« first day (1209 days earlier)      last day (4107 days later) » 

19:00
stick a USB hub in the USB port, then go: USB CD/DVD and USB HDD, boot and dd
@allquixotic Oh I thought security risk mainly...
This'll make a good QOTW post when it's all figured out
hopefully it has drivers. One of the reasons I cannot use it is because it is missing a bunch of drivers / has services installed that require passwords to stop
HAMMERTIME!
@rlemon there should be enough drivers on a modern Linux to get it to a terminal... you do have a keyboard with it, right?
19:01
nope
I mean, I have keyboards. none are bt
@rlemon ack... that's a problem
@allquixotic You can connect USB CD/DVD to tablet? OTG?
well how about plugging a USB keyboard into the USB hub you connect to the tablet?
all you need is a 3-port USB hub and a USB host port on the tablet
19:02
this is the tablet.
btw
brb need to find normal -> micro usb converter
does it have a USB host port? if not, you'll need a USB OTG cable
this could quickly get expensive
USB OTG -> powered USB hub -> { keyboard, CD/DVD, HDD }
hackish, but it'll work
depends on how much of that you'll have to buy
@nhinkle the thought processes that we've been employing to narrow down the issue bit by bit, and then gather data as you've done on google docs, is useful for all sorts of different problem domains
@Boris_yo I see a mayor problem if some user use it as user image...
showing this stuff "in use" with a real problem is valuable to someone who isn't familiar with it all
Hi I would like a secon opinion on this before I open a new question : superuser.com/questions/681323/…
any insights on what might be causing this Very frustrating issue
@Braiam Mesa 9.1.x is rather recent compared to my ancient driver from ~2010; aside from that, R300 is extremely well-maintained and low-bug by now
also, worth mentioning that someone in the comments on the SU Q about the FF bug said that he can't reproduce the problem
would love to get the specifics on that one and correlate
19:07
no OTG, do have powered hubs. need to connector converters :/
and here he's!
@MariusMatutiae o/
Let me explain a little more. I have a not so bad connection (20mb/10mb) and my neighbourd has the same conenction from a different provider. He can log into a server withouth any lag but I get 1500 of ping... my network is empty. Router just been fully restored. Modem is reset.
the server is not far from my home but I still get an awfull ping. I play all types of game and my ping is always around 20-30... maybe 50 but never 1500
I would blame different routes used by both ISP's...
When I ping the server from CMD I get a 20ms response
and/or network/protocol throttling
19:10
AND the trace routes form me and my friend from the server ( also the opposite server to us) are pretty much the same... 12 hops or so every time
from*
how could I check the Network/protocol throttling ?
How to shoot yourself in the foot. A) Provide array of crasy DRM garbage that messes with every file. B) be evasive about why it is there and how C) have it act as much like a virus as possilble D) have everyone wanting to turn it off, or delete it E) now try to sell them your music and videos When they cant Watch them anymore. F) the grade you get for your insane protection scheme.
@nhinkle I updated the question; hope I'm not out of line
3
Q: Visiting a site turns the system unresponsive

That Brazilian Guy Important: Several SuperUser regulars are currently trying to narrow down the scope and cause of this issue. We are currently seeking willing volunteers to fill out the form here after performing the following test: Install or start a recent version of Firefox (say, 24.x or 25.x; beta would ...

@MariusMatutiae can you please post your about:support in Firefox (click the "Copy text to clipboard" button) to gist.github.com or paste it straight into chat? thanks....
http://superuser.com/questions/681323/can-a-software-cause-a-huge-ping-response/681327?noredirect=1#comment863259_681327

For more information on the test I made already
This is so frustrating I have no idea what I can do from here... also the ISP I use is a reseller of my friends ISP so it's pretty much the same server
@phadaphunk Did you run pathping or at least traceroute for your and your neighbors connections yet?
