Report site bugs here

Binkyt11

Sleep isn't real.

@Background Pony #EC18
You can edit that under settings. Go to Uploading and uncheck the options that have fancy editing come up first for uploading or editing.

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

Yoshimon1

yay(){ yay|yay & };yay

@Binkyt11
Yeah, about that…

The settings for the fancy tag editor seem to be broken. Every tag editor suddenly defaults to the plain editor for me, even though I’ve never touched those settings and both fancy checkmarks are still set.

Posted 5 years ago Report

Link Quote Reply

Litrojia

@Yoshimon1
Same situation here.

Posted 5 years ago Report

Link Quote Reply

Binkyt11

Sleep isn't real.

@Yoshimon1
@Litrojia
Wait, so that’s a site wide problem? Shit, Firefox has been doing it to me and I thought it was was FF being FF.

Just checked Chrome. Shit.

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

LightningBolt

Senior Moderator

Undead inside

hahaha, get on my level, scrubs, plain tags are better.

I mean, huh.

Posted 5 years ago Report

Link Quote Reply

Princess Luna

Senior Moderator

Site Developer

Tag Czar

@Yoshimon1
Basically, a bug was squashed, but in a manner that destroyed the entire purpose of the code responsible. I’ve put in place a solution that is bugfree and works, if a tad clunky.

Posted 5 years ago Report

Link Quote Reply

Delta_Mk9

catto unicorn

Thanks guys

Posted 5 years ago Report

Link Quote Reply

Meanlucario

Time to get spooky

Nevermind, it’s working again.

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

I miss the show so much

I hate to complain like this. But…

First off, it’s awesome that image pages no longer randomly scroll down when you open them. The tag editor no longer steals focus.

Unfortunately, it does not get focus at all anymore, not even when you open it and you want it to get focus. It used to. To edit tags, you have to click the tag input box (which can be anywhere, so screw your muscle memory), only then you can start typing. This isn’t how it used to work. I tested this with and without the fancy editor and both have the same problem: completely missing keyboard focus unless you click the gray tag box first.

Moments ago, while trying to edit tags in several images, I ended up faving and upvoting and opening their sources, until I noticed my typing isn’t going into the tag box, but the site interprets them as hotkeys. The “L” hotkey still works normally, but it’s useless unless you also reach for the mouse.

Posted 5 years ago Report

Link Quote Reply

stsyn

Moderator

Lastest Chrome (69.0.3497.81 64 bit) crashes when you click on any non full size webm to enlarge it.
Example: >>1483108

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

Wiimeiser

(Foil Hat)

StSyaN
Actually surprised it didn’t crash my whole PC there. Seems to be related to the unnecessary UI overhaul. What happens if you use, say, Javascript to load a WebM in the middle of a page, does the crash happen there, too? How about a Youtube embed?

EDIT: Relevant

chrome has this newer UI for their video-player when watching a video of either mp4 or webm format

EDIT2: Also relevant

Looks like Chrome’s ditched support for WebM…

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

stsyn

Moderator

@Wiimeiser
Not sure what exactly happens here. What I tried:

Disabled every extension
Opened incognito tab with linked webm
After clicking spoiler, large.webm shows (it’s important, I think)
Clicked webm
Goodbye tab
If I have dev console open, tab die too.

Firstly noticed in context of my fullscreen userscript. I made it to try load full version, not thumbnailed ones:
Opened tab with linked image
After clicking spoiler, full.webm shows
Clicked webm
In about 90% cases goodbye tab

If I have dev console open, tab usually not die.

Posted 5 years ago Report

Link Quote Reply

Wiimeiser

(Foil Hat)

StSyaN
Does the console print an error? If so, posting it might be a good idea…

Posted 5 years ago Report

Link Quote Reply

stsyn

Moderator

@Wiimeiser
ADUp crash is not related and occurring at startup
And no scripts

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

Wiimeiser

(Foil Hat)

StSyaN
These lines seem to be relevant:

Synchronous XMLHttpRequest on the main thread is deprecated

Cannot read property ‘appendchild’ of Null

By the looks of things, a feature was deprecated and its coding removed, so the code that runs when loading full size WebM files is calling a nonexistent routine in Chrome. Somehow, it also tries to load the “Shrodinger’s Pony” page as the file instead, but cannot find said page and returns a 404 for the 404 page. At least that’s from my understanding, I don’t know code…

EDIT: The no-script version being empty is concerning, though…

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

stsyn

Moderator

@Wiimeiser
Probably site devs won’t be able to do anything atm. I should somehow prevent source changing on my side. WebM Volume Toggle may fix it as well.

