Owen Taylor otay...@redhat.com writes:
I guess the basic question here is whether most stuff is updating the
text of a label or whether most stuff is adding 1 elements to a
GtkTreeView. If most things that applications are doing in response to
timeouts/IO completions/IPC are cheap and
Matthias Clasen matthias.cla...@gmail.com writes:
On Thu, Sep 2, 2010 at 7:24 PM, Havoc Pennington h...@pobox.com wrote:
I was thinking about the problem of wasting memory, writing to one
internal representation and not changing the other, etc. How about
this: we only keep one
Matthias Clasen matthias.cla...@gmail.com writes:
On Fri, Sep 3, 2010 at 3:22 AM, Soeren Sandmann sandm...@daimi.au.dk wrote:
Not that it matters that much, but if you use cairo_mask() with the
pixbuf data as both source and mask, the source as RGB24 and the mask
as ARGB32, you could
Yasushi SHOJI ya...@atmark-techno.com writes:
I'm for one wandering how to port my application to gtk+ 3.0. My
problem is that the current gdk pixbuf nor cairo does not have a
client side pixel buffer other than RGB888 or RGBA.
Note that cairo 1.10 (which GTK+ 3.0 will depend on, as I
Dana Jansens d...@cg.scs.carleton.ca writes:
There are 2 other BSTs that would work effectively to provide a
persistent BST, while also maintaining a very efficient implementation
for when persistence is not used. They are:
- Treaps
- Red-Black Trees
Treaps can be expected to outperform
Matthias Clasen matthias.cla...@gmail.com writes:
Next I implemented sorting. Because I use an array, I can use qsort(),
which is fast. I had thought about switching to GSequence to get ever
closer to GtkListStore, but I did not do that, and one of the reasons
is that GSequence uses
frederico schardong frede@gmail.com writes:
When I click with the mouse on the drawing area the callback
button_pressed is called and the g_print message shows, and when I
move the window the callback expose_event is called and shows its
g_print. The only difference between both g_print
Behdad Esfahbod [EMAIL PROTECTED] writes:
Also, my only quick feedback is that you are hooking up on the font
size, while you should instead hook on screen dpi. Think of screen dpi
as a global scaling factor, and font size is relative size of the fonts
to the other elements.
I'll repeat
Adam Jackson [EMAIL PROTECTED] writes:
On Wed, 2008-07-23 at 12:12 -0400, Matthias Clasen wrote:
On Wed, 2008-07-23 at 11:12 -0400, Adam Jackson wrote:
I hacked up a proof-of-concept addition to the protocol along these
lines:
At
http://www.daimi.au.dk/~sandmann/aggregates.patch
there is a patch to add aggregates to GSequence.
In a tree data structure, an 'aggregate' is information stored in a
node which is computed from that node's children. This field from
GtkRBTree:
/* count is the number of nodes
Alexander Larsson [EMAIL PROTECTED] writes:
A possibly interesting approach would be to put all the deprecated code
in another ELF section and place that section at the end. This will mean
that the deprecated stuff will be in one continous place and might make
paging in non-deprecated code
Bastien Nocera [EMAIL PROTECTED] writes:
On Wed, 2007-11-21 at 00:39 +0100, Soeren Sandmann wrote:
Hi
I have been doing some work on adding RandR 1.2 support to GTK+. There
is an initial patch here:
http://www.gnome.org/~ssp/randr/gtk.patch
Was that based on the patch
Kalle Vahlman [EMAIL PROTECTED] writes:
What's the purpose of gtk_menu_attach_to_widget ? And is there any
bad effect
if I create a gtk menu without using this function ?
In short, it's convenience function for tying the menu's lifecycle to
a widget. I think the only thing to
Hi
I have been doing some work on adding RandR 1.2 support to GTK+. There
is an initial patch here:
http://www.gnome.org/~ssp/randr/gtk.patch
The interesting part is perhaps the API:
A new signal on GdkScreen:
void (*monitors_changed) (GdkScreen *screen);
that is emitted whenever
BJörn Lindqvist [EMAIL PROTECTED] writes:
Incidentally, blitting pixbufs is slower than it has to be because its
format rarely matches the X11 server which uses either xRGB32 or
ARGB32.
I don't disagree with anything else you say, but this performance
issue is really a non-issue. We are
Alex Jones [EMAIL PROTECTED] writes:
A few years ago there used to be a distributor patch in Gentoo to enable
this, and it was sweet. What happened, here?
As far as I know the patch was never reviewed - and the bug was put on
future with no other comments. The bug in question is 103811.
Behdad Esfahbod [EMAIL PROTECTED] writes:
The main issues I see with it are (other than just the large number of
functions for something as simple as a sequence):
- GSequenceIter is used both as a node reference and as an iterator.
For example, the way to get/set the value of a node is to
Morten Welinder [EMAIL PROTECTED] writes:
That depends on the data and how much we have seen before, but if
we initially did a fair-sized sample -- last I looked we were doing 1000
rows -- then the probability is low and hence it is not annoying.
