Add a setting for enabling region based columns.

Add a setting for enabling region based columns.

The intention is to enable this if the --enable-experimental-webkit-features
command line switch is specified. That needs to be done from the top-level
chromium repo, and there'll be a separate patch for that.

BUG=242905
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=152304

[mstensho (USE GERRIT), 2013-05-22 14:21:00]

I have a corresponding patch for the chromium repo ready, but I suppose I'll
have to wait until this one has landed.

[mstensho (USE GERRIT), 2013-05-28 12:11:01]

On second thought, lumping this together with
--enable-experimental-webkit-features seems like a bad idea. It seems that
run-webkit-tests somehow supplies this switch automatically, and that's going to
hurt, because:

1. We have reftests that compare the old and new multicol implementation (using
<script>internals.settings.setRegionBasedColumnsEnabled(true)</script>)

2. Many other multicol tests will still fail or simply crash with the region
based multicol implementation.

In other words, if --enable-experimental-webkit-features enables region based
columns, it will effectively be enabled for everyone running tests.

Should I submit a chromium patch that adds --enable-region-based-columns
instead, then?

[abarth-chromium, 2013-05-28 17:13:03]

It's hard to answer that question without seeing the Chromium-side CL.

Whether you want to lump this in with other experimental features depends on how
stable this code is and what sorts of feedback you're looking for.  If the code
is unstable and under development, it should be in the "testing" category where
its enabled during testing.  If it's more stable and you're looking for
developer feedback, it should be in the "experimental" category so that it's
easy for developers to enable as an experiment.

[mstensho (USE GERRIT), 2013-05-28 19:22:18]

I can submit the Chromium-side CL, but it won't obviously compile without this
patch. Also, it's not really interesting. It's just about making some command
line flag call WebSettings::setRegionBasedColumnsEnabled(), and nothing else.

The region based multicol code is neither stable nor finished, and the current
multicol implementation is still more feature-complete. I'm working on this new
region based multicol implementation together with David Hyatt. He started
working on that last year.

Developers do not want to enable this just yet. But I want to, and I need a way
to do it. :) One way is a command line flag of some sort. Another option could
be to check some environment variable. Other suggestions?

[abarth-chromium, 2013-05-29 01:03:56]

We generally prefer command line flags to environment variables.  Whatever you
decide to do in Chromium, this CL LGTM.

[commit-bot: I haz the power, 2013-06-12 08:28:32]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mstensho@opera.com/15744005/1

[commit-bot: I haz the power, 2013-06-12 08:28:33]

Failed to apply patch for Source/WebKit/chromium/public/WebSettings.h:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file Source/WebKit/chromium/public/WebSettings.h
  Hunk #1 FAILED at 99.
  1 out of 1 hunk FAILED -- saving rejects to file
Source/WebKit/chromium/public/WebSettings.h.rej

Patch:       Source/WebKit/chromium/public/WebSettings.h
Index: Source/WebKit/chromium/public/WebSettings.h
diff --git a/Source/WebKit/chromium/public/WebSettings.h
b/Source/WebKit/chromium/public/WebSettings.h
index
fad73e0d435857ffeb1b005f26a861d71017ff58..d6c772c5a0a5f66ccac300a4ab76bf8eaa7a6ee8
100644
--- a/Source/WebKit/chromium/public/WebSettings.h
+++ b/Source/WebKit/chromium/public/WebSettings.h
@@ -99,6 +99,7 @@ public:
     virtual void setEnableTouchAdjustment(bool) = 0;
     virtual void setExperimentalCSSCustomFilterEnabled(bool) = 0;
     virtual void setExperimentalCSSGridLayoutEnabled(bool) = 0;
+    virtual void setRegionBasedColumnsEnabled(bool) = 0;
     virtual void setCSSStickyPositionEnabled(bool) = 0;
     virtual void setExperimentalCSSVariablesEnabled(bool) = 0;
     virtual void setExperimentalWebGLEnabled(bool) = 0;

[mstensho (USE GERRIT), 2013-06-12 09:41:44]

Do I have to wait for another LGTM, or can I just go ahead and commit?

[abarth-chromium, 2013-06-12 18:42:28]

LGTM

[commit-bot: I haz the power, 2013-06-12 19:26:45]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mstensho@opera.com/15744005/9001

[commit-bot: I haz the power, 2013-06-12 21:34:18]

Change committed as 152304