Posted 5 years ago Report

Link Quote Reply

byte[]

Philomena Contributor

@JP
Sorry about that, was a copypasta-style bug

Posted 5 years ago Report

Link Quote Reply

byte[]

Philomena Contributor

StSyaN
Not our bug.

`
(gdb) attach 31546
Attaching to process 31546
[New LWP 31547]
[New LWP 31548]
[New LWP 31549]
[New LWP 31550]
[New LWP 31551]
[New LWP 31552]
[New LWP 31553]
[New LWP 31554]
[New LWP 31555]
[New LWP 31556]
[New LWP 31557]
[New LWP 31558]
[New LWP 31586]
[New LWP 31587]
[New LWP 31588]
[New LWP 31589]
[Thread debugging using libthread_db enabled]
Using host libthread_db library “/lib/x86_64-linux-gnu/libthread_db.so.1”.
pthread_cond_timedwait@@GLIBC_2.3.2 () at ../sysdeps/unix/sysv/linux/x86_64/pthread_cond_timedwait.S:225
225 ../sysdeps/unix/sysv/linux/x86_64/pthread_cond_timedwait.S: No such file or directory.
(gdb) c
Continuing.
[New Thread 0x7f4831b93700 (LWP 31635)]
[New Thread 0x7f4832394700 (LWP 31636)]
[New Thread 0x7f4832b95700 (LWP 31638)]

Thread 1 “chrome” received signal SIGSEGV, Segmentation fault.
0x000056136e4c1280 in blink::MediaControlOverlayPlayButtonElement::IsMouseEventOnInternalButton(blink::MouseEvent const&) const ()
(gdb) bt
#0 0x000056136e4c1280 in blink::MediaControlOverlayPlayButtonElement::IsMouseEventOnInternalButton(blink::MouseEvent const&) const ()
#1 0x000056136e4c1021 in blink::MediaControlOverlayPlayButtonElement::DefaultEventHandler(blink::Event&) ()
#2 0x000056136dbdd0ca in blink::EventDispatcher::DispatchEventPostProcess(blink::Node*, blink::EventDispatchHandlingState*) ()
#3 0x000056136dbdcabb in blink::EventDispatcher::Dispatch() ()
#4 0x000056136dce6166 in blink::MouseEvent::DispatchEvent(blink::EventDispatcher&) ()
#5 0x000056136dbdbf05 in blink::EventDispatcher::DispatchEvent(blink::Node&, blink::Event&) ()
#6 0x000056136df78696 in blink::MouseEventManager::DispatchMouseEvent(blink::EventTarget*, WTF::AtomicString const&, blink::WebMouseEvent const&, WTF::String const&, blink::EventTarget*, bool) ()
#7 0x000056136df78a31 in blink::MouseEventManager::DispatchMouseClickIfNeeded(blink::EventWithHitTestResults<blink::WebMouseEvent> const&, blink::Element&) ()
#8 0x000056136df70d1d in blink::EventHandler::HandleMouseReleaseEvent(blink::WebMouseEvent const&) ()
#9 0x000056136e229ce0 in blink::PageWidgetEventHandler::HandleMouseUp(blink::LocalFrame&, blink::WebMouseEvent const&) ()
#10 0x000056136dcfd172 in non-virtual thunk to blink::WebViewImpl::HandleMouseUp(blink::LocalFrame&, blink::WebMouseEvent const&) ()
