code review 163058: Integrated feedback by Ken.

Integrated feedback by Ken.
Easy stuff in this round, more to come.

[gri, 2009-12-01 22:56:04]

Hello iant, rsc, r, ken2 (cc: gri),

I'd like you to review the following change.

[rsc, 2009-12-01 23:18:57]

LGTM

http://codereview.appspot.com/163058/diff/3/4
File doc/go_spec.html (right):

http://codereview.appspot.com/163058/diff/3/4#newcode542
doc/go_spec.html:542: The value of an n-bit integer is n bits wide and
represented using a
s/a two's complement system/two's complement/

http://codereview.appspot.com/163058/diff/3/4#newcode3204
doc/go_spec.html:3204: int(3.14)      // 3.14 truncated to int 3
This comment suggests a truncation is happening - I had
to look back a few lines to see that this was a list of illegal
expressions.

// 3.14 cannot be represented as an int

http://codereview.appspot.com/163058/diff/3/4#newcode3306
doc/go_spec.html:3306: A statement list can always be terminated with a
semicolon in effect
s/semicolon/&,/

http://codereview.appspot.com/163058/diff/3/4#newcode3467
doc/go_spec.html:3467: IfStmt    = "if" [ [ SimpleStmt ] ";" ] [ Expression ]
Block [ "else" Statement ] .
SimpleStmt can be EmptyStmt, so I think the extra [ ]
are unnecessary.  They're fine if they're just for emphasis.

[r, 2009-12-01 23:25:54]

http://codereview.appspot.com/163058/diff/3/4
File doc/go_spec.html (right):

http://codereview.appspot.com/163058/diff/3/4#newcode352
doc/go_spec.html:352: and two-digit hexadecimal (<code>\xnn</code>) escapes
represent individual
i prefer the original

http://codereview.appspot.com/163058/diff/3/4#newcode450
doc/go_spec.html:450: For instance, <code>3.0</code> can be given any integer
and any
s/and/or/

http://codereview.appspot.com/163058/diff/3/4#newcode542
doc/go_spec.html:542: The value of an n-bit integer is n bits wide and
represented using a
n should be <i>

[r, 2009-12-01 23:26:03]

LGTM

[gri, 2009-12-01 23:41:27]

PTAL.

http://codereview.appspot.com/163058/diff/3/4
File doc/go_spec.html (right):

http://codereview.appspot.com/163058/diff/3/4#newcode352
doc/go_spec.html:352: and two-digit hexadecimal (<code>\xnn</code>) escapes
represent individual
On 2009/12/01 23:25:54, r wrote:
> i prefer the original

made the n's <i>

http://codereview.appspot.com/163058/diff/3/4#newcode450
doc/go_spec.html:450: For instance, <code>3.0</code> can be given any integer
and any
On 2009/12/01 23:25:54, r wrote:
> s/and/or/

Done.

http://codereview.appspot.com/163058/diff/3/4#newcode542
doc/go_spec.html:542: The value of an n-bit integer is n bits wide and
represented using a
On 2009/12/01 23:18:57, rsc wrote:
> s/a two's complement system/two's complement/
> 

technically this is not correct: two's complement is a specific value (see the
link).

http://codereview.appspot.com/163058/diff/3/4#newcode542
doc/go_spec.html:542: The value of an n-bit integer is n bits wide and
represented using a
On 2009/12/01 23:25:54, r wrote:
> n should be <i>

Done.

http://codereview.appspot.com/163058/diff/3/4#newcode3204
doc/go_spec.html:3204: int(3.14)      // 3.14 truncated to int 3
On 2009/12/01 23:18:57, rsc wrote:
> This comment suggests a truncation is happening - I had
> to look back a few lines to see that this was a list of illegal
> expressions.
> 
> // 3.14 cannot be represented as an int
> 

Done.

http://codereview.appspot.com/163058/diff/3/4#newcode3306
doc/go_spec.html:3306: A statement list can always be terminated with a
semicolon in effect
On 2009/12/01 23:18:57, rsc wrote:
> s/semicolon/&,/
> 

Done.

http://codereview.appspot.com/163058/diff/3/4#newcode3467
doc/go_spec.html:3467: IfStmt    = "if" [ [ SimpleStmt ] ";" ] [ Expression ]
Block [ "else" Statement ] .
On 2009/12/01 23:18:57, rsc wrote:
> SimpleStmt can be EmptyStmt, so I think the extra [ ]
> are unnecessary.  They're fine if they're just for emphasis.
> 
Sigh. We went back and forth here. Reverted.

[rsc, 2009-12-02 00:00:24]

> technically this is not correct: two's complement is a specific value
> (see the link).

how about "two's complement arithmetic"? (wikipedia-approved)
the "a" in "a two's complement system"
makes it sound like there are other,
different two's complement systems.

[gri, 2009-12-02 00:14:58]

sounds good.

On Tuesday, December 1, 2009, Russ Cox <rsc@golang.org> wrote:
>> technically this is not correct: two's complement is a specific value
>> (see the link).
>
> how about "two's complement arithmetic"? (wikipedia-approved)
> the "a" in "a two's complement system"
> makes it sound like there are other,
> different two's complement systems.
>

[gri, 2009-12-02 00:15:55]

*** Submitted as http://code.google.com/p/go/source/detail?r=e6cf652c263f ***

Integrated feedback by Ken.
Easy stuff in this round, more to come.

R=iant, rsc, r, ken2
http://codereview.appspot.com/163058