aboutsummaryrefslogtreecommitdiffstats
path: root/NEWS
blob: ac3af07da4685bd878b37821ff33f50817e97888 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
- monitor-get-edid-using-vbe:
  o remove all x86emu and vga softbootloader code, and always use the
    LRMI interface which was previously the backup one on 32-bit x86;
    libx86 is now used instead of liblrmi, allowing the use of LRMI
    interface on non-x86 hosts as well (fixes Mandriva bug #53866, which
    was caused by a bug in the removed code)
  o fix the retrieval of VBE vendor strings when using the LRMI interface
  o retrieve up to 4 EDID extension blocks
  o add --skip-vbe-check for skipping call for VBE info; useful if
    calling the program multiple times for different ports
  o on failure, return exit status 2 if no successful VBE calls were made
  o never retry in console mode if the card reports that the port does
    not support DDC (usually this means that the port has no display
    devices connected)

- monitor-parse-edid:
  o print EDID version and the number of EDID extension blocks
  o accept EDID data with multiple extension blocks
  o add support for Detailed Timing Descriptors in CEA EDID Timing
    Extension
  o add support for Short Video Descriptor formats 1-64 (as defined by
    EIA/CEA-861) in Video Data Blocks of CEA EDID Timing Extension
  o set Interlace flag for ModeLines created for interlaced Detailed
    Timings (and note it in the ModeLine comment)
  o correctly parse some interlaced video timings if the Detailed Timing
    Descriptor specifies field vertical parameters instead of the frame
    parameters (at least Fujitsu-Siemens Myrica VQ32-1T testcase)
  o ignore 1x1 modes (at least Nokia Valuegraph 447w testcase)
  o add support for standard timing descriptors in monitor descriptors
  o add comment "Monitor preferred modeline" only for the modeline the
    monitor reports as preferred, and add "Monitor supported modeline"
    comment for the other modelines

- monitor-edid, monitor-get-edid:
  o do not get duplicate EDIDs if the same EDID is available via
    multiple methods
  o allow retrieval of multiple EDIDs via VBE

- monitor-edid:
  o add --first support for stopping processing after one EDID has been
    found

Version 2.5 - 18 October 2009, by Anssi Hannula

- monitor-probe-using-X:
  o disable glx module to speed up X startup, especially if a proprietary
    glx module is in use (fixes a timeout with slow filesystems)
  o use -sharevts for X server when plymouth is active, as VT switching
    does not work at that point (fixes Mandriva bug #53736)

Version 2.4 - 1 October 2009, by Anssi Hannula

- monitor-parse-edid:
  o ignore the Manufacturer Specified Range Timing descriptor if the
    first detailed timing descriptor appears to violate it (this probably
    means that the descriptor is actually something else or in a
    different format; this fixes Lenovo W500 detection, reported by Udo
    Rader)

Version 2.3 - 6 September 2009, by Anssi Hannula

- monitor-get-edid-using-vbe:
  o update x86emu from xserver git (fixes issues at least with newish
    NVIDIA cards)
  o check if the port supports DDC before trying to read EDID data (this
    should prevent problems when probing nonexistent ports on old cards)
- monitor-edid, monitor-get-edid:
  o probe ports 0-2 by default instead of just 0-1 (many NVIDIA cards
    seem to be using port 2 and the changes in monitor-get-edid-using-vbe
    should make it safe for old problematic NVIDIA cards as well)

Version 2.2 - 16 August 2009, by Anssi Hannula

- monitor-edid, monitor-get-edid:
  o support getting EDID from kernel DRM when kernel mode-setting is
    enabled
- monitor-probe-using-X:
  o use resolution from LVDS initial mode when X reports that it is exact,
    making it work with nouveau driver

Version 2.1 - 7 July 2008, by Pixel

- monitor-parse-edid:
  o compute and display the "dpi" of the preferred modelines
  o handle parsing of EDIDs found in "xrandr --prop" or Xorg.log
- monitor-get-edid-using-vbe:
  o fix checking the current vt
  o instead of checking vt >= 7 to know if X is running, 
    check wether $DISPLAY is set

Version 2.0 - 8 April 2008, by Pixel

- change license from LGPLv2+ to GPLv3+
- drop cvt (xorg now bundles a similar version)
- remove internal lrmi.c, build with external liblrmi instead
  (Remi Collet and Ville Skyttä)

Version 1.16 - 10 March 2008, by Pixel

- do not install monitor-get-edid-using-vbe on archs where VBE is not
  available (Remi Collet)

Version 1.15 - 23 January 2008, by Pixel

- monitor-probe:
  o probe "using DMI" before "using X"
- monitor-probe-using-X:
  o in last resort, get Intel BIOS mode when "BIOS panel mode is bigger than
    probed programmed mode"

Version 1.14 - 10 January 2007, by Pixel

- monitor-edid, monitor-get-edid:
  o call monitor-get-edid-using-vbe with a range of ports, it stops on first
    success (by default it tries port 0 then port 1) (fixes mdvbz#38799)

Version 1.13 - 8 January 2007, by Pixel

- monitor-get-edid: 
  o skip /proc/acpi/video/**/EDID files which can't be valid (mdvbz#34417)
- monitor-get-edid-using-vbe: 
  o minimal support for getting EDID from different DDC port
    (experimental, need testing before using it in monitor-edid)