[Gtk-sharp-list] Design for Indexer for GLib.List to allow [] based addressing.

Mike Kestner mkestner@speakeasy.net
Sat, 25 Oct 2003 00:00:55 -0500

On Fri, 2003-10-24 at 17:38, Daniel Kornhauser wrote:
> The change is only four miserable lines in ListBase.cs:
> 		internal ListBase (System.Type element_type)
> 		{
> 			list_ptr = (IntPtr) 0;

list_ptr = IntPtr.Zero;

> 			this.element_type = element_type;
> 		}
> one line in SList.cs :
> 	public SList (System.Type element_type) : base (element_type)
> 		{
> 		}

You can put those braces on the line.  i.e.  : base (element_type) {}

> I tested it and it works alright should I commit this too ?
> (When/If I get miguel's cvs account)

You should submit patches in the form of the output from diff -u.  That
makes it easier for the reviewer to apply the code to their own tree to
try it out.  It also makes the reviewer more confident that it's been
compiled and tested.  ;-)  I believe that you have done so, btw. Just
suggesting the best practices for patch submission, which are pretty
standard across open source projects.

> Sorry for the stupid question, but should I even be asking if I have to
> make the commit, since mkenster didn't make a negative comment on Joe's
> suggestion ?

All patches should be submitted to the list, or attached to an open
bugzilla bug, or both. Nothing should be committed until an approval
comes from the maintainer.  I try not to be fascist in my review, and
love to get patches, but I also want to make sure the binding is as
clean as possible. 

I refrained from comment on Harold's delicate sensibilities, and you're
right that I would've said something if I thought it was a bad idea, but
somebody still needs to submit a patch.

> I want to avoid making unnecessary noise on the list.

Patches are wonderfully necessary noise.  I think everyone should make
some noise.  I'm officially declaring the next 7 days Gtk# patch week.
Don't miss out on the fun, everyone.

Mike Kestner <mkestner@speakeasy.net>