… | |
… | |
43 | TODO: think about daemon delaying urxvt client exit. |
43 | TODO: think about daemon delaying urxvt client exit. |
44 | IDEA: use harfbuzz to render combining sequences, maybe also complex scripts |
44 | IDEA: use harfbuzz to render combining sequences, maybe also complex scripts |
45 | TODO: rclock horribly time-biased |
45 | TODO: rclock horribly time-biased |
46 | TODO: rclock iso date mode? |
46 | TODO: rclock iso date mode? |
47 | TODO: printf "\e[4:3m" - "undercurl" - should probbaly be interpreted as underline, rather than \e[43m |
47 | TODO: printf "\e[4:3m" - "undercurl" - should probbaly be interpreted as underline, rather than \e[43m |
|
|
48 | TODO: feature where I can just click on (y/p/n) when prompted after pasting text to the terminal? |
|
|
49 | |
|
|
50 | TODO: 08:11:16 <dg> in some cases printf "\e]50;\a" can be a crash |
|
|
51 | 08:11:50 <dg> it seems to depend a bit on setup, two in a row maybe needed |
|
|
52 | - what font dpes urxvtr flal back ion this case? why not fixed? |
|
|
53 | 07:28:56 <dg> schmorp: try printf '\e]50;\a\xe7\xaf\xad' |
|
|
54 | 07:29:15 <dg> i'm guessing the builtin font doesn't have that character (篭) |
|
|
55 | 07:29:40 <dg> on one machine with 9.30 that segfaults, with what looks like a null pointer deref |
|
|
56 | 07:29:58 <dg> on latest dev, but a different machine it hangs |
|
|
57 | 07:33:26 <dg> hmm, and on openbsd it doesn't even hang with that reduction, but the actual case the fuzzer |
|
|
58 | found does hang |
|
|
59 | 07:34:02 <dg> curl -o - dg.cx/f12-afl-urxvt-id0-src517.txt |
48 | |
60 | |
49 | TODO: nmeum reports that when the wm resizes urxvt while it opens, the prompt is now somewhere in the middle |
61 | TODO: nmeum reports that when the wm resizes urxvt while it opens, the prompt is now somewhere in the middle |
50 | of the window, rather than near the top. this behaviour change wss part of the wrapmode fix, and should normally |
62 | of the window, rather than near the top. this behaviour change wss part of the wrapmode fix, and should normally |
51 | prevent from readline jumping around on resizes. running urxvt with the exact geometry that the wm will assign "fixed" this. |
63 | prevent from readline jumping around on resizes. running urxvt with the exact geometry that the wm will assign "fixed" this. |
52 | https://github.com/exg/rxvt-unicode/commit/1eb2c326cb88ee7bae01e995a16829746bfe374e |
64 | https://github.com/exg/rxvt-unicode/commit/1eb2c326cb88ee7bae01e995a16829746bfe374e |
|
|
65 | |
|
|
66 | - fix memory leak when a background expression uses the muladd |
|
|
67 | or focus_fade operators (reported by Erik-Jan). |
|
|
68 | - fix reference counting of values returned by some perl bindings. |
|
|
69 | - fix OSC responses with 7-bit ST (reported by Stuart Henderson). |
|
|
70 | - fix mouse reporting of motion notify events to report 3 as |
|
|
71 | the button number when no button is pressed, same as xterm. |
|
|
72 | Previously urxvt reported the number of the last button |
|
|
73 | pressed or an invalid value, before the first button press |
|
|
74 | (reported by Uli Schlachter). |
|
|
75 | - change the encoding of buttons 8-11 in mouse reporting |
|
|
76 | sequences to match xterm's and enable reporting of release |
|
|
77 | events for them (patch by Pierre-Marc Fournier). |
|
|
78 | - clamp coordinates in mouse reporting sequences to the number |
|
|
79 | of rows and columns (patch by Nick Hanley). |
|
|
80 | - added Alt_L/ALt_R and updated some keys in iso 14755 5.2 mode. |
|
|
81 | - perl 5.38 does not handle empty locale names, use "C" instead |
|
|
82 | (patch by sxzzsf). |
|
|
83 | - add emoji test character, so that emoji-only fonts can be used without |
|
|
84 | any tricks. |
|
|
85 | |
53 | 9.31 Mon 02 Jan 2023 18:13:16 CET |
86 | 9.31 Mon 02 Jan 2023 18:13:16 CET |
54 | - implement a fix for CVE-2022-4170 (reported and analyzed by David Leadbeater). |
87 | - implement a fix for CVE-2022-4170 (reported and analyzed by David Leadbeater). |
55 | While present in version 9.30, it should not be exploitable. It is exploitable |
88 | While present in version 9.30, it should not be exploitable. It is exploitable |
56 | in versions 9.25 and 9.26, at least, and allows anybody controlling output to |
89 | in versions 9.25 and 9.26, at least, and allows anybody controlling output to |
57 | the terminal to execute arbitrary code in the urxvt process. |
90 | the terminal to execute arbitrary code in the urxvt process. |