gsautter
commented
2 years ago You're right, this turn-of-the-year behavior may seem a bit counter intuitive at the beginning ... however, there's a simple idea behind this: the finer granularities don't exist without the coarser ones. In particular, it means this:
- if there is no
monthly, it implies there is noweeklyordaily - if there is no
weekly, it implies there is nodaily
This allows for a very simply download/import logic:
- check the full dump, and download it if your local version has an earlier timestamp
- check the
monthly:- if there is none, you're done right here
- download it if your local version doesn't exist or has an earlier timestamp
- check the
weekly:- if there is none, you're done right here
- download it if your local version doesn't exist or has an earlier timestamp
- check the
daily, and download it if your local version doesn't exist or has an earlier timestamp
In particular, it saves you even checking for the daily if you don't find the weekly, etc.
The weekly and daily will become available in the next couple of days, and then things will go back to their normal course.
If you have a different idea how the naming logic should behave, please let me know ... the current one is simply what I figured might be easiest to handle for client code (using above logic) ... and I might well be wrong ...
@gsautter
You explained this to me before but I am confused still… please clarify for me how the creation and naming of the zip archives work. If I look today at the dumps (https://tb.plazi.org/dumps/) I see the following
Since it has been only day since the full archive was created, why is the second archive called "monthly" and not "daily"? From what you explained to me, I would have expected there to be a daily archive for 6 days (from full archive creation), then on the 7th day, a weekly archive would have been created, then on the 30 days there would be a monthly archive… what's going on?