# lspci -vnn | grep -A 12 VGA
00:02.0 VGA compatible controller [0300]: Intel Corporation 3rd Gen Core processor Graphics Controller [8086:0166] (rev 09) (prog-if 00 [VGA controller])
Subsystem: Toshiba America Info Systems Device [1179:0003]
Flags: bus master, fast devsel, latency 0, IRQ 45
Memory at e0000000 (64-bit, non-prefetchable) [size=4M]
Memory at d0000000 (64-bit, prefetchable) [size=256M]
I/O ports at 3000 [size=64]
Expansion ROM at <unassigned> [disabled]
Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
19:15
I'm trying that as we speak over teamviewer ill post the resulsts in a few seconds
Alternative to shooting yourself in the foot: A) Provide Player that plays DRM files, as its purpose in life. B) have the player itself have the proper de-coding methods and licencing (not the whole computer system) C) provide player for free D) charge for the media F) Tell people that is its purpose in life G) enjoy profits
@phadaphunk if the ping within Minecraft is sending a packet that is read and decoded by Minecraft's main packet-processing thread(s), and it's high, that could indicate that the server that's hosting the Minecraft is experiencing significant load, causing queues to fill up for processing commands, eventually leading to delay in processing packets.
I use chromium
So SSDs consist of small emmc's?
@MariusMatutiae OK, but we are specifically troubleshooting a problem in Mozilla Firefox; it is well-known that Chrome/Chromium are not symptomatic. In order for you to provide us the info we need, please try Firefox or Iceweasel or similar.
Something based on the "Gecko" rendering engine.
19:17
@allquixotic We are hosting the server on a linux CloudAtCost using McMyAdmin as a manager
@allquixotic We are two users on the server and even when I'm on my own I get a 1500 ping
@phadaphunk do any other users viewing the ping through the minecraft client see the ping at a more normal level?
@allquixotic Yes they do (my neighbourg sees it aroun 30)
@phadaphunk and that "30" is not from the ping command on cmd.exe, but from minecraft itself, right? just making sure
@OliverSalzburg 11 Hops with the pathping .. still computing statistics for 275 seconds
@phadaphunk Yeah, let's see what the RTTs are after that
19:19
@allquixotic Yes it is. It's always a little higher there but still I get 1500 and feel like 1500 in game (when I can manage to log in)
@Psycogeek 1.34v is a lot for Sanyo NiMH 2500mAh?
it's unlikely, but not impossible, that your ISP might be traffic shaping you; meaning that it might detect the specific pattern or style or protocol of traffic and decide to give it very low priority
most ISPs already do this for the Bittorrent protocol, but I haven't heard of anything specifically against Minecraft.
@allquixotic Didnt know that was possible.. I can check that out by contacting them and ask why I get a slow connection when accessing this specific server ?
@phadaphunk you could. my guess is that they will be as unhelpful as humanly possible if you contact them
you might even wonder if they're human when you get through with them
@phadaphunk Other Minecraft servers work fine for you?
19:21
@Boris_yo 1.34V is low for charging, normal for being charged. in fact that is a very good post charge level, before it gets any discharging.
@OliverSalzburg Yes they do. Haven't tried since I got that issue yesterday night though. I would have to find a server to test because it's the only one I know that is up right now.
@OliverSalzburg Wow just got the results.. what do you want from that ? theres various info
@phadaphunk everything >:D
@allquixotic which question?
@phadaphunk Just paste the whole thing on gist.github.com
@OliverSalzburg I have Various % for every hops lol
19:23
@phadaphunk The RTT is what's relevant I guess
But if your ICMP ping in general is fine, I guess there won't be any major surprises :P
There it is
@OliverSalzburg Paste in the link. I have no ideas if it's good or not
Would reinstalling an OS magically fix my "INF file not found" error?
@nhinkle the main Q
3
Q: Visiting a site turns the system unresponsive

That Brazilian Guy Important: Several SuperUser regulars are currently trying to narrow down the scope and cause of this issue. We are currently seeking willing volunteers to fill out the form here after performing the following test: Install or start a recent version of Firefox (say, 24.x or 25.x; beta would ...

