code review 170750043: doc/go1.4.html: final library changes

doc/go1.4.html: final library changes
First draft now complete.

[r, 2014-10-29 20:36:50]

Hello golang-codereviews@googlegroups.com,

I'd like you to review this change to
https://code.google.com/p/go

[rsc, 2014-10-29 21:55:35]

LGTM

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html
File doc/go1.4.html (right):

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode564
doc/go1.4.html:564: so the split function can generate a final empty token.
as the documentation has always promised, 

or something like that, to make clear this is just doing what was always
documented.

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode569
doc/go1.4.html:569: handle empty tokens at EOF as desired.
The most common failure mode for a bad split function is an infinite loop
generating empty tokens at EOF. Programs that used to 'work' now hang. I saw
this a couple times updating Google code. Should we make the Scanner
implementation count empty tokens at EOF and panic if it sees 100 or something
like that?

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode716
doc/go1.4.html:716: as it already does for the other systems.
s/does/did/?

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode755
doc/go1.4.html:755: implementations on Linux of
s/.*/implementation on Linux, the/

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode767
doc/go1.4.html:767: func TestMain(m *<a
href="/pkg/testing/#M"><code>M</code></a>)
s/M/testing.M/

[r, 2014-10-29 22:34:59]

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html
File doc/go1.4.html (right):

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode564
doc/go1.4.html:564: so the split function can generate a final empty token.
On 2014/10/29 21:55:34, rsc wrote:
> as the documentation has always promised, 
> 
> or something like that, to make clear this is just doing what was always
> documented.
> 

Done.

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode569
doc/go1.4.html:569: handle empty tokens at EOF as desired.
On 2014/10/29 21:55:34, rsc wrote:
> The most common failure mode for a bad split function is an infinite loop
> generating empty tokens at EOF. Programs that used to 'work' now hang. I saw
> this a couple times updating Google code. Should we make the Scanner
> implementation count empty tokens at EOF and panic if it sees 100 or something
> like that?

https://code.google.com/p/go/issues/detail?id=9020

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode716
doc/go1.4.html:716: as it already does for the other systems.
On 2014/10/29 21:55:34, rsc wrote:
> s/does/did/?

Done.

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode755
doc/go1.4.html:755: implementations on Linux of
On 2014/10/29 21:55:34, rsc wrote:
> s/.*/implementation on Linux, the/

Done.

https://codereview.appspot.com/170750043/diff/1/doc/go1.4.html#newcode767
doc/go1.4.html:767: func TestMain(m *<a
href="/pkg/testing/#M"><code>M</code></a>)
On 2014/10/29 21:55:34, rsc wrote:
> s/M/testing.M/

Done.

[r, 2014-10-29 22:35:52]

*** Submitted as https://code.google.com/p/go/source/detail?r=3f6f728549d3 ***

doc/go1.4.html: final library changes
First draft now complete.

LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews
https://codereview.appspot.com/170750043

[gobot, 2014-10-29 22:44:47]

This CL appears to have broken the openbsd-386-rootbsd builder.
See http://build.golang.org/log/59b31cffb4237b8572a3bff1b1aee0b4d994c05c