We're happy to announce a new set of inputs! First, the ones that support events from zipped files in the storage location of your choice:
For those storage locations, as well as for S3, we've also added the option to load Excel files (.xls or .xlsx files)! This isn't available in the UI yet, so let us know if you'd like us to enable it for your input.
One additional input we've added is HubSpot, which syncs every 30 minutes to add to our super-fast input lineup.
Ever added an input, only to realize it was improperly configured and we didn't stop you? No more! Our inputs now validate as you click "Finish", so you get a chance to fix your input definition before it's created. This one was a long time coming, and we hope it makes input creation a lot easier.
We've also improved some existing inputs:
- Gone are the days of needing to ask support to dump and load an initial snapshot of a MongoDB collection - the MongoDB input now gives you the option (selected by default) to dump and load all collections for an initial snapshot.
- Google Adwords was updated to the latest API.
- Microsoft SQL Server was improved to also allow varchars as the update indicator column for incremental loads.
We've added some handy support for some of our outputs:
- We heard your feedback! We changed our auto-mapping logic for the case when a table with the event type name already exists. Instead of creating a new table with _1 appended, we'll check if the table that exists has the exact same properties as the table we're attempting to create. If it does, we'll use it. If it doesn't, we'll fail the auto-map and you'll have to choose a table manually.
- MySQL output now maps varchars/strings that are more than 256 characters to the Text data type, allowing you to support much more data in your MySQL.
- We've added to Snowflake the Timestamp_LTZ type.
- We made improvements across the board to speed up UI loading (you may also notice we added a quote on the load screen - your daily dose of data inspiration!).
- We've made some consolidation improvements for fewer failures, including making sure that consolidations still run successfully, even if the source input is temporarily unavailable.
- We also made the change to stop running consolidations after an event type is deleted or discarded, since those will fail anyway.
- We noticed that the output node in the plumbing looked a bit weird when no output was configured because the Live tab was hidden and the error indication was wrong. So, we improved that.