desbest wrote in post #1125087:
> Timestamps are a number that counts the number of seconds from the epoch
> date 1/1/1970.
> I read somewhere on the
> internet<
http://en.wikipedia.org/wiki/Year_2038_problem>,
> that timestamps will expire in the year 2038.
> This is why I always use *yyyy-mm-dd* and *yyyy-mm-dd HH:ss* to show the
> date and I don't use timestamps or time database columns.
>
> Is it a good idea to use timestamps, or should I continue with
> yyyy-mm-dd
> HH:ss ?
This is only the case where the timestamp is stored as a 32 bit integer
value. AFAIK most modern systems use 64 bit integer values for this as
do not have the year 2038 bug.
I assume by "yyyy-mm-dd" you mean that you're storing dates as string
values. That is significantly less efficient that using datetime fields.
I would not worry about the 2038 bug. This problem will be fixed before
it becomes a problem if it is not already fixed in your database and
other related systems.
For example PostgreSQL stores its timestamps in 8 bytes (64 bits) and
has a range from the year 4713 BC to 294276 AD. So I think one would be
safe when using PostgreSQL timestamp.
Also of note is that the epoch for PostgreSQL is midnight 2000-01-01 and
not 1970-01-01. Not all systems use the same epoch.
In fact you'll find there are many different Epochs:
http://en.wikipedia.org/wiki/Epoch_(reference_date)#Notable_epoch_dates_in_computing
--
Posted via
http://www.ruby-forum.com/.