i put a bit of a header at the top of the Q
I've been trying to fix this problem for two weeks and nothing that I've found online has been working.
@Yuki depends on what's saying that; if you install the same programs in the same way on the reinstalled OS, you might encounter it again
19:28
@allquixotic I'm trying to install the driver for my new graphics card.
LOST PACKETS, LOST PACKETS, LOST PACKETS, LOST PACKETS, LOST PACKETS, GOD SAVE HIM!
1
Q: Unable to install - INF file not found

Jason LI just got a new graphics card, AMD Radeon HD 7790, and I'm having trouble installing the driver. I uninstalled my all of my old graphics drivers and used Driver Sweeper. I then downloaded the appropriate driver from the AMD site and attempted to install it, but I ran into errors when the install...

welp. now just need to track down some USB-B -> MINI5A converters or female to female USB-A connector.
@Braiam To many losses ?
@allquixotic Got the bug on Kubuntu and Arch, with Firefox: memory goes from 2.3GiB to 4.8GiB, CPU1 at 100%, no network traffic.
19:29
@MariusMatutiae fill out the form too please docs.google.com/forms/d/…
@Yuki good troubleshooting steps taken so far -- given that information, I would say that, most probably, a clean install of Windows would not produce that error message
I'd love to be able to tell you how to fix your existing Windows, but a reinstall sounds very prudent if you can't figure it out
ohh and I filled out the survey ... but what I saw was (win7, FF25) -> locked, locked, locked, asking me to wait or quit (I waited), back to normal
locks for ~5-10 seconds
@allquixotic All right then. I guess I'll back-up my files and do a clean install.
@MariusMatutiae I had figured that you would be able to reproduce it once you switched to Firefox :-) thanks for trying it!
@rlemon when you reopen the window (after it's been minimized) does it freeze all over again?
19:30
@allquixotic heh, fixing things the MSTF way?
@nhinkle didn't try.
why do I loose so many packets...
but I was able to close it without killing FF / other tabs
@phadaphunk anything over 1% is too many
@rlemon did you check task manager? did memory usage shoot up or oscillate at all?
19:31
yea, spiked for a second
@phadaphunk I can run a comparison on my end, but it'll take a while
same with cpu
the "I'm totally not affected by this at all" behavior is: you open the tab, load the link, it loads within less than a second, and CPU/RAM usage remains more or less constant (it might increase by say 50-100 MB, and a very short-lived CPU spike, but nothing like 5-10 seconds)
if you don't have that exact pattern then you're impacted to some extent
@Braiam How could I fix that ? Is it the router ? THe ISP ?
What is causing so many losses
I'm not saying i'm not affected. I'm just saying it wasn't as bad for me
19:33
@rlemon understood
work pc. so intel onboard graphics.
@OliverSalzburg Not necessaRY just ran the test on my friends setup and there
is 0% loss everywhere
@phadaphunk ISP...
not everyone is experiencing a full-on browser crash or OOM
but I do have a good cpu / lots of ram
19:33
the OOM'ers seem to be limited to GNU/Linux
Windows is impacted in the sense that RAM usage is way higher than it should be, and the browser temporarily freezes
and on Mac it appears to be as bad as on GNU/Linux
when I said that I OOMed the kernel didn't started the OOM kill but I ;)
@phadaphunk Yeah, same from the box I just tested on
mtr is so much cooler than pathping :P
@phadaphunk Given that you have packet loss on every hop, it's either a problem with your networking equipment or your ISP
mine just gave up D:
> Too many hops: pmtu 144
If you have the option, you could tether to your mobile phone and see what that gives you
@allquixotic sorry about the misunderstanding, guys
19:38
@OliverSalzburg Alright thanks guys :(
@Psycogeek I don't know what it should be for charging but what I specified is post-charge.
@MariusMatutiae it's ok dude, your datapoint helps
!!should I go for a smoke before I nuke my coffee my bot man?
@rlemon No way
ok, hot coffee whilst I smoke!
