Thanks Jarod and team this has a lot of good stuff in it.
Very good written. Also, it is very helpful to have a little practical example given in almost each of the new features and functions.
Idea: Maybe it would be good to have a recommended keyword to search for in the online help in order to get very fast to the most useful information. But it is not a must. ;-)
Thanks for this good quality.
Oooops! Just found this:
GA or not GA?
This is a feature that has received a lot of interest from some users. It is currently offered as a technology preview and not suitable for all use cases, but we did feel that it was valuable to include in the release in order to allow interested users to experiment with this new service in POC or development environments.
I saw a lot of text and explanations in the Install Guide for this feature and I wonder if there is something we should be aware about BEFORE we install the cluster. I mean something like disk capacities or other things like this, which is very difficult to be changed after the cluster is deployed without the required settings/configs?
Another new feature question: ... "raw, unprocessed content fields"... What could this be?
Some of the parsers used by our metadata extraction stage will parse out multiple versions of the same piece of metadata, including a "raw" version and a processed, or cleaned up version. We typically filter out these "raw" fields as they are redundant and generally less useful than what we keep. In order to be more flexible, however, we added the option for this stage to not do that filtering and "open up the fire hose" to attach every piece of metadata they find to documents if users desire.
Do you have an example of this "raw" data compared to the "cleared up version" for better understanding?
If you want to see the difference the option makes, I'd recommend testing a pipeline with an email file. Try it with that option both off and on, and compare the difference.
Retrieving data ...