00:51:31 <buc> tomman: njsg: frg: Regarding https://ircbot.comm-central.org:8080/seamonkey/20220822#c197807 ...
00:51:37 <buc> Reverting 1485179-2 fixes the issue.
00:53:21 <buc> Strange commit. First, they increase bitfield size from 3 to 4, but then eliminate this change using "& 0x7" in an assignment.
00:54:24 <buc> But do it in one place only. If you do it (& 0x7) in all two places (which seems the same as reverting of the patch), the problem gone.
00:56:31 <buc> I believe it is better to reverth this one, since it affects SM now. If somebody have a time, please bisect nightly Firefox builds to determine where and when this issue gone there.
00:56:36 <tomman> tables are hard, let's go shopping~
00:59:25 * a-865 waves to buc
00:59:41 <buc> a-865: Hi
01:02:11 <buc> Instead of 1485179-2 reverting, just add "& 0x7" into https://gitlab.com/seamonkey-project/seamonkey-2.53-mozilla/-/blob/2_53_14_beta_01/layout/tables/nsTableFrame.cpp#L5371, the same way as it done in https://gitlab.com/seamonkey-project/seamonkey-2.53-mozilla/-/blob/2_53_14_beta_01/layout/tables/nsTableFrame.cpp#L5330
01:06:54 <buc> tomman: Do you have a reliable test case for https://ircbot.comm-central.org:8080/seamonkey/20220824#c197839 ? I opened the link, but there was no TypeError messsage in my case.
01:07:33 <tomman> buc: it seems they updated their pile of Javascript vomit
01:08:40 <tomman> I still get the weak map key error, but this time caught inside a Unhandled Promise rejection from their Javascript framework vomit
01:09:03 <tomman> so far it's the only site where I've seen the weak map key issue