search

OpenCosmics / fits-evaluation

We will use hisparc's existing data to evaluate NASA's FITS format as an option for a generic open cosmic data format.
Apache License 2.0
6 stars 4 forks source link

Implement Frank's mind map #11

Closed onyb closed 9 years ago

onyb commented 9 years ago

https://www.mindmup.com/#m:a150006750edb50132dfcf1222860cd564

RaoOfPhysics commented 9 years ago

DST is a social construct, which doesn't change the position of the sun. So, that's tricky. I would think that the times should always be the non-DST ones. @berghaus to comment?

153957 commented 9 years ago

Since we are looking at cosmic rays (mostly not from the Sun) would GMST (or LST) then not be more appropriate than local time?

onyb commented 9 years ago

@153957: I let it remain as local time. The program parsing this JSON can obviously convert the timestamp to GMT or any other format.

So far Daylight Savings Time is concerned, I am looking into it right now.

onyb commented 9 years ago

@153957, @RaoOfPhysics: Daylight Savings Time changes is already handled by the API that I have used.

153957 commented 9 years ago

Handled how? You only give it the location so I assume it returns the current DST, not the DST at time of the event..?

onyb commented 9 years ago

@153957: I am afraid you are right. This needs more work than I thought.

153957 commented 9 years ago

The main reason we do not use a local time (or UTC) is because of DST (and leap seconds). Date times are not unique in those cases. This increases the probability for errors when determining event rates or combining it with other data sources (for instance weather data). It is easiest to keep it as GPS (or UTC) time, prevents some headaches.. (we use GPS time with the UNIX epoch)