You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 29, 2019. It is now read-only.
At some point between 4.16.1 and 4.21, using a simple string such as "2018-03-01" started acting a bit strange. The event shows up a bit before the date instead of on the actual date.
Notice that the event line is exactly on the line where the 20th starts.
But using timevis 4.21 I get this:
Notice that now the event is pinned to a bit before the 20th. I realize I can probably change my javascript code to account for timezones, but I've ported this library to R (and it's quite popular), so people are simply writing date strings in R and they get passed down to visjs, and it would suck to have to tell everyone to change their code.
The text was updated successfully, but these errors were encountered:
I can confirm that this change happened before 4.19.1, though I can't exactly say in which version it happened because that's the oldest version of the library I could find in my archive. I found some hints that it might happened in 4.19.0 or 4.18.0 but I can't say for sure without testing those versions. At least that issue should be easily fixed by defining your own moment constructor.
At some point between 4.16.1 and 4.21, using a simple string such as "2018-03-01" started acting a bit strange. The event shows up a bit before the date instead of on the actual date.
Here's a minimal HTML code to reproduce:
Running this with visjs 4.16.1 looks fine:
Notice that the event line is exactly on the line where the 20th starts.
But using timevis 4.21 I get this:
Notice that now the event is pinned to a bit before the 20th. I realize I can probably change my javascript code to account for timezones, but I've ported this library to R (and it's quite popular), so people are simply writing date strings in R and they get passed down to visjs, and it would suck to have to tell everyone to change their code.
The text was updated successfully, but these errors were encountered: