OpenEuphoria

RE: Time conversion...

• Posted by leviathan at uswest.net Feb 26, 2001
• 533 views

> Hi again "LEV",
> 
> In my opinion, if your going to wrap a time function like that
> you should go all the way and include hours as well, in case
> someone inputs a high number for the seconds parameter.

Or perhaps two routines to do it, one with hours, and one without? 
Or even an entire time conversion routine? (Minutes into seconds, 
hours+minutes into seconds, even perhaps msecs?)

  This
> way you get the output
>   05:02:03  {hours:mins:secs}
> for an input of
>   5*60*60 + 2*60 + 3
> instead of an output of
>   302:03   {mins:secs}
> 

For the program I'm working on, I only get things that are usually 
below 60 minutes, and even ones above that, I don't bother with 
anyway, because that'd be one big freaking MP3 file! :)

> Secondly,
> using an integer as input parameter means the user has to
> round the seconds input themselves.  Rounding within the 
> function (passing atom) means you can get
>   119.99 seconds
> to come out as
>   00:02:00  {hours:mins:seconds}
> without bothering the user with this detail.
> 

Again, in my case, I don't come across decimal seconds, but again, 
yeah, passing atoms instead of integers is a better idea.

> The shortest function isnt always the best, as the most versatile
> functions are almost always longer then any others and are more
> suited to inclusion into a library.
> Consider a second parameter, that specifies whether or not to
> output
>   hours:mins:secs
> or just
>   mins:secs
> or even
>   days:hours:mins:secs
> or even several other formats to cover many possible applications.
> 

Which is what I was just thinking about too :)

> Good luck with it.

Luckily, I found exactly what I needed, however, I know full well that 
I'll need some stuff that manipulates time like that later on, so I 
should just go and wrap some time functions :) (Perhaps after I 
learn how to wrap dlls)

--"LEVIATHAN"