I see no problem with that date. yy-mm-dd ensures that files are always ordered correctly by date.

Not true in 2100.

This is another reason why we should call this year 12017 [1], the next 12018, and so on.

[1] http://www.popularmechanics.com/science/archaeology/a24195/c...

The formula (year - 2000) * 10000 + month * 100 + day can perfectly accommodate any future dates, until we change our calendar.

But without zero-padding, it will sort wrong when using lexiographic order (as is common). 1180101 would sort before 171204 in many contexts, even though it should clearly come behind the latter.

however if we invent time travel, and go back into the year 2000 we wouldn't be able to store any event data. furthermore, to get back, if that data field was used to store event data and the coordinates to get back to the current location of earth, could send you to a divide by zero coordinate and you wouldn't be able to get back. The solution is to just ban time travel to the year 2000 until a patch comes out

I don't know, but it is possible that the GW encodes something that increases once a century or faster. There is definitely more information in two letters than two digits.

GW = gravitational wave

https://en.wikipedia.org/wiki/List_of_gravitational_wave_obs...

What you don't think the date string 120521 is unambiguous? Oh or was that 051221 or was that 210512...?

The date when it was discovered is irrelevant - no astronomer cares about that. You should think of it as a unique string associated with some data + coordinates.

But it's not unique. What if we get two gravitational hits on one day? Or what if we get one on January 1, 2019 and also one on January 1, 2119?

GRBs are named in a similar fashion with simply having an additional character at the end, for example GRB171010B is the second GRB of 2017-10-10. And the naming scheme will simply have to change in 2100 (actually probably already 2090 because the system was in use before 2000)..

Supernovae on the other hand are named with the full year followed by one or more characters to disambiguate them - for example SN1987A or SN2013EE. This is handled centrally by the IAU and takes some time, which usually means that the same object is known under many different temporary names until the official name is assigned (and only if it is an actual supernova!).

It’s bad, and for a number of alternative dates using the same formatting it would be unreadable. They were lucky in this specific case.

Not so much for others... for example, ‘GW011012’

Nice...

Isn't it just an identifier that then points to all the meta-data?

GW - gravitational wave

170817 - one of the worse ways to format a date.

It makes you wanna cry..