19:42
@rlemon you'll find that Cavil is somewhat less cooperative than Caprica Six
he likes "No" as an answer in most situations
I think it's behavioral conditioning from all the times we've made him post grumpy cat and grumpy fox saying no
Interestingly @allquixotic the simpler image makes FF really laggy but it doesn't completely crash.
(!!no and !!foxno)
isn't the "random" 50/50?
!!foxno
19:43
!!yes
wow that's creepy.
@OliverSalzburg Once I had packet loss of 100% in one of starting hops which I measured with Pong Plotter but I was still able to send and receive data from server.
!!foxyes
@nhinkle That didn't make much sense. Use the !!/help command to learn more.
19:44
@nhinkle but it does crash on the original?
@allquixotic not a full crash, but uber-freeze
I had to force-kill it to recover.
!!halp
@nhinkle That didn't make much sense. Maybe you meant: help
@Boris_yo Some servers simply don't respond to ICMP PING requests
We need to teach it halp
@nhinkle that's very interesting. perhaps the memory usage is exponential or something, relative to the number of "bad" directives (whichever directive that may be)
19:44
more accuraltely low TTL
we might be able to engineer an artificially terrible image by copy and pasting the offending directives several times in an image, in order to "magnify" the problem and crash even more of the systems, to make it all the more obvious who's impacted
@Braiam ohh it is not '50:50
I'm pretty sure that even if we did that, I wouldn't crash, but most of the people who merely get a 5-10 second lag would then crash
or just hang indefinitely
@terdon? :P
@OliverSalzburg So there is 2 types of pings? One is ICMP and another is data ping?
@OliverSalzburg definitively mtr is more cool for traceroutes D:
19:48
@Boris_yo There are an indefinite amount of pings. It's just a concept
the ICMP ping is the by-default ping test of the internets... you don't support ICMP you are not in the internetz
@OliverSalzburg How do you distinguish between packet loss of data and packet loss of ping?
@Braiam well, no... GNU/Linux's ping command defaults to UDP
@allquixotic you kidding me...
 ping, ping6 - send ICMP ECHO_REQUEST to network hosts
HA!
@Braiam eh, I don't remember which platform defaults to udp... either linux or windows does
19:55
in Ask Ubuntu General Room, yesterday, by Braiam
I BET YOU ARE USING WINDOWS!
@Braiam im using windows ;)
@Boris_yo think of a real-world example: you have a sports stadium with just one entrance gate, where they take your ticket and give you the ticket stub... only one person can enter at a time... but behind that first gate are a bunch of second gates, in a wide line, and the person who takes your ticket can take as much time as they specifically need... the fast guy just lets you right on through; the slow guy stares at your ticket, analyzes it, bites it, etc
the first gate is the hardware level -- the time it takes to get a packet - any packet - to the remote host; the second set of gates is the application layer; one application might have no queue and thus a very fast response time, while another application might have a delay
ICMP is a second-level gate where the guy barely even glances at your ticket; he just tears off the bottom half and gives it right back to you as fast as he can
Minecraft would be a second-level gate where the guy dips your ticket in paint and waits for the paint to dry before giving it back
20:14
can someone with the FF bug please try this? (source)
ideally someone for whom the original image doesn't outright OOM you
I'm out....
@Braiam out?
you can try it anyway :P it might just crash you worse
no nice value saves me! D:
nice firefox

(process:7959): GLib-CRITICAL **: g_slice_set_config: assertion `sys_page_size == 0' failed
Terminated
firefox killed itself...
@allquixotic ^
@Braiam nice :)
Bob
Bob
20:44
huh
@allquixotic The VServer guest downs't seem to have a shutdown
anyway, I've stopped the terraria server for now
Hi all, first time here. My question superuser.com/questions/681439/building-pc-for-programming is on hold is this room where somebody can help me?
@1110 probably
fairly low activity at the moment in here but will try
@Bob correct -- no shutdown -- you have to use vserver stop on the host (vserver --help)
Bob
Bob
@allquixotic :\
also @Bob read the scrollback, the star wall, etc
amazing stuff today
Bob
Bob
@allquixotic Uh... I'll do it later. Leaving for work now.