If you did that, then you could also
Jean-Yves Lefort [EMAIL PROTECTED] writes:
Maybe you can explain how the name sequence is more appropriate for
your container than for a GList, GSList, GQueue, GString, GArray,
GPtrArray or GByteArray?
It is not. I don't think GSortedList is accurate either, since a
GSequence does not have to
Johan Dahlin [EMAIL PROTECTED] writes:
Attached is a patch that adds a generic cclosure marshaller using libffi to
gobject. The reason for this is to allow signals of GObjects written in a
language binding to call signal callbacks in C.
It's not just a theoretical case, Edward Hervey recently
Soeren Sandmann [EMAIL PROTECTED] writes:
If accepted, I'll write API documentation and port GtkListModel to use
it. There is already a through testsuite SVN.
I wrote API documentation and committed it.
Changes since last time:
- Fixed a couple of bugs where the sequence would try
Hans Petter Jansson [EMAIL PROTECTED] writes:
(In retrospect using a splaytree was probably not a wise decision. A
red/black tree might have been better, even though it would have a
code-complexity cost in some areas).
Could we change the implementation (say, to RBT) without introducing
Hans Petter Jansson [EMAIL PROTECTED] writes:
I'm not sure it's correct to call it an iterator when it's just a
pointer to an internal data structure node. When a node is freed, any
iterator pointing to that node will be freed as well, and that might not
be what the user expects (e.g. she
Peter Robinson [EMAIL PROTECTED] writes:
I am writing a server application (using glib) in which I want to be
able to (frequently) turn polling on and off for given sockets.
You may want to take a look at this:
http://www.daimi.au.dk/~sandmann/watch.h
http://www.daimi.au.dk/~sandmann/watch.c
Havoc Pennington [EMAIL PROTECTED] writes:
I think retained mode is interesting when the objects have two
properties:
- Does the object contain state that the application doesn't *care*
about.
- Is the object generic enough that many applications would need it
That
Yevgen Muntyan [EMAIL PROTECTED] writes:
Much more people would make use of simple things like draw a
rectangle and drag it around if it wasn't so hard.
But applications generally need to do much more than draw a rectangle
and drag it around. If dragging rectangles around was really all there
Jim George [EMAIL PROTECTED] writes:
Am I doing something wrong? Is this a glade bug? I noticed this in
programs which use glade code-generation as well as libglade. Or could
it be because the popup menu is a top-level window in it's own
right?
Yes, the menu is a toplevel window in its own
Someone posted a link to the minutes of the gtk+ meeting at GUADEC
which reminded me that I promised to write up some of the ideas I had
for a canvas-type widget.
The reason model/view is a useful model is that it allows applications
to keep a data structure (the model) containing all the
Havoc Pennington [EMAIL PROTECTED] writes:
It depends on the goals of a canvas right - you're thinking of it as
drawing area plus hit testing while it can be thought of as solving
a wider set of problems.
In particular I think of it as a more flexible sort of widget system
more like HTML or
Tim Janik [EMAIL PROTECTED] writes:
whether you benefit in terms of memory size by using GSlice instead of
malloc/free depends heavily on your application. simply watch your
memory requirements with G_SLICE=always-malloc and without it.
In general, I'd recommend just always using
Tim Janik [EMAIL PROTECTED] writes:
In general, I'd recommend just always using
g_slice_set_config (G_SLICE_CONFIG_ALWAYS_MALLOC, 1);
This way the various tools, valgrind and memprof, will work with your
application, and you don't get any surprises with pathological cases
Matthias Clasen [EMAIL PROTECTED] writes:
On 4/25/06, Tim Janik [EMAIL PROTECTED] wrote:
i don't think there's a point in implementing grouping unless the above
question has been answered, i.e. we've gotten a resonable usage scenario.
so until that happens, it's probably best to leave
Tim Janik [EMAIL PROTECTED] writes:
That's because that feature simply doesn't work. Moving the mouse to a
tooltip-less area does not turn off fast mode as it should, regardless
of grouping.
well, given that behaviour is implemented, there's no need for seperate
groups, right?
I am not
Federico Mena Quintero [EMAIL PROTECTED] writes:
Spreading the is_composited/composited_changed API among GdkScreen and
GtkWidget means that people will just ignore the latter's signals, since
they won't ever see them.
I don't think so; they are much more likely to ignore the signal on
Soeren Sandmann [EMAIL PROTECTED] writes:
The attached patch implements all the properties dealing with
compositing managers that were recently added to the EWHM spec.
And here is the patch.
Soren
? birnan
? emfle
? gtkcm
? gtklabel.ac
? net-wm-all-apr1
? net-wm-all-apr12.patch
? net-wm
Johan Dahlin [EMAIL PROTECTED] writes:
Index: gdk/gdkwindow.h
+ GDK_WINDOW_TYPE_HINT_DESKTOP,
+ GDK_WINDOW_TYPE_HINT_DROPDOWN_MENU, /* A drop down menu (from a
menubar) */
+ GDK_WINDOW_TYPE_HINT_POPUP_MENU, /* A popup menu (from right-click) */
+
Matthias Clasen [EMAIL PROTECTED] writes:
Comments on the patch:
- The registration for the signal does not match the signature in the
class struct
(misses the boolean parameter).
