frg_Away: do we still have a ghost window problem in seamonkey?

cause UXP found .. a solution but it seems a bit risky to me

nsITobin We have a ghost window but I don't see any problems with it. Mozilla recenztly kicked it out more or less. I think remains only on macOS but not sure. I would not want to tinker with it unless there is actually a problem.

i mean ghost windows that stay active because of scripts when they shouldn't

github windows chruning away long after the tab is closed

repo.palemoon.org/MoonchildProducti…265d63ee4231937ff1c1dd23529652df0c4 and repo.palemoon.org/MoonchildProducti…37a3181ff7b1aa5f3e8ad7fcc71137ae024 plus fe code to do something with it

comments indicate bad things could happen and then they added an idleobserver to their fe to just kill em off

and its intended to be debug code

nsITobi oh you mean the ghostbuster. I now and then see this but verxy rare. Ebay mostly and not as excessive

according to the .. mess of an issue seems as more stuff is added the worse it is gonna get

but then they say its a minority of sites..

and that it seems specifically masked by e10s multiprocess but in single process .. yeah so maybe a good idea I just question enabling debug code as a front line feature especially knowing this bit of the code has never been highlighted before

the PROPER fix i would think is to make the ghostbuster code better not just force it when idle

nsITobin btw. when it comes to MS Bullsh*t. Server 2025 is latest Windows and does not need later cxpu nor uefi nor secure boot nor tpm. So just arbitrary requirements to force people to upgrade hardware. Installed it in a vm for testing but don't plan to use it.

Status meeting in 34 minutes - wiki.mozilla.org/SeaMonkey/StatusMeetings/2024-11-10

frg_Away: the issue is not if you can bypass their arbitrary restrictions but the arbtary timing of when one of those MIGHT be technologically required cause they got around to refactoring or adding the target component

hardware crypto for instance

yeah but the will stay with Server 2025 till 2034. They can't change these.

wonder how badly compramised the win32 api is now

I really don't like the idea that a mere update will change functionality

with how they stagger out feature enablement

Status meeting in 15 minutes - wiki.mozilla.org/SeaMonkey/StatusMeetings/2024-11-10

hiya tonymec

hi IanN

hi .*

tonymec: how was Oktoberfest?

IanN: AFAIK that's Bavarian, not Belgian. Here we drink the same amounts of beer at all times of the year.

tonymec: i'm sure the Mozcamp or something we did in Belgian including attending some sort of Oktoberfest type event

greetings IanN

hello nsITobin

Status meeting in 4 minutes - wiki.mozilla.org/SeaMonkey/StatusMeetings/2024-11-10

hi there *

hi rsx11m

I’m here

hi WG9s-iPad

fancy

are there any real irc apps for an ipad?

Status meeting time - wiki.mozilla.org/SeaMonkey/StatusMeetings/2024-11-10

Hi all

op no time for that now the meeting is starting :P

Hello frg

Who's taking minutes?

me and late again I suspect

thanks frg

Nominees for Friends of the Fish Tank?

WG9s-iPad you had one I think

I nominate for for helping me to get about:credits so work on the official website

i second for aka frg

I'm for for 2

y0

fourthing for

thanks btu was also a group effort.

a collective "well done"

Action Items

bau

Status of the SeaMonkey Infrastructure

can we see if we can get mozilla to extend their infra cutloose until 115 is ended?

bau here too

ewong has been creating some new boxes, need to do some testing

Status of the SeaMonkey Source Tree

Finally did some reviews and central checkins. Still lots of cZ reviews outstanding but given that I see no problems in my daily use I will probably mostly rubberstamt then after a brief look. Please test 2.53.20b1 pre.

yes, please test a recently nightlt if you use cZ

*nightly

Did some more places backports to bring it up to 60 / 2.57. My error with drag and rop seems to be gone but also pelase test. tags and keywords got the most changes. We do not plan to remove them btw.

are there any more planned places changes that would affect the work I am doing to make the go menu the history menu?

