SECNDS broken?

[dbruceg]

You might have a look at the SECNDS routine. The intrinsic function gives different results than the IFPORT version, and it appears that BOTH of them are broken.

Bruce

[Steven_L_Intel1]

Exactly what do you see? I see this:

- The intrinsic returns a correct value (with an exception noted later) to hundredths of seconds
- The ifport routine always returns an integer valued result (no hundredths)
- If the intrinsic would return zero after subtracting the argument, it returns 86400 instead - the ifport routine doesn't.

The latter two I'll agree are wrong and I'll report that.

[dbruceg]

I haven't got the stuff in front of me, but that looks just like what I was getting last night. I got the 86400's, though I didn't notice the correlationwith a returned zero. Could very well be, though, and that gives me a workaround. I noticed that the IFPORT routine returned no hundredths, but I also suspected that the integral value was wrong, since things that should have clocked a few seconds (I think) reported zero ET. That, however, is only a suspicion.

Bruce

[Steven_L_Intel1]

The integral value was just fine in my tests. Truncated, but otherwise correct. I have reported these to the developers.