|
Name |
|
Date |
Size |
#Lines |
LOC |
| .. | | 03-May-2022 | - |
| data/ | H | 30-Oct-2005 | - | 588 | 465 |
| debian/ | H | 07-May-2022 | - | 172 | 100 |
| intl/ | H | 30-Oct-2005 | - | 7,703 | 5,262 |
| pixmaps/ | H | 03-May-2022 | - | 275 | 209 |
| po/ | H | 07-May-2022 | - | 238 | 192 |
| src/ | H | 30-Oct-2005 | - | 2,607 | 1,908 |
| AUTHORS | H A D | 03-Jun-2005 | 62 | 5 | 3 |
| COPYING | H A D | 03-Jun-2005 | 17.6 KiB | 341 | 281 |
| ChangeLog | H A D | 30-Oct-2005 | 375 | 13 | 11 |
| INSTALL | H A D | 03-Jun-2005 | 7.6 KiB | 183 | 143 |
| Makefile.am | H A D | 05-Jun-2005 | 567 | 27 | 16 |
| Makefile.in | H A D | 30-Oct-2005 | 13.6 KiB | 454 | 370 |
| README | H A D | 05-Jun-2005 | 3.7 KiB | 127 | 82 |
| TODO.tasks | H A D | 03-Jun-2005 | 161 | 7 | 6 |
| acconfig.h | H A D | 03-Jun-2005 | 316 | 15 | 14 |
| acinclude.m4 | H A D | 03-Jun-2005 | 39 | 2 | 1 |
| aclocal.m4 | H A D | 30-Oct-2005 | 311.6 KiB | 8,852 | 8,013 |
| autogen.sh | H A D | 07-Jun-2005 | 3.6 KiB | 128 | 108 |
| config.guess | H A D | 30-Oct-2005 | 42.4 KiB | 1,454 | 1,257 |
| config.h.in | H A D | 03-Jun-2005 | 8.1 KiB | 303 | 214 |
| config.sub | H A D | 30-Oct-2005 | 30.7 KiB | 1,567 | 1,426 |
| configure | H A D | 30-Oct-2005 | 837.1 KiB | 27,412 | 22,688 |
| configure.in | H A D | 30-Oct-2005 | 3.2 KiB | 110 | 91 |
| gtkperf.anjuta | H A D | 05-May-2005 | 1.8 KiB | 55 | 54 |
| install-sh | H A D | 03-Jun-2005 | 6.3 KiB | 270 | 153 |
| ltmain.sh | H A D | 30-Oct-2005 | 179.7 KiB | 6,427 | 5,058 |
| missing | H A D | 03-Jun-2005 | 6.3 KiB | 199 | 159 |
| mkinstalldirs | H A D | 03-Jun-2005 | 722 | 41 | 23 |
| stamp-h.in | H A D | 07-Jun-2005 | 10 | 2 | 1 |
README
1GtkPerf
2==========
3
4What?
5GtkPerf is an application designed to test GTK+ performance. The point is to
6create common testing platform to run predefined GTK+ widgets (opening
7comboboxes, toggling buttons, scrolling text etc.) and this way define the
8speed of device/platform.
9
10Why?
11I know that bencmarking tools (including GtkPerf) can be fooled and don't give
12real-life results. Yet, I belive that GtkPerf can be helpful to solve for
13example this kind of things:
14
15 * How fast is my software platform compared to others?
16 * How fast is my hardware compared to others?
17 * How fast/slow GTK+ is with different themes?
18 * Is there any notably slow widgets when using my GTK+ theme(engine)?
19 * How much does GTK+ performance increase when moving e.g. from GTK+ 2.4 -> 2.6?
20 * How much does GTK+/GDK perfomance increase when updating X-server?
21 * Can GTK+ be efficiently used in embedded devices?
22
23Tests
24=====
25
26The tests are separate from each other with the same main logic: Making single
27test operation (opening combo box, drawing n circles, etc.) and waiting for that
28to get processed are returned to g_idle. So every test is jumping between test
29function and g_idle as fast as possible.
30
31These tests are provided (in 0.30).
32
33GtkEntry (on_idle_gtkentry_test)
34-----
35Switches between "Test String" and "Longer test string" texts for [count] times.
36
37
38GtkComboBox (on_idle_gtkcombobox_test)
39-----
40GtkComboBox contains 10 entrys "Selection 1"..."Selection 10". This test opens
41and closes GtkComboBox [count] times while selecting next entry.
42
43
44GtkComboBoxEntry (on_idle_gtkcomboboxentry_test)
45-----
46GtkComboBoxEntry contains 10 entrys "Selection 1"..."Selection 10". This test
47opens and closes GtkComboBoxEntry [count] times while selecting next entry.
48
49
50GtkSpinButton (on_idle_gtkspinbutton_test)
51-----
52GtkSpinButton value is increased with one [count] times. When 1000 is reached,
53value is changed back to 0.
54
55
56GtkProgressBar (on_idle_gtkprogressbar_test)
57-----
58GtkProgressBar is increased with 1% [count] times. Whenever bar gets full,
59its value is set to 0.
60
61
62GtkToggleButton (on_idle_gtktogglebutton_test)
63-----
64GtkToggleButton state is toggled between on/off [count] times.
65
66
67GtkCheckButton (on_idle_gtkcheckbutton_test)
68-----
69GtkCheckButton state is toggled between on/off [count] times.
70
71
72GtkRadioButton (on_idle_gtkradiobutton_test)
73-----
74Two GtkRadioButtons inside same group are witched to be selected alternately
75[count] times.
76
77
78GtkTextView - Add text (on_idle_gtktextview_addtext_test)
79-----
80Text "Future is Open. " is appended to GtkTextView [count] times.
81
82
83GtkTextView - Scroll (on_idle_gtktextview_scroll_test)
84-----
85Scrolls the text added in earlier test GTK_SCROLL_STEP_FORWARD and when end is
86reached GTK_SCROLL_STEP_BACKWARD for [count] times.
87
88
89GtkDrawingArea - Lines (on_idle_gtkdrawingarea_lines_test)
90-----
91Draws random size and color lines to GtkDrawingArea for 1000*[count] times.
92
93
94GtkDrawingArea - Circles (on_idle_gtkdrawingarea_circles_test)
95-----
96Draws random size and color circles to GtkDrawingArea for 1000*[count] times.
97
98
99GtkDrawingArea - Text (on_idle_gtkdrawingarea_text_test)
100-----
101Draws random size and color "Future is Open." strings with PangoLayout to
102GtkDrawingArea for 100*[count] times. Font used is "Sans Bold 16".
103
104
105GtkDrawingArea - Pixbuf (on_idle_gtkdrawingarea_pixbufs_test)
106-----
107Draws Duck.png to random place in GtkDrawingArea for 10*[count] times.
108
109
110
111maemo
112=====
113
114To make GtkPerf Hildonized" (see http://www.maemo.org) and make it work nicelu in Nokia 770, do this:
115
1161) Enable this line in configure.in:
117#MAEMO_MODULES="libosso hildon-base-lib hildon-lgpl hildon-libs"
118
1192) Enable this line in appdata.h:
120//#define HILDONIZED
121
122Now compile normally and GtkPerf uses maemo-specific changes in code!
123
124
125
126
127