Open alexanderdean opened 8 years ago
Assigning to @ihortom to populate the release table...
@alexanderdean - here's my view
Release | Column count | Additions | Removals | Renamings |
---|---|---|---|---|
>=R73 (v0.8.0) | 128 | unstruct_event contexts derived_contexts |
||
R71-R72 (v0.7.0) | 131 | true_tstamp event_vendor event_name event_format event_version event_fingerprint |
dvce_tstamp renamed to dvce_created_tstamp |
|
R63-R70 (v0.6.0) | 125 | tr_currency tr_total_base tr_tax_base tr_shipping_base ti_currency ti_price_base base_currency geo_timezone mkt_clickid mkt_network etl_tags dvce_sent_tstamp refr_domain_userid refr_dvce_tstamp derived_contexts domain_sessionid derived_tstamp |
||
<=R62 (v0.5.0) | 108 |
Looks great, though I don't think we ever removed derived_tstamp
:
https://github.com/snowplow/snowplow/blob/master/4-storage/redshift-storage/sql/atomic-def.sql#L181
According to release notes we renamed derived_tstamp
to dvce_created_tstamp
Have you got a link to that Ihor? I think we renamed dvce_tstamp
to dvce_created_tstamp
...
You are right, amended the table
Looks great, thanks. Actually fewer iterations that I was remembering...
Is there a reason not to add a renmaings
column, in addition to additions
and removals
? That will make it a little clearer that dvce_tstamp
was renamed to, not replaced with, dvce_created_tstamp
.
Good idea, let's make that change
Updated the table
We would use the count of fields to determine which version of the enriched event we are working with. See also: #2, #4.