org/model_info?model_id=NMLCH000058 results in internal redmine error

I just found that the following models have duplicate authors/translators - the Person_ID is the same, too.

NMLSY000080
NMLSY000081
NMLSY000082
NMLSY000083

Requires a sweep through code, refactoring passwords out so they're not in code files.

From Suzanne:

I found 4 publications in the database that are not related to a model:
6000021 - pubmed/15525801
6000022 - pubmed/15525801
6000023 - pubmed/15525801
6000033 - pubmed/12574468

NeuroML/NeuroMLWebsite#2

We need a plan for cleaning up spike. It is too confusing to have a lot of old code there.

TO be assigned to Suzanne when db authors have been cleaned up

However, I just discovered a channel NMLCH000054 that is in the models table with metadata
but it is NOT in the channels table and returns an error when retrieving by model_id.
<<

Nif: Random
Ours: as dicussed

As reported by @dietrichasuedu

When I was just testing out the NIF integration, I clicked on the “View HTML” for several of the models that I chose randomly, and the HTML view does not work.

For example, NMLCL000073 shows:

Cell: L5PCtemplate_0 exported from NEURON ModelView

and

NMLCL000005 shows:

Cell: soma exported from NEURON ModelView

I agree that code for neuroml-db and curation tool probably should be combined. But I think it is best to have everything for neuroml.org separate since we host it but are not entirely responsible for maintaining it. Should be possible to always clearly separate the two. Please add wording to reflect this in documentation.

We either need to delete this or fix it. I'm not sure what it is doing but maybe a keyword search with AND, OR, NOT and no return of associated models. I seem to see cells only.

HTML of the following cell does not display "friendly"

http://neuroml-db.org/model_info?model_id=NMLCL000001

So we know when the site is most and least busy, as well as where traffic is coming from.

This affects:

http://www.neuroml-db.org and
https://www.neuroml-db.org

In the https case, results in a SSL certificate error.

Checking apache config files, this appears to be happening before the request gets to apache. I can't access the domain.com registrar, but this may be an issue with the www sub-domain setup. Asked @scrook for access to look at the www entry on domain.com.