[01:39] <sarnold> hello; does launchpad do anything to influence how pageup and pagedown function? when viewing a long bug report, pageup always jumps to the very top of the bug report and then appears to try to go higher; pagedown will jump to the very top of the bug report and then go downwards. I've got launchpad.net set to trusted in noscript, privacy badger doesn't appear to be blocking anything, and I've got
[01:39] <sarnold> tridactyl installed but it seems pretty well behaved..
[01:39] <sarnold> eg a bug report where pageup and pagedown down work as I expect: https://bugs.launchpad.net/ubuntu/+source/cryptsetup/+bug/1879980
[01:39] <ubot5> Ubuntu bug 1879980 in cryptsetup (Ubuntu Focal) "Fail to boot with LUKS on top of RAID1 if the array is broken/degraded" [Medium,In progress]
[01:39] <sarnold> and a very long page elsewhere that works as I expect: http://sled.rs/perf.html
[01:40] <sarnold> firefox 81.0+build2-0ubuntu0.20.04.1
[01:44] <sarnold> (yes, I could use ^U or ^Y or ^B thanks to tridactyl, but on web pages I'm more accustomed to pageup and pagedown)
[09:56] <cjwatson> sarnold: I've noticed that and it drives me nuts because I cannot work out what's causing it
[09:56] <cjwatson> sarnold: in the past when I've tried to debug it it's gone away
=== zyga_ is now known as zyga
[17:54] <sarnold> cjwatson: argh :) I hate bugs that go away when you try to look at them; it happens often enough for me that I'd even suspect it happens all the time. I'm happy to help debug remotely, such as I can, but it'd really require a lot of hand-holding, I'm very unfamiliar with anything beyond about html 1.1
[21:01] <cjwatson> sarnold: Yeah, I think the main problem was that it went away when I tried to reproduce _with devtools switched on_
[21:01] <cjwatson> But it was a while ago
[21:02] <sarnold> cjwatson: hmm, if that fixes it.. :)
[21:07] <cjwatson> Maybe something to do with https://git.launchpad.net/launchpad/tree/lib/lp/app/templates/base-layout-macros.pt#n150
[21:09] <cjwatson> Or not, doesn't seem to be called
[21:10] <cjwatson> It does feel like a focus bug though
[21:10] <sarnold> focus feels right, half the time I need to click into the page to use pgdown in the first place
[21:10] <cjwatson> Right, exactly
[21:10] <cjwatson> Happens on merge proposals too
[21:11] <cjwatson> At some point I may get cross enough with it to have another go
[21:11] <sarnold> :)
[21:12] <cjwatson> But it may be possible to work it out by seeing if you can get it to reproduce with devtools and setting breakpoints on all the something.focus() calls
[21:13] <sarnold> just skimming this function .. what happens if there's no elements with the name (and thus no focus() calls)? what happens if there's two or more? Should there be a 'break;' after that node.focus()?
[21:14] <sarnold> oh, well, there kind of is. hmm.
[21:14] <cjwatson> Yeah I'm not going to debug it at 10pm esp when I don't think it's called here
[21:14] <sarnold> :D
[21:14] <sarnold> thanks cjwatson, have a good night
[21:15] <cjwatson> :)
[21:55] <tomreyn> have you considered to raise the anti bot measures bar a little? it feels like you must be reverting spam bot edits 50% of your every day at work.
[21:57] <cjwatson> tomreyn: Hardly, I spend a few minutes a day on it
[21:58] <cjwatson> tomreyn: You only see what slips through
[21:58] <tomreyn> to the latter, that's what i was thinking, yes
[21:59] <sarnold> it's bursty, most days I never see anything, and then once in a while I'll see two or three in one day
[21:59] <tomreyn> i'd still get annoyed having to spend some minutes on it every day, but i'm glad you're more relaxed. ;-)
[21:59] <sarnold> compared against email, it's pretty staggeringly good :)
[22:00] <tomreyn> ok, i rest my case then ;)
[22:00] <cjwatson> It's not ideal, but there are always tradeoffs
[22:00] <cjwatson> (I don't think comparing against email is much of a bar though)
[22:01] <tomreyn> let's quickly fix e-mail once and for all, then we can get to launchpad.
[22:01] <sarnold> well, the other comparison that comes to mind is irc, which is also super bursty... some months, nothing, and then one day last week I spent half an hour cleaning up after hundreds of bots...
[22:01] <cjwatson> I find it more interesting to improve API facilities to make it easier to deal with various categories of spam, which is a thing we do from time to time
[22:01] <cjwatson> Usually when one of us gets annoyed at workarounds for lack of same :)
[22:02] <sarnold> hehe