Apple borked macOS 15.1 and SeaMonkey and a bunch of other programs will no longer start normally. Solution seems to be self signing the code. We don't have time to look into notarization yet.

frg: cZ (2.53.20b1pre 20241106113419): no mode symbols or colored balls in the user list, and all names are in bold

tonymec please use the latest version and try again

Oh I see it is the latest. ianN to the rescue :)

frg: AFAIK it is the latest version.

tonymec: anything in the error console or stdout?

Code is in bug 1927582

Release Train

Should we do the beta this week? I would probably let the cZ code in Bug 1927582 let mature a bit more for the next release. The others seem to be fine.

oh, no quit message?

frg: potentially

IanN have a bunch of personal stuff scheduled this week but will see that it goes forward.

IanN: don't know where is the error console. stdout for today (preceded by "date" output) is as follows:

Sun 10 Nov 04:22:37 CET 2024

/root/.mozilla/seamonkey/r0hh40it.frg-2018+/user.js:21: prefs parse error: expected ',' after pref name

/root/.mozilla/seamonkey/r0hh40it.frg-2018+/user.js:152: prefs parse error: unknown keyword

info: 0, warning: 0, error: 0

Done

info: 0, warning: 0, error: 0

Done

info: 0, warning: 0, error: 0

Done

info: 0, warning: 0, error: 0

Done

info: 0, warning: 0, error: 0

Done

oh god

tonymec: uh. 20241106122222 shows balls and no bold to me; which theme are you using? (Bold could also be just font size, or different font); sizing is different but I guess that's what was discussed before

tonymec: do you use any special themes for cZ or cZ plugins?

oh wow I was scrolled up

frg: I need to test the new builder this week

No user.js in 2.53.20b1 pre. Probably a different add-on needs a fix.

IanN I won't manage to use 2019 for the next build but 2012 R2 is still fine for 2.53. Sorry. Too much real world stuff.

frg: what puzzles me is the "frg" there, was that just nick autocomplete?

IanN: no cZ plugins; I always used theme chatzilla.hacksrus.com/motifs/output-light-faces.css

tonymec: or is your profile named "frg-2018+"?

njsg: yes

tonymec the theme probbaly needs an update, User tree switched from tree to listbox.

frg: ah

tonymec: try with chrome://chatzilla/skin/output-marble.css or chrome://chatzilla/skin/output-light.css and see if it works

Extensions Tracking

bau

Maybe a hint. The places updates might affect old extensions using the old sync apis. I doubt there is one left still working after switching already a few releases back but you never know.

wb tonymec|away

2.Next, Feature List, Planning & Roundtable

bau

frg sounds like a broken record .. 8-)

IanN: user list didn't change. No faces (that's to be expected) in the main log.

rsx11m yeah. Pretty much the same lately. Backporting and support and now and then some fixups.

no news is sometimes good news ;-)

tonymec: so it's working now, or still no mode symbols / coloured balls?

IanN: still no mode symbols or balls, still all bold, and still (forgot to say) sorted by mode

which SeaMonkey theme? or did you say that already?

tonymec: when you get chance, can you test with a new profile

modes do show up with modern, but tabs other than the active one don't extend to the full height of the tab bar

njsg_: theme MicroMonkey [converted] 2.0.27

oof, missed the meeting?

tomman: almost done

tonymec: I see mentions of user-list and treechildren in chatzilla.css for micromonkey, so it might be that

can we conclude AOB first and you continue debugging after the meeting?

ah Chatzilla

hm, after the meeting finishes I'll try with Sm buitin themes (but not Safe Mode bcz that means no cZ).

worth reminding the weird regression on the user list

where every time someone joins or parts a channel, the nick list on the current tab gets populated with ALL THE NICKS from ALL THE CHANNELS :D

rsx11m: yes, indeed

tomman: tried with a recent build (from the last 4 or 5 days)?