20:51
Ok
what does the fox say? :D
Bob
Bob
...I'm not even going to bother.
:D
I'm not allowed to use !!foxno
only the fox can
@1110 Z87 motherboard and Haswell i3 looks to be about right
see if you can get a quad core though, that'd be my only suggestion for programming
So you think that amd would be waste of time?
@1110 AMD APUs are really for people with a razor-thin budget who want a decent gaming PC -- the CPU throughput and parallelism is way behind Intel
right now AMD is struggling to compete with Intel's 3 generations old Sandy Bridge architecture on the CPU front
against Haswell or Ivy Bridge there's no contest; Intel wins
the only thing AMD APUs have going for them is very good, affordable gaming GPU
here's a really nice Haswell: ark.intel.com/products/75047/…
costlier, but it's still not an i7
a good $100 cheaper than the top
and what about mother board I saw this one asrock.com/mb/Intel/Z87M%20Pro4/?cat=Specifications
21:00
@1110 pretty much any Z87 mobo is going to suffice
just make sure it has the ports/connectors you're interested in using to connect your devices
then get whatever's cheapest I guess, or maybe not if you care about build quality etc
ok thank you for the help :) Maybe I will start with i3 as I think it's much faster than my 775 D805 :D
@1110 well, whatever; it's your choice; but you said 5-6 years... in 5-6 years a Haswell i3 will be as old and slow as a Pentium 4 today
usually if you're going to upgrade very infrequently like that, you want to spare no expense and get the best stuff
high-end lasts longer
I didn't thought on that :) then I will get that other cpu you suggest me, and find cheaper motherboard, motherboard is actualy my main concern
@1110 there's not much to motherboards these days -- especially if you're not getting a dedicated GPU
you plug the RAM and CPU in, you plug the power in, you plug your display into the motherboard's HDMI or DisplayPort port, and that's it
it would matter more if you depended on the motherboard for things like WiFi, Bluetooth, PCI Express slots for RAID controller and GPU, etc
I depend only on pci express :)
Thanks for you help I am now more confident to not go to amd :)
21:11
@1110 for your use case I strongly recommend against AMD
@Bob @OliverSalzburg this is awesome O_O
@allquixotic that looks like a koding clone
can anyone flag this as spam?
21:29
@allquixotic If I change IP to VPN on the fly when being on any webmail service and send email to someone, what IP will recipient see?
@Boris_yo email recipients generally aren't able to see the IP address of the web mail sender... they can get the IP address of the SMTP server that sent the email, but that's about it
@allquixotic But software will be less demanding. I am talking MS Office, internet surfing, email checking and all office stuff. But video games is different story.
@allquixotic Strange, I am sure when I sent email to myself, I saw my IP. I will check this again with sending from yahoo mail to outlook mail.
@Boris_yo he's a programmer
needs plenty of ram
@allquixotic To run VMs?
21:42
@allquixotic This is interesting. Apparently compared to Outlook and Gmail, Yahoo discloses senders IP.
This is the line:
Received: from [MY IP] by web126102.mail.ne1.yahoo.com via HTTP; Wed, 27 Nov 2013 13:35:36 PST
@Boris_yo this is something special about yahoo
it's not required by the email system at all
@allquixotic You knew this?
Bob
Bob
reading that conversation. the weird characters in the svg filename are a zero width space... In the file extension o.O
@Bob turns out that didn't make a difference
the file with the weird characters isn't even the one that ended up having the problem
21:46
@allquixotic Interesting. All messages are shown in Sent folder that I sent today, except one which is when I sent from VPN. Ha
22:06
Hi guys!
hi @Overflowh
@allquixotic Good afternoon!
There's someone that could help me analizing S.M.A.R.T. data from a disk? Finally I got a "new" one!
@Overflowh sorry, a little tapped out from the saga earlier today ;p
Bob
Bob
22:22
> The SI TSS v1.2.1.41 TCS service depends on the TPM Base Services service which failed to start because of the following error:
The operation completed successfully.
