• Official data to be hosted on zenodo
• Convert files from json to h5 (much smaller)
• Move results into a public (sandbox) zenodo for testing; no DOI generated
• Test automated upload

• Masses should be in the source frame as thats what the paper is showing (add info to labels; otherwise document somewhere that these are source frame masses)
• TOC at the top of each event page; e.g. header for each plot/table so the user could see whats throughout the page
• NR Fits -> events
• Add offical LVK posteriors. Possibly switch to toogle on/off the different models?
• Summary page plots on the home page
• NRSur7dq2 --> NRSur7dq4
• For reproducibility: (i) yaml files for which the runs were done. PSD files and calibration envelopes should be easy to get to (either we give the data, or we provide link to where this can be found -- point to 2.1 and 3 data whenever possible)
• Have versions and configurations (e.g. https://pycbc.org/3-ogc/posteriors/4._config_files/4.01_gw150914_095045/)

@tousifislam has a new file format for the results which we need to be able to parse/work with

Are the public posteriors compatible with pesummary (in addition to the custom nr-catalog package?)

Should we store the results as h5 instead of json? h5 files are much smaller in size i think...

We should probably investigate the size difference first.

I can't tell who submitted the zenodo record, but I figured this might be as good a place as any to suggest this: That record needs at least a little more detail in the description. It says, "Refer to the paper for more details. " But what paper? There should at least be a link to the arxiv version.

for users who want to copy/paste some code to start using all the posteriors. On the main page a few Python commands for how the entire catalog can be downloaded and, say, make a plot of some posterior.