From: Yehuda Katz <wycats@...>
Date: 2009-02-25T02:48:43+09:00
Subject: [ruby-core:22422] Re: YASNP (Yet Another Selector Namespace 	Proposal)

--001636ed63f40f101a0463adbd30
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Matz,

I'm fairly amused by your response. If you go back to my very first post on
this topic, the OP was for file-based selector namespaces. Charlie had
already hacked together a working implementation before the discussion on
this list.

I agree entirely that using a block is not ideal, because it's not
*actually* a block (and might make people think there are block semantics
involved when there aren't actually any). The alternative I proposed (using
class/method/while syntax to lexically enclose the code to use the
namespace) is acceptable to me, but more granular than I think will actually
be used in real life.

The salient part of my proposal is that the heavy lifting could be done at
parse-time (a special call node that remembers the namespace or a special
call instruction with the namespaces in it when compiled) and then cached,
so that a namespaced call would be roughly the same speed (and cachability)
as a regular method call.

-- Yehuda

2009/2/24 Yukihiro Matsumoto <matz@ruby-lang.org>

> Hi,
>
> In message "Re: [ruby-core:22322] Re: YASNP (Yet Another Selector Namespace
>     Proposal)"
>     on Sun, 22 Feb 2009 11:35:41 +0900, Aaron Patterson <
> aaron@tenderlovemaking.com> writes:
>
> |+1 as well, but I like the block.  For me, it seems to be more clear as
> |to what is going on.
> |
> |  use Merb::String => String do
> |    "bla blah".some_special_merb_only_method
> |  end
> |
> |  "bla blah".some_special_merb_only_method # => doh!  doesn't exist
>
> +1 to have some kind of selector namespace (it's planned for 2.0
> anyway), but -1 for using block (or block-like syntax) to denote
> namespace scope, because
>
>  * it's not really a block
>  * too small granularity for namespace switching might bring confusion
>
> I'd prefer per file namespace switching.
>
>
>                                                        matz.
>
>


-- 
Yehuda Katz
Developer | Engine Yard
(ph) 718.877.1325

--001636ed63f40f101a0463adbd30
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Matz,<br><br>I&#39;m fairly amused by your response. If you go back to my v=
ery first post on this topic, the OP was for file-based selector namespaces=
. Charlie had already hacked together a working implementation before the d=
iscussion on this list.<br>
<br>I agree entirely that using a block is not ideal, because it&#39;s not =
*actually* a block (and might make people think there are block semantics i=
nvolved when there aren&#39;t actually any). The alternative I proposed (us=
ing class/method/while syntax to lexically enclose the code to use the name=
space) is acceptable to me, but more granular than I think will actually be=
 used in real life.<br>
<br>The salient part of my proposal is that the heavy lifting could be done=
 at parse-time (a special call node that remembers the namespace or a speci=
al call instruction with the namespaces in it when compiled) and then cache=
d, so that a namespaced call would be roughly the same speed (and cachabili=
ty) as a regular method call.<br>
<br>-- Yehuda<br><br><div class=3D"gmail_quote">2009/2/24 Yukihiro Matsumot=
o <span dir=3D"ltr">&lt;<a href=3D"mailto:matz@ruby-lang.org">matz@ruby-lan=
g.org</a>&gt;</span><br><blockquote class=3D"gmail_quote" style=3D"border-l=
eft: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left:=
 1ex;">
Hi,<br>
<br>
In message &quot;Re: [ruby-core:22322] Re: YASNP (Yet Another Selector Name=
space =A0 =A0 Proposal)&quot;<br>
<div class=3D"Ih2E3d"> =A0 =A0on Sun, 22 Feb 2009 11:35:41 +0900, Aaron Pat=
terson &lt;<a href=3D"mailto:aaron@tenderlovemaking.com">aaron@tenderlovema=
king.com</a>&gt; writes:<br>
<br>
|+1 as well, but I like the block. =A0For me, it seems to be more clear as<=
br>
|to what is going on.<br>
|<br>
| =A0use Merb::String =3D&gt; String do<br>
| =A0 =A0&quot;bla blah&quot;.some_special_merb_only_method<br>
| =A0end<br>
|<br>
| =A0&quot;bla blah&quot;.some_special_merb_only_method # =3D&gt; doh! =A0d=
oesn&#39;t exist<br>
<br>
</div>+1 to have some kind of selector namespace (it&#39;s planned for 2.0<=
br>
anyway), but -1 for using block (or block-like syntax) to denote<br>
namespace scope, because<br>
<br>
 =A0* it&#39;s not really a block<br>
 =A0* too small granularity for namespace switching might bring confusion<b=
r>
<br>
I&#39;d prefer per file namespace switching.<br>
<br>
<br>
 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0matz.<br>
<br>
</blockquote></div><br><br clear=3D"all"><br>-- <br>Yehuda Katz<br>Develope=
r | Engine Yard<br>(ph) 718.877.1325<br>

--001636ed63f40f101a0463adbd30--