Re: [PATCH] sleep: support subtraction of sleep amounts

[Bernhard Voelker]

On 08/03/2014 11:02 PM, Bob Proulx wrote:
> I know that Jim already opened the door by saying he would accept the
> feature if it were written.  For my part I see this as creeping
> featurism and would vote against it.  With the "camel in the tent"
> parable in mind I expect sleep to inevitably include full human date
> parsing routines at some point.  If it will accept "1d -5s" then why
> not "all night" or "until the next blue moon"?

Considering the bug reports we've received for date(1) in the past, it
seems that we'll need to refactor the date parsing code some day anyway
... so camels welcome! ;-)

Just kidding - okay, it seems there's not much consense for adding this.
I just took that old idea to provide a patch, but honestly would probably
never use "1d -1s" myself.  Therefore, I'd just drop it.

Nonetheless, I think some basic tests for sleep(1) should be added for
proving that the addition of times works.  I'll try to provide a patch
for this soon (stripping off the subtraction tests).

Thanks & have a nice day,
Berny