Joel Jaeggli wrote:
> The 2038 issue is more similar to the gps rollover (ie week 1024 midnight
> august 21 1999) then it is to y2k datefield issues.
...
GPS is a weak analogy. Linux keeps date codes on files and cares
about them. As far as I know GPS doesn't. It is possible to have
very old files in a Linux system. If you let the counter roll over
and just adjust the definition of the epoch (as you have suggested)
then any files you have lying around are now dated in the future.
Very ugly.
Bill Wendling:
> Also sprach Matti Aarnio:
>} On Wed, Jan 05, 2000 at 03:37:06PM -0600, Bill Wendling wrote:
>} ...
>} > I'll put my foot squarly in my mouth and predict that, in 38 years,
>} > people will think that running anything on a machine less than 1GHz speed
>} > for a production machine is nothing short of insanity.
Even today there are plenty of computers that run 8 bits at
low clock frequencies: embedded systems. Embedded systems
don't use the biggest CPU simply because it is available.
They used the smallest, cheapest CPU that gets the job done.
Linux IS being used in embedded systems today. In the future
that may be the majority of Linux systems. I think that
Linux embedded systems will be around long after Linux is
no longer used for personal computers.
> I'm sorry, but the computer industry just doesn't seem to be concerned
> with what does the job. They are interested in fast, sexy hardware and
> cool apps with lots of eye candy. Not that that's bad or anything.
This is simply not true of embedded systems.
The sooner the Linux kernel supports dates beyond 2038, the
sooner embedded systems will be created that can survive beyond
that date. And the fewer broken embedded systems there will
be in 2038.
I don't mean this needs to happen in the about-to-be-released
kernels. But I think it would be wise to include this in the
next generation. One way or another, this is clearly Linus's
decision. I look forward to hearing how he weighs in on this
issue.
-- Allen Brown not representing my employer- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Jan 07 2000 - 21:00:05 EST