I am wondering what kind of issues QNX4 will have in 2038? Is it possible
to utilize time_t as a unsigned number? Will the internals of QNX support
such a thing?
Thanks.
Jeff Adler
Automation Services
“Jeffrey Adler” <jeffadler.at.bigfoot.dot.com@nowhere.com> wrote in message
news:cvo3q9$at9$1@inn.qnx.com…
I am wondering what kind of issues QNX4 will have in 2038? Is it possible
to utilize time_t as a unsigned number? Will the internals of QNX support
such a thing?
No it’s not possible. In 2038 you will be getting problems. The issue is
not just with OS but also with compiler/libraries etc. I wouln’t count on
QSS fixing fix this, ever.
Thanks.
Jeff Adler
Automation Services
Mario Charest postmaster@127.0.0.1 wrote:
“Jeffrey Adler” <> jeffadler.at.bigfoot.dot.com@nowhere.com> > wrote in message
news:cvo3q9$at9$> 1@inn.qnx.com> …
I am wondering what kind of issues QNX4 will have in 2038? Is it possible
to utilize time_t as a unsigned number? Will the internals of QNX support
such a thing?No it’s not possible. In 2038 you will be getting problems. The issue is
not just with OS but also with compiler/libraries etc. I wouln’t count on
QSS fixing fix this, ever.
This would be a major amount of work to fix.
I sometimes wonder if there’ll be any hardware QNX 4 will run on,
or any network/graphics cards QNX 4 will have drivers for, or
how many hacks IDE will go through to support bigger drives, and
whether QNX 4 will work with them, either.
QNX 6 uses an unsigned interpretation of time_t in it’s libraries
(good 'til 2100 or so) and uses 64-bit nanoseconds internally (2500
ish).
I don’t know what QSS will do, it seems unlikely to be under the
same management (at least the same people) by that time. I know
we did put some effort into dealing with the QNX2 date expiry
problem in 1996, and in dealing with Y2K issues in 1999.
-David
David Gibbs
QNX Training Services
dagibbs@qnx.com