On 2008-07-24, at 08:54EDT, André Bargull wrote:
We should eventually support that in our parser (and prohibit super
calls in the body of a constructor).
Is there already a JIRA-ticket for this, or will you create one?
Need one.
Rather than move the inits completely out of the constructor and in
to the `construct` method, why not just move them to after the
super call?
That won't work, because the LzNode constructor calls its construct-
method, so everything after the super-call in the constructor will
happen after LzNode#construct() has been called. Which means the
LzDatapath "take datapath-slot of immediateparent"-logic did already
happen.
Ah. Right.
There's a bunch of changes in here that are not related to the
title of the patch. They look ok, but maybe you should at least
make a comment about them in your change message?
- added typing for LzReplicationManager#getCloneForNode(..)
- added typing for LzReplicationManager#setVisible(..)
- removed LzReplicationManager#op_orderf, because it was superfluous
(the same information as in LzReplicationManager#orderpath)
- renamed LzReplicationManager#comp_orderf to
LzReplicationManager#comparator
- moved LzReplicationManager#mergesort() (simply for legibility and
consistency, first __LZHandle*Nodes() and then helper-methods)
- added doc LzReplicationManager#mergesort()
(I've added these information to the change textfile).
Great. Please check in!
On 7/24/2008 2:37 PM, P T Withington wrote:
In fact, in JS2, you will not have the option of calling super at
random points in your constructor; it will automatically be called
first. The reason is, you should not be touching the `this` object
until it is at least a fully initialized member of your
superclass. In JS2, the only thing you can do in a constructor is
to specify how your constructor's arguments are mapped to your
superclass constructors parameters. The syntax is something like:
function MyClass(a, b, c) : super(b, a) { ... }
We should eventually support that in our parser (and prohibit super
calls in the body of a constructor).
Rather than move the inits completely out of the constructor and in
to the `construct` method, why not just move them to after the
super call?
There's a bunch of changes in here that are not related to the
title of the patch. They look ok, but maybe you should at least
make a comment about them in your change message?
Otherwise, approved.
On 2008-07-23, at 17:45EDT, André Bargull wrote:
Change 20080708-bargull-rwY by [EMAIL PROTECTED] on
2008-07-08 20:14:02
in /home/Admin/src/svn/openlaszlo/trunk
for http://svn.openlaszlo.org/openlaszlo/trunk
Summary: replication-manager takes datapath-slot in swf9
New Features:
Bugs Fixed: LPP-6539
Technical Reviewer: ptw
QA Reviewer: hminsky
Doc Reviewer: (pending)
Documentation:
Release Notes:
Details:
moved some code and learned that swf9 initializes _all_ class
members in the constructor (in constrast to a prototype based
system).
Tests:
attached at bugreport
Files:
M WEB-INF/lps/lfc/data/LzReplicationManager.lzs
M WEB-INF/lps/lfc/data/LzLazyReplicationManager.lzs
Changeset: http://svn.openlaszlo.org/openlaszlo/patches/20080708-bargull-rwY.tar
