time.UnixNano() error -- Possible bug

[senior7515]

http://play.golang.org/p/MmC-ltROsq

Increate the time on your host to say 1 second.

It prints the same result.

[Brad Fitzpatrick]

The playground has a fixed clock for security-ish reasons.

It works on a real machine.

[David Symonds]

The playground is a limited environment, and its time does not
advance. Try running that code locally.

Dave.

[David Symonds]

Wait, that's not it. You're never changing the 't' variable after init.

[chris dollin]

Not here.

(i and j will of course be ==; they're a conversion of the
same time to a unix-nanoseconds-count.)

Chris

--
Chris "allusive" Dollin

[Brad Fitzpatrick]

On Wed, Mar 7, 2012 at 10:38 PM, David Symonds <dsym...@golang.org> wrote:

Wait, that's not it. You're never changing the 't' variable after init.

Hah, I missed that too.  I guess the playground does do time now:

http://play.golang.org/p/mCtC_vIrBW

(modulo source -> result caching)

[senior7515]

I know that. I am saying that on my fbsd box it does in fact print the
same number.

I also mentioned it to increase it to 1 second before testing. I think
there is another limit to 1 sec on appengine or smth like that.

On Mar 8, 1:43 am, Brad Fitzpatrick <bradf...@golang.org> wrote:

[senior7515]

Oh wait i didn't say I used it on an FreeBSD box. sorry.

On Mar 8, 1:43 am, Brad Fitzpatrick <bradf...@golang.org> wrote:

[senior7515]

Chris: even after sleeping ?

time now in nano + sleep + time now in nano

Did I miss something ? maybe i don't get it.

On Mar 8, 1:41 am, chris dollin <ehog.he...@googlemail.com> wrote:

[senior7515]

Edit:

I ran this code with 2 seconds of sleeping on a real machine. FBSD 9.0
amd 64. Latest weekly i just rebuilt my $GOROOT/src

According to http://weekly.golang.org/pkg/time/#Time.UnixNano

It returns the number of nano since epoch. This should change after
sleeping for 2 seconds for sure.

It doesn't on my system.

I also understand that there are appengine limitations. That's why i
mentioned to test it on your *host*.

Thank you. Let me know if I can provide any further clarification.
Sorry for being ambiguous the first time. It's late.

[Brad Fitzpatrick]

UnixTime returns the UnixTime of the receiver, t, not the the unix time of "Now", which you only called once, in your init.

You need to call Now() twice (before & after sleeping), if you want to see movement in time.

[David Symonds]

On Thu, Mar 8, 2012 at 5:58 PM, senior7515 <galleg...@gmail.com> wrote:

> I ran this code with 2 seconds of sleeping on a real machine. FBSD 9.0
> amd 64. Latest weekly i just rebuilt my $GOROOT/src
>
> According to http://weekly.golang.org/pkg/time/#Time.UnixNano
>
> It returns the number of nano since epoch. This should change after
> sleeping for 2 seconds for sure.

UnixNano is a method on an object. If you don't change the object, it
will report the same value each call.

Read the docs more carefully: t.UnixNano() is the number of
nanoseconds since epoch of *that* time, not of the current time. You
need to call time.Now().UnixNano() to get the current value.

>
> It doesn't on my system.

For clarity, please paste the *exact* code you are using.

[senior7515]

David: Thank you. Also, that is the exact* code except i sleep it 1e9
as mentioned in the post.

Brad: Thanks. False alarm.

On Mar 8, 2:02 am, David Symonds <dsymo...@golang.org> wrote:

> On Thu, Mar 8, 2012 at 5:58 PM, senior7515 <gallego.al...@gmail.com> wrote:
> > I ran this code with 2 seconds of sleeping on a real machine. FBSD 9.0
> > amd 64. Latest weekly i just rebuilt my $GOROOT/src
>

> > According tohttp://weekly.golang.org/pkg/time/#Time.UnixNano

[senior7515]

Just re read them. To quote : UnixNano returns the Unix time, the
number of nanoseconds elapsed since January 1, 1970 UTC.

How does that tell me is for the receiver t without looking at the
source ?

I suppose one can argue that since the receiver is t, it would make
sense.

I'll try not to post silly questions without looking at the source,
but I wanted to clarify that the docs didn't read clear to me.

[David Symonds]

On Thu, Mar 8, 2012 at 6:11 PM, senior7515 <galleg...@gmail.com> wrote:

> Just re read them. To quote : UnixNano returns the Unix time, the
> number of nanoseconds elapsed since January 1, 1970 UTC.
>
> How does that tell me is for the receiver t  without looking at the
> source ?
>
> I suppose one can argue that since the receiver is t, it would make
> sense.
>
> I'll try not to post silly questions without looking at the source,
> but I wanted to clarify that the docs didn't read clear to me.

You're right, it could be clearer. The other methods refer to the
receiver t, but Unix/UnixNano don't. Can you please file a bug?
http://code.google.com/p/go/issues/list

Dave.