Re: Support variable changing in SDK interface and v8 native (issue 11662019)

[peter...@gmail.com]

ping

https://codereview.chromium.org/11662019/

[apa...@chromium.org]

LGTM


https://codereview.chromium.org/11662019/diff/13001/plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java
File
plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java
(right):

https://codereview.chromium.org/11662019/diff/13001/plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java#newcode307
plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java:307:
return new RelayOk() {
Just occurred to me. Why don't you keep a singleton instance for this?

https://codereview.chromium.org/11662019/

[peter...@gmail.com]

https://codereview.chromium.org/11662019/diff/13001/plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java
File
plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java
(right):

https://codereview.chromium.org/11662019/diff/13001/plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java#newcode307
plugins/org.chromium.sdk/src/org/chromium/sdk/internal/v8native/DebugSession.java:307:
return new RelayOk() {

On 2013/01/10 13:59:20, apavlov wrote:
> Just occurred to me. Why don't you keep a singleton instance for this?

Mostly because I don't want to spoil outer namespace with a constant
that is never used (see fixed condition above).

https://codereview.chromium.org/11662019/

[apa...@chromium.org]

Not sure what you mean. I implied something like RelayOk.instance().

https://codereview.chromium.org/11662019/

[peter...@gmail.com]

> Not sure what you mean. I implied something like RelayOk.instance().

This doesn't work like this. This value is of symbolical meaning. I have to
prove to myself and to everybody reading this that relay operation was
finished
correctly. I make it in form of returning a token (that is never used on
runtime, but demanded by compiler). It's the comment inside that matters the
most.

https://codereview.chromium.org/11662019/

[apa...@chromium.org]

Yes, I remember its meaning, and since the actual token identity is never
checked, yet there may be lots and lots of similar tokens generated at
runtime,
I thought it would be plausible to use a single token instance everywhere.

https://codereview.chromium.org/11662019/