Uncaught TypeError: Cannot read property 'scrollHeight' of undefined

[yarikoptic]

Happens for http://nwbexplorer.opensourcebrain.org/nwbfile=http://datasets.datalad.org/dandi/dandisets/000027/sub-RAT123/sub-RAT123.nwb URL - it seems to get and load that really basic file , but then Python tab has just a spinning wheel. Console log logs an error:

[filippomc]

Hi @yarikoptic, thanks for reporting this. The error appears when passing the filename as parameter. Can workaround for the moment by going to http://nwbexplorer.opensourcebrain.org and writing the file url in the form.

[yarikoptic]

Thank you @filippomc for working on the fix!
Sorry I have not provided more context: the idea is to directly link to nwbexplorer and just provide a url to the target file, so we could provide "integration" with it from dandiarchive.org and a good/old datasets.datalad.org -- you can even see it in "action" within context menu (vertical dots to the left from file) on http://datasets-dev.datalad.org/?dir=/dandi/dandisets/000027/sub-RAT123 sample. ATM just named "View in Renderer", will be made more specific.

[pgleeson]

@yarikoptic (and @satra) that's exactly the type of usage we'd encourage, making it as easy as possible for uses to see the contents of the NWB files using NWBE.

There have been some recent updates to the graphical front end to make more data types accessible in the viewer (e.g. a Processing tab in addition to Acquisition and Stimulus), but the fallback is always looking at the data using NWB Widgets in the python console.

While this route still requires that the NWB files aren't too large (<1GB), the integration of NWBE into OSBv2 with inbuilt JupyterLab and file persistence should help loading times and general management/analysis of multiple files.

[filippomc]

btw, the fix is not deployed and works from the link from the datalad page.

[yarikoptic]

I guess you meant to say "now deployed" since it works for me now even on the phone! ;-) awesome, thank you!!