AOB

IanN: this build is 20241030113351, so I may try a newer one later

broken record bau

so next meeting is in December

soon be Xmas!

next meeting would be December 1st

tomman was a problem in older builds.

that's the Thanksgiving weekend here

fun, December starts on a Sunday

which means the year ends... in mid-week

that kinda sucks :D

okay, thanks for your time today, next meeting in 3 weeks time, same bat channel, same bat time

ok, I can post the agenda and invitation but won't be there

we will decide at next meeting when the following meeting is, as I am away 22nd and 29th

so we keep it on the Thanksgiving weekend?

or shall we move by a week? (Nov 24 or Dec 8)

we could do Dec 8 and then Jan 5

ok with me

or Nov 24, Dec 15 and Jan 5

i'm easy either way

not bad either

me too

someone has a dice?

heads and tails

how's the mozilla infrastructure situation, any reason that'd need more meetings? OTOH that can just be discussed outside of a meeting, yes

[coin lands on its edge]

njsg_ no change for now.

Nov 24 is probably to early to be discussing a release, whereas Dec 8 would be better

Lets do 08/12 and 01/05. Have a bunch of personal stuff for the next weeks.

sounds good to me

okay 8th and 5th

Can always attend but ias stated mostly bau lately.

so next meeting is in 4 weeks time

I've got no strong feelings one way or the otherTM, so ok with that one

user list is OK with both builtin themes

bye all!

bye and here

cya all

At least with modern some users at the end of the list have a "box" around them

uh, seems to be mouse selection related somehow

but it only affects the last few at the end, and only if I click on one there

njsg_: strange

indeed, more likely to be somehow a local issue, although it's puzzling that it affects only one end of the list and seems to be reproducible

njsg might be gfx driver related

IanN: so the culprit is my _SeaMonkey_ theme. No problem with my usual cZ theme and either of the Sm builtin themes. Sorry.

tonymec|away: not a problem, just need to fix the theme

modern tab height issue present in 20241010 2100 too

IanN: huhuh. "By Alfred Kayser, last updated April 6, 2017, No updates found."

tonymec|away: :(

njsg_: in cZ?

does it have a repository or site somewhere? I didn't see a link in the addons site

IanN: yes; maybe that's how the tabs are supposed to look, but I'd say it looks like the height needs to be extended, at least for what I'm used to see as the rendering of a tab bar

no idea for how long this has been like it is now, sadly I don't test Chatzilla often

njsg: dunno

njsg: it says "Homepage: addons.mozilla.org/en-US/seamonkey/addons/309/developers ". Obsolete, I suppose, but maybe ported to the Thunderbird addons site ?

njsg_: do the active tab looks lots bigger than the inactive tabs?

*does

Hm, maybe the idea *is* to make them enlarge on hover

the active tab has a background which looks bigger in height than that of the other tabs

njsg_: definitely been like that for a while

njsg_: I think it could be made to look better

playing a bit with the CSS in the DOM inspector, it looks like the color bar that indicates activity might be part of why it is like thi, that is, the height might be smaller to account for it

s@thi@&s@

njsg_: I think the main issue is the background colour for the inactive tabs

njsg_: IanN2: I see the cZ active tab maybe a millimetre or so higher than the other tabs. Same level at bottom. Intended I suppose. For the bgcolor I have user CSS.

tonymec|away: check with modern

njsg: with Modern the difference (top _and_ bottom) is a little bigger, with a bigger font. Maybe 2 mm difference either way.

tonymec|away: ah, I see, I should try also with the "default" resolution (dpi) to see if scaling is making it worse

uh. this time it put all as the quit message, I was wondering if it needed a network name as argument somehow

njsg: I see I have a bunch of "ChatZilla" lines in userChrome.css

I am able to build 102 but 91 needs a patch well one so far

yeah anything older than 102 likely needs alt deps cause artifacts are unavailable largely cbindgen and lower clang i'd expect

guess i better get good at compiling clang and rust

102 on linux el8-ish builds with 115 artifacts i picked rust 1.65 as a random guess so ye

so I am on a fool's errand.. try and get dom/xbl code to build.. not hooked up.. but build

on esr102

well that was fun never the less and informative..