Dear all,
We recently published a fairly massive dataset with around 10,000 files at QDR. While our Dataverse, with some initial creaking and whining, is handling this mostly OK, we did notice that the generated JSON-LD file is massive, so we're loading a ~4MB JSON file into the site header for everyone who looks at the dataset. This seems.... suboptimal (it also appears to break google's parsing of the file).
Do any metadata experts here have an opinion on a better option for this -- like a shortened JSON-LD that takes it easy on the related items or so? Or even better, any of the repositories with large deposits like this, do you have a strategy for this?
Thanks in advance!
Sebastian
--