On 07/29/2014 03:46 PM, Bernhard Voelker wrote:
> On 07/29/2014 09:38 PM, Eric Blake wrote:
>> As I understand the comments on that bug, the question there was whether
>> other implementations allow 'sleep inf', and not whether strtod() is
>> able to parse "inf".  It may well be that there is a non-GNU sleep that
>> parses "inf" but not "infinity" because it hand-rolls its own parser
>> instead of using strtod().
> 
> ... which we wouldn't have a problem with, right?

Correct - regardless of whether other implementations support 'sleep
inf' as an extension, we do because we use strtod() and parse our sleep
value as a double.  It shouldn't matter what other implementations use
to do their parsing, as long as they meet the minimum portable behavior,
and as long as known extensions between their implementation and ours
don't appear to conflict.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to