#11 0x000056136e229ab6 in blink::PageWidgetDelegate::HandleInputEvent(blink::PageWidgetEventHandler&, blink::WebCoalescedInputEvent const&, blink::LocalFrame*) ()
#12 0x000056136dd019b3 in blink::WebViewImpl::HandleInputEvent(blink::WebCoalescedInputEvent const&) ()
#13 0x000056136eebd02a in content::RenderWidgetInputHandler::HandleInputEvent(blink::WebCoalescedInputEvent const&, ui::LatencyInfo const&, base::OnceCallback<void (content::InputEventAckState, ui::LatencyInfo const&, std::__1::unique_ptr<ui::DidOverscrollParams, std::__1::default_delete<ui::DidOverscrollParams> >, base::Optional<cc::TouchAction>)>) ()
#14 0x000056136eeb0985 in content::RenderWidget::HandleInputEvent(blink::WebCoalescedInputEvent const&, ui::LatencyInfo const&, base::OnceCallback<void (content::InputEventAckState, ui::LatencyInfo const&, std::__1::unique_ptr<ui::DidOverscrollParams, std::__1::default_delete<ui::DidOverscrollParams> >, base::Optional<cc::TouchAction>)>) ()
#15 0x000056136e98473d in content::QueuedWebInputEvent::Dispatch(content::MainThreadEventQueue*) ()
#16 0x000056136e983dc0 in content::MainThreadEventQueue::DispatchEvents() ()
#17 0x000056136b02c019 in base::debug::TaskAnnotator::RunTask(char const*, base::PendingTask*) ()
#18 0x000056136b068be9 in base::sequence_manager::internal::ThreadControllerImpl::DoWork(base::sequence_manager::internal::ThreadControllerImpl::WorkType) ()
#19 0x000056136b02c019 in base::debug::TaskAnnotator::RunTask(char const*, base::PendingTask*) ()
#20 0x000056136b02b3fe in base::MessageLoop::RunTask(base::PendingTask*) ()
#21 0x000056136b02b792 in base::MessageLoop::DoWork() ()
#22 0x000056136b02dd3a in base::MessagePumpDefault::Run(base::MessagePump::Delegate*) ()
#23 0x000056136b048385 in base::RunLoop::Run() ()
#24 0x000056136eee4501 in content::RendererMain(content::MainFunctionParams const&) ()
#25 0x000056136acf3057 in content::RunZygote(content::ContentMainDelegate*) ()
#26 0x000056136acf419d in content::ContentMainRunnerImpl::Run(bool) ()
#27 0x000056136ad2789b in service_manager::Main(service_manager::MainParams const&) ()
#28 0x000056136acf2661 in content::ContentMain(content::ContentMainParams const&) ()
#29 0x000056136907c1b3 in ChromeMain ()
#30 0x00007f48453fa2e1 in __libc_start_main (main=0x56136907c130 <main>, argc=3, argv=0x7ffc50d71d78, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7ffc50d71d68)
at ../csu/libc-start.c:291
#31 0x000056136907c02a in _start ()
`

