Re: 2048 bug
Re: 2048 bug
- Subject: Re: 2048 bug
- From: Esteban Bodigami <email@hidden>
- Date: Tue, 14 Jul 2009 05:59:18 -0600
i was thinking of octal or ternary computers for that era... and based on optical systems (optical Apple CPU), but yeah... 64 bits FTW!
Terry Lambert
<email@hidden>
On Jul 13, 2009, at 2:59 PM, Esteban Bodigami <
email@hidden> wrote:
2552/7/13 Kevin Van Vechten:
From: Esteban Bodigami
Date: July 13, 2009 11:16:33 AM PDT
Subject: 2048
cann't we just count the minutes after the 1st of January of 1970 instead of seconds after it?
No. In general it is not possible to redefine existing programming interfaces after the fact without disrupting existing applications.
If one were to define a new interface, it might make sense to partition the units into minutes and seconds as you describe.
We've already bumped time_t to be a 64-bit value on x86_64 which extends the resolution quite a bit further into the future for that platform.
I know about that solution... but what about 32-bit computers?
They become obsolete in 29 years. Or we slide the epoch. Or we lose binary compatibility. As the person more or less responsible for making it size variant so it was 64 bits for 64 bit systems, I leaning towards obsolete for 32 bit systems.
-- Terry
--
Esteban Giuseppe Bodigami Vincenzi
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
| References: | |
| | >2048 bug (From: Esteban Bodigami <email@hidden>) |
| | >Re: 2048 bug (From: Kevin Van Vechten <email@hidden>) |
| | >Re: 2048 bug (From: Esteban Bodigami <email@hidden>) |
| | >Re: 2048 bug (From: Terry Lambert <email@hidden>) |