Re: Predictable Internet Time

["Patrik Fältström " <[e-mail]>]

On 3 Jan 2017, at 7:43, Eliot Lear wrote:

> Good morning, Patrik,
>
> On 1/3/17 7:29 AM, Patrik Fältström wrote:
>> I think personally, as long as we do have leap seconds:
>>
>> - we should have the leap second information available somewhere in clear machine readable format. Some suggestions exists, including encoding it in A-records in DNS ;-)
>>
>> - we should look at having the time since epoch really be the number of SI-seconds since the epoch
>>
>> - we should have translation between number of seconds and UTC take leap seconds into account
>
> The TZDB at IANA contains the file "leapseconds" which is machine readable, and provides a history of all leapseconds.  This provides you the 1st and 3rd of your requirements.

What is important is that the file include also known future leap seconds (and not) like the official memo. So that one can prep the local system with some predictability for the next say 5 months.

Maybe it does. My apologies for being lazy.

>> - we should fix the code that do not accept 61 seconds in a minute
>>
>> Now, we can also say we should stop having leap seconds, but I feel that is a _different_ matter and different discussion. I am myself not clear over what is the correct thing regarding leap seconds.
>>
>> What I am sure of is that I think most of the problems we have is because of bad programming (including in old UNIX days the priorities although correct at the time have continued to let the time_t definition continue to be wrong).
>
> I'm of two minds as well.  On the one hand, the current system makes it quite easy to do the wrong thing, and that's not good.  There are utility functions such as pytz's normalize() and localize() functions that people just don't know about, for instance.  On the other hand, it's not just leap seconds.  It's timezones in general, it's DST transitions, etc.  Getting rid of leap seconds only marginally improves things, and gets really philosophical really quickly...

The problem I have is that we now implement things like smearing which to me is a bad fix for something that is wrong in the first place (definition of time_t).

   paf