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply

I miss the show so much

@byte[]
Oh thank goodness. You have no idea how glad I am to see it working again.

Posted 5 years ago Report

Link Quote Reply

Wiimeiser

(Foil Hat)

@byte[]
So basically what was happening was it was receiving null data and interpreting the file as being bigger than it’s allocated memory? Or what? I know that error also pops up when it tries to read an image with a null size and assumes the resulting 0x0 to be 18446744073709551616x18446744073709551616…

Posted 5 years ago Report

Link Quote Reply

byte[]

Philomena Contributor

@Wiimeiser
No.

Dump of assembler code for function \_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE:
   0x0000557f3a81b210 \<+0\>:	push   \%rbp
   0x0000557f3a81b211 \<+1\>:	mov    \%rsp,\%rbp
   0x0000557f3a81b214 \<+4\>:	push   \%r15
   0x0000557f3a81b216 \<+6\>:	push   \%r14
   0x0000557f3a81b218 \<+8\>:	push   \%rbx
   0x0000557f3a81b219 \<+9\>:	push   \%rax
   0x0000557f3a81b21a \<+10\>:	mov    $0x1,\%al
   0x0000557f3a81b21c \<+12\>:	cmpl   $0x0,0xe8\(\%rsi\)
   0x0000557f3a81b223 \<+19\>:	jne    0x557f3a81b36a \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+346\>
   0x0000557f3a81b229 \<+25\>:	mov    \%rsi,\%rbx
   0x0000557f3a81b22c \<+28\>:	mov    \%rdi,\%r15
   0x0000557f3a81b22f \<+31\>:	mov    0x200\(\%rdi\),\%rdi
   0x0000557f3a81b236 \<+38\>:	callq  0x557f39b5d610 \<\_ZN5blink7Element21getBoundingClientRectEv\>
   0x0000557f3a81b23b \<+43\>:	mov    \%rax,\%r14
   0x0000557f3a81b23e \<+46\>:	mov    0x10\(\%r15\),\%ecx
   0x0000557f3a81b242 \<+50\>:	lea    0x38\(\%r15\),\%rax
   0x0000557f3a81b246 \<+54\>:	test   $0x4000000,\%ecx
   0x0000557f3a81b24c \<+60\>:	jne    0x557f3a81b267 \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+87\>
   0x0000557f3a81b24e \<+62\>:	test   $0x1,\%cl
   0x0000557f3a81b251 \<+65\>:	je     0x557f3a81b256 \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+70\>
   0x0000557f3a81b253 \<+67\>:	mov    \(\%rax\),\%rax
   0x0000557f3a81b256 \<+70\>:	mov    \(\%rax\),\%rax
   0x0000557f3a81b259 \<+73\>:	mov    \(\%rax\),\%rax
   0x0000557f3a81b25c \<+76\>:	test   \%rax,\%rax
   0x0000557f3a81b25f \<+79\>:	je     0x557f3a81b278 \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+104\>
   0x0000557f3a81b261 \<+81\>:	mov    0x18\(\%rax\),\%rax
   0x0000557f3a81b265 \<+85\>:	jmp    0x557f3a81b280 \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+112\>
   0x0000557f3a81b267 \<+87\>:	test   $0x1,\%cl
   0x0000557f3a81b26a \<+90\>:	je     0x557f3a81b26f \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+95\>
   0x0000557f3a81b26c \<+92\>:	mov    \(\%rax\),\%rax
   0x0000557f3a81b26f \<+95\>:	mov    \(\%rax\),\%rax
   0x0000557f3a81b272 \<+98\>:	mov    0x8\(\%rax\),\%rax
   0x0000557f3a81b276 \<+102\>:	jmp    0x557f3a81b280 \<\_ZNK5blink36MediaControlOverlayPlayButtonElement28IsMouseEventOnInternalButtonERKNS\_10MouseEventE+112\>
   0x0000557f3a81b278 \<+104\>:	mov    \%r15,\%rdi
   0x0000557f3a81b27b \<+107\>:	callq  0x557f39b62e00 \<\_ZNK5blink7Element28NonLayoutObjectComputedStyleEv\>
=\> 0x0000557f3a81b280 \<+112\>:	mov    0x30\(\%rax\),\%rax
   0x0000557f3a81b284 \<+116\>:	mov    0x10\(\%rax\),\%rax
   0x0000557f3a81b288 \<+120\>:	movss  0x44\(\%rax\),\%xmm0
   0x0000557f3a81b28d \<+125\>:	movss  \%xmm0,-0x1c\(\%rbp\)
   0x0000557f3a81b292 \<+130\>:	mov    0x20\(\%r15\),\%rax
   0x0000557f3a81b296 \<+134\>:	mov    0x10\(\%rax\),\%rax
   0x0000557f3a81b29a \<+138\>:	mov    0x7c0\(\%rax\),\%rdi
   0x0000557f3a81b2a1 \<+145\>:	callq  0x557f3a47f7f0 \<\_ZNK5blink10LayoutView10ZoomFactorEv\>

I don’t have the luxury of full symbols, but the crash here is from a null dereference. If I work back through the assembly, the culprit is most likely this line where ComputedStyleRef() (monomorphized here to blink::Element::NonLayoutObjectComputedStyle, called at +107) is returning a reference to nullptr in rax, which crashes immediately upon use.

Posted 5 years ago Report
Edited 2 years ago

Link Quote Reply

byte[]

Philomena Contributor

@byte[]
Someone filed a bug report here. Thanks.

Posted 5 years ago Report

Link Quote Reply

NamelessNP

Starlight and Sunny Fan

When I go to edit tags in my filters, the hidden tags default to plain text editing. When I switch to fancy editing and try clicking in the box to enter a new tag to be hidden, the cursor jumps up to the box for entering tags to be spoilered. I have to use the keyboard to tab through the entire list of tags to get the cursor in the new hidden tag box.

Adding on this: Tried the mobile version, does the same thing there so I can’t fancy edit filters on my tablet at all, since there’s no tab key to get around this.

Posted 5 years ago Report
Edited 5 years ago

Link Quote Reply