Yeah, a last minute change. I want to remove the boolean parameter
since the information can be found through
Kristian Rietveld [EMAIL PROTECTED] writes:
1. Let's start out with a somewhat more general change. We need to be able
to cancel currently running asynchronous operations. For this we need
to have a handle on an operation, so we need to introduce a
GtkFileSystemHandle. The
Soeren Sandmann [EMAIL PROTECTED] writes:
Also, the callback is always called, even if you cancel the operation,
right?
If it isn't, then
- How do you free the user data you passed in?
- What good is the handle passed to the callback? You can't do
anything
One thing I and many other people regularly need is to be notified
whenever something happens on a file descriptor. This is needed when
you write networked applications, or when you with to communicated
with spawned child processes, and in many other circumstances.
The way you are supposed to do
Havoc Pennington [EMAIL PROTECTED] writes:
So random thoughts.
A few more random thoughts:
- Ability for one canvas to view more than one model at the same time
is useful so that you can have eg., the coordinate axes and the
graph be separate models for a graphing display.
- Zooming is
Benoit Carpentier [EMAIL PROTECTED] writes:
lines 199 to 209 :
if (GTK_WIDGET_MAPPED (widget)
paned-child1-allocation.height
child1_allocation.height)
{
gtk_widget_size_allocate (paned-child2,
child2_allocation);
gtk_widget_size_allocate (paned-child1,
Paul Pogonyshev [EMAIL PROTECTED] writes:
However, it seems that in the new (2.4 and up) toolbar API there is no
easy way to add a random widget to a toolbar (well, I could subclass
GtkToolItem, but...)
Any suggestions? A different design might solve a problem as well as a
simple way of
Lorena Salinas [EMAIL PROTECTED] writes:
Here is part of the function...
valid = gtk_tree_model_get_iter_first (GTK_TREE_MODEL (my_model),
iter);
This function also needs to be protected by the GTK+ lock, because the
GtkListStore rearranges the data structure every time it is accessed.
In bug 302486
http://bugzilla.gnome.org/show_bug.cgi?id=302486
there is a patch that adds the ability to get crossing events on a
widget basis, instead of on a window basis. What this means is that a
widget will receive enter/leave events even if it is NO_WINDOW widget
or covered by a
Bill Haneman [EMAIL PROTECTED] writes:
The ONLY time that an application should choose a specific color (or
even a specific value, i.e. 'black' or 'white') is when red means
RED, i.e. if the application is a painting application, an image
processing application, or something similar, where
[EMAIL PROTECTED] (Morten Welinder) writes:
fchmod if available?
The patch does use fchmod(). Are there systems we care about where
fchmod() is not available?
I don't think we care about the permissions being the same as the target
file's while the file is being created. In fact, might
Tor Lillqvist [EMAIL PROTECTED] writes:
Soeren Sandmann writes:
The patch does use fchmod(). Are there systems we care about where
fchmod() is not available?
Win32... No such thing in the C library. As far as I know, there isn't
even anything in the underlying Win32 API that would
[EMAIL PROTECTED] (Morten Welinder) writes:
1. It is hideously expensive.
This is already a fairly expensive function. And fork() is really not
that bad. My system can fork more than 20 times per second. If
you need to write tons of small files quickly, you should (a) not use
this
Paul Pogonyshev [EMAIL PROTECTED] writes:
Ah, indeed. The problem is that I use KDE as the desktop.
It seems to work when `gnome-settings-daemon' is running, however,
for some reason ``Priority text beside icons'' works exactly as
``Text below icons'' (it works fine when set from my
Soeren Sandmann [EMAIL PROTECTED] writes:
I think this is a serious problem. The file should be created with the
umask permissions. You can do this:
#ifndef G_OS_WIN32
{
mode_t mode = umask (0);
umask (mode);
chmod (filename, mode);
}
#endif
but then there is a race
Alexis S. L. Carvalho [EMAIL PROTECTED] writes:
+static gchar *
+write_to_temp_file (const gchar *contents,
+ gsize length,
nitpick
gssize to be consistent with g_file_replace?
/nitpick
I have changed it, but I don't see why it
matters. write_to_temp_file() is not called
Is it useful to add g_file_write() to complement
g_file_get_contents()? It seems a feature many applications could use.
Patch attached (not tested on Windows). It is also available at
http://www.daimi.au.dk/~sandmann/filewrite.patch
Søren
Index: gfileutils.c
Havoc Pennington [EMAIL PROTECTED] writes:
Should it do the atomic write thing? (write to a temporary file
alongside the new file, then rename() to the final filename)
The minus is that you need double the size of the file to write it, but
the plus is that you aren't screwed if you run out
54 matches
Mail list logo