code review 1436041: Fix typo in gob docs. They were introduced in revision ...

Fix typo in gob docs. They were introduced in revision 3199778baf
"change the encoding of uints...".

[nigeltao, 2010-05-31 22:50:40]

Hello r (cc: golang-dev@googlegroups.com),

I'd like you to review this change.

[r, 2010-05-31 22:57:00]

I'd be surprised if these are wrong. I used a full example and worked on the
documentation until I could explain every byte.   Please show me a
counterexample.

I'm not saying you're wrong, just that I need more evidence.

[nigeltao_gnome, 2010-05-31 23:40:05]

On 31 May 2010 15:57,  <r@golang.org> wrote:
> I'd be surprised if these are wrong. I used a full example and worked on
> the documentation until I could explain every byte.   Please show me a
> counterexample.
>
> I'm not saying you're wrong, just that I need more evidence.

The documentation is comprehensive and mostly right. It's just that,
about a year ago, the uint encoding changed in
http://code.google.com/p/go/source/detail?r=3199778baf171d6d1b6a29bfdc5d62c6e...
and not all of the comments were updated at the same time.

Test program:
-----------------------
package main

import (
    "bytes"
    "fmt"
    "gob"
    "math"
)

type foo struct {
    i int
    f float
}

func main() {
    b := bytes.NewBuffer(nil)
    e := gob.NewEncoder(b)
    e.Encode(foo{})
    b.Reset()
    e.Encode(foo{i:-129, f:17.0})
    fmt.Printf("% x\n", b.Bytes())
    fmt.Printf("17.0's bits are %x\n", math.Float64bits(17.0))
}
-----------------------

Output:
-----------------------
$ ./8.out
0b ff 82 01 fe 01 01 01 fe 31 40 00
17.0's bits are 4031000000000000
-----------------------

[r, 2010-06-01 00:01:01]

Aha, I see now. Thanks.

LGTM

[nigeltao, 2010-06-01 00:36:23]

*** Submitted as http://code.google.com/p/go/source/detail?r=0e45c9a6e789 ***

Fix typo in gob docs. They were introduced in revision 3199778baf
"change the encoding of uints...".

R=r
CC=golang-dev
http://codereview.appspot.com/1436041