https://sourceforge.net/p/ruby-gnome2/feature-requests/2011-12-20T11:07:02ZRecent changes to feature-requests2011-12-20T11:07:02Z2011-12-20T11:07:02ZMichel Demazurehttps://sourceforge.net/u/demazure/https://sourceforge.net48194357710ca97b1330ebab8d286b9318ba8b06

When using FileChooserDialog in gtk2-1.0.3-x86-mswin32, I get warnings : the hi-color theme is not found, nor the gtk-file icon. This theme should be included.

2009-12-28T13:58:04Z2009-12-28T13:58:04ZVincent Carmonahttps://sourceforge.net/u/vinc-mai/https://sourceforge.net5a8a1cf3423ca96bae3e3718589e481ca1bf0e0a

Hi team. It seems that there is no way to parse tag. Tag messages are emited on the bus attached to a pipeline but nothing can be done with them. I see no usefull public method for Gst::Message::TAG. A parse method could be handy. There was a parse\_tag method for the old binding. See rb\_gst\_message\_parse\_tag on https://trac.luon.net/ruby-gstreamer0.10/browser/trunk/ruby-gstreamer0.10/src/rbgstmessage.c. Do not know if it can help. Vincent.

2009-03-06T15:55:54Z2009-03-06T15:55:54ZRoger Packhttps://sourceforge.net/u/rogerdpack/https://sourceforge.netfb56c035e8d8a8873f01951cf22aa375916c32d8

Is there a windows installer executable for the latest version? Might be nice, since it's confusing to not see one \[apparently\] anywhere. Thanks\!

2008-10-23T04:44:07Z2008-10-23T04:44:07ZVincent Carmonahttps://sourceforge.net/u/vinc-mai/https://sourceforge.nete2a1f0edeb565d25851b7d9ebec2f9b9e03fd653

Hi. I would like to display file with matching different extensions in a Gtk::FileChooserDialog. I try to use Gtk::Filter with the add\_pattern method. Given the documentation a shell style glob must be used: http://ruby-gnome2.sourceforge.jp/hiki.cgi?Gtk%3A%3AFileFilter\#add\_pattern With bash, ls \*.\{extension1,extension2\} works fine but the same pattern does not work with Gtk::Filter.add\_pattern. I join a file to demonstrate this \(code is dirty\). Gtk::BUILD\_VERSION => \[2, 14, 3\]

2007-02-21T12:07:51Z2007-02-21T12:07:51ZMartin Nordholtshttps://sourceforge.net/u/enselic/https://sourceforge.net5d8badc9a79bb41121e5868649714ecf9eeb4c88

It is convenient to connect an Gtk::Action directly to a signal in one line, like this: actionGroup.add\_action\( Gtk::Action.new\( params \)\).signal\_connect\( "activate" \) \{ self.on\_some\_action \} To do this, Gtk::ActionGroup needs to return the added Gtk::Action instead of self when add\_action is called. When \(if\) the patch is applied, I will edit the documentation wiki.

2006-08-30T17:20:12Z2006-08-30T17:20:12ZGustav Munkbyhttps://sourceforge.net/u/grddev/https://sourceforge.netc4aed6a7ebf74c61ad15f7907a8171e8d3d495e5

I currently have an implementation of a directed adjacency graph, which I would like to display in a TreeView. Currently, I'm forced to copy my values and the adjacency information into a TreeStore. The TreeStore however does not know that the the same node can be reached by many different paths in a graph, and therefore synchronizing the content of the treestore with the graph becomes burdensome. In this case it would be more natural with a new implementation of TreeModel, which would allow me to define the appropriate behaviour for the different methods. A straight-forward implementation of such a device would mean implementing the GtkTreeModel interface and delegating all functions to a Ruby object. Since I know very little about both Ruby C programming as well as GTK C programming, I don't know how complicated this request is. Judging by the following tutorial, the second part doesn't seem so complicated. http://scentric.net/tutorial/sec-custom-models.html In a dream world, the interface would be "Rubyfied" and one would be able to write a nice "Rubyish" implementation of a tree model, where iters would be replaced by normal Ruby objects, and get\_children should return an enumeration rather than a pointer, and so on. This I guess would be more complicated, and not really "neccessary". I'm writing this as a feature request because a search for TreeModel on the mailing lists returned lots of threads about other people wanting to implement TreeModels in Ruby.

2006-07-06T20:33:41Z2006-07-06T20:33:41ZAnonymoushttps://sourceforge.net/u/userid-None/https://sourceforge.net4f04dbc6bbd6b6e4a69be7e793c6de9a4647100a

GladeXML.new doesn't connect signals to methods defined in the glade file unless there's an empty block is given. For example suppose we have a glade file with a button on a window and we connect "clicked" signal of "button1" to hide method of "window1" using Glade. When I load the file with the command: GladeXML.new\("hello.glade"\) this connection does not work. But if I use the command: GladeXML.new\("hello.glade"\) \{ \} it works. Forcing the user of the API to supply an meaningless empty block seems like a bug. GladeXML.new should connect signals defined in the glade file in case no block is given. Regards, O. Murat Homurlu, ozgurmurath \{at\} gmail com

2006-06-16T12:26:39Z2006-06-16T12:26:39ZMichal Suchanekhttps://sourceforge.net/u/hramrach/https://sourceforge.net050be94f4107f94fca1c5b3274020a239a7ac71a

Hello I would like to use a filter on the data passed to the appliaction running in the terminal. The current interface offers two solutions: a\) run an application in a pty manually, and forward data between a Terminal instance and the application. However, this cannot use the nice convenient fork\_command. b\) use fork\_pty to run the filter inside a new ruby process. But this does not allow easy manipulation of the filter from the original terminal application. One use for such filter would be to use romkan to convert the input from romaji to kana. Thanks Michal

2006-04-05T14:12:14Z2006-04-05T14:12:14ZMathew Schafferhttps://sourceforge.net/u/schapht/https://sourceforge.net2579023a8a4af111ded77ae5fea890b3d9855899

Granted at version 0.14, I wouldn't expect a gem any time real soon, but it'd be a great future enhancement.

2005-09-17T20:42:01Z2005-09-17T20:42:01ZAnonymoushttps://sourceforge.net/u/userid-None/https://sourceforge.net895d4d4c192ce88e1da2ae294e0fa328a875a4f9

If I execute the attached piece of code ruby-gnome2 tells me this: >>> snip <<< ./test.rb:10:in \`initialize': fundamental type GBoxed isn't supported \(TypeError\) from ./test.rb:10:in \`new' from ./test.rb:10 >>> snap <<< An error warning like this would be more useful: the class MyModel does not include the Gtk::TreeModel interface