|Summary:||[patch] x11-toolkits/vte3 update to 0.62.0 broke mate-terminal|
|Product:||Ports & Packages||Reporter:||rkoberman|
|Component:||Individual Port(s)||Assignee:||freebsd-gnome (Nobody) <gnome>|
|Severity:||Affects Only Me||CC:||rkoberman|
Description rkoberman 2020-10-05 04:52:40 UTC
After update of fribidi to , mate-terminal fails with seven repetitions of: (mate-terminal:45583): VTE-CRITICAL **: 14:04:47.253: void vte_terminal_match_set_cursor_type(VteTerminal *, int, GdkCursorType): assertion 'tag >= 0' failed The failure occurs after fribidi and vte3 are updated tp 1.0.10 and 0.62.0, respectively. No change to mate-terminal, itself. x11/roxterm works with the updated fribidi and vte3. roxterm has not been updated (except for the -fno-common issue) since before the glib20 update. Still, it looks like an issue between vte3-62 and fribidi-1.0.10 triggered by some cursor related call.
Comment 1 rkoberman 2020-10-07 02:41:04 UTC
Created attachment 218579 [details] patch to upgrade x11-toolkits/vte3 to 0.62.1 Patch to x11-toolkits/vte3 to correct error when used by mate-terminal.
Comment 2 Baptiste Daroussin 2020-10-09 21:18:28 UTC
I took the upgrade, but I won't mfh it as on 2 different boxes I didn't experience the issue you had with mate-terminal. If someone else also repo than I will MFH it.
Comment 3 commit-hook 2020-10-09 21:19:03 UTC
A commit references this bug: Author: bapt Date: Fri Oct 9 21:18:11 UTC 2020 New revision: 551830 URL: https://svnweb.freebsd.org/changeset/ports/551830 Log: update to 0.62.1 PR: 250130 Submitted by: email@example.com Changes: head/x11-toolkits/vte3/Makefile head/x11-toolkits/vte3/distinfo head/x11-toolkits/vte3/pkg-plist
Comment 4 rkoberman 2020-10-09 21:58:23 UTC
(In reply to Baptiste Daroussin from comment #2) I believe that the failure I saw was due to my customized configuration and most will not see it. It involved a gtk3 routine to set the cursor dependent on its location. That routine was deprecated some time ago but was only replaced in vte3 recently. Unless someone else hits the problem, I see no likelihood that many will see it. Thanks for the commit, in any case!