> The WvPCR service depends on the TPM Base Services service which failed to start because of the following error:
The operation completed successfully.
@allquixotic Dispite I haven't really understood what you wrote, don't worry :P it's not so compelling
22:37
@Bob I fail to do things because I did them successfully all the time
this seems perfectly normal
@JourneymanGeek you missed The Saga, too D:
the star wall pretty much says it all
Bob
Bob
finally figured out the intermittent hangs on my work machine
well, I say 'figured out', but this is practically spoon feeding :P
@allquixotic: yeah, I was there for the start of it ;p
@Bob any Firefoxen to test the resource-flooding SVG file?
Bob
Bob
22:44
@allquixotic tested at home, nicely oscillating memory usage (1.4 GB => 2.6 GB => 1.4 GB etc)
@Bob yuck -- so you're basically in the same boat.
Bob
Bob
gota firmware update to do here first, then I might test it
also, GTX 560 Ti, so should be more than capable... unless driver bug
would be interesting to see what happens with the even more pathological image i posted a while ago
it's basically a copy and paste of the worst offending lines from the original SVG, but with the values and ids slightly tweaked, to make the SVG renderer do it oh so slightly differently each time
so assuming it completes one iteration, it has 19 more to do
@allquixotic you darn terrorist... using pathological weapons ¬_¬
3 hours ago, by allquixotic
can someone with the FF bug please try this? (source)
22:47
btw, there was a CVE about SVN files the other day...
naturally it completes in less than 1 second on my system :D
@Bob if you are worried about losing your tabs or other profile data, starting FF in safe mode should work
it almost certainly will crash
> CVE and NVD are both sponsored by the National Cyber Security Division of the U.S. Department of Homeland Security.
fear!
Bob
Bob
@allquixotic Again the varying RAM usage.
It never rendered (white screen). I just closed the window after a while.
o.O
It's still just a white page, but Aurora finished almost instantly
@Bob that specific SVG never renders anything, ever
Bob
Bob
wait no, it's spiking again
22:56
it would be spiking pretty badly but that's just from it rendering, effectively, nothing
when you are actively looking at the tab, it's probably eating 100% CPU
Bob
Bob
@allquixotic Aurora takes about 5 secs
when you tab away, the lag disappears
Bob
Bob
and rerenders when i switch to it
if I just sit there, it stabilises
it seems to render every single frame separately
Bob
Bob
memory goes from 150MB to 600MB and then back down
22:58
it stabilizes? hmm. hinkle was observing earlier that it was specifically unstable
I think he refers stable as "follow a pattern"
Bob
Bob
@allquixotic I'm on 27.0a2 2013-11-06
it's a bit old - updating now, actually
Bob
Bob
@allquixotic Is that Aurora?
see the spikiness? whenever he's switched to the tab with the offending SVG, firefox continuously allocates and deallocates around a gigabyte of RAM, fluctuating up and down
@Bob no, most of our testing today was with FF 25.0.1 stable
Bob
Bob
23:00
@allquixotic Yea, that's what I get in FF25
I'm saying Aurora stabilises.
maybe FF 27 has stopped the continuous re-rendering of SVGs, which would be nice
Bob
Bob
It's effectively a different browser, if you're looking at rendering bugs.
at least that way once it successfully renders it once, it'll stop doing it over and over again, which would improve performance for particularly slow SVG draws
this one's in a category of its own, though, because very few SVGs are constructed with such a concise and potent markup that throws between 1 and 4 GB of junk on the heap
Bob
Bob
@allquixotic Again, only 150MB => 600MB => 150MB
Bob
Bob
23:02
i.e. Aurora is far better than FF stable for this
so the total memory usage is also less in Aurora
ok
Bob
Bob
I wonder if Beta is like this too
if the other people who were having the problem also test nightly or aurora and find it to be much better, Mozilla might end up just closing our bug as "fixed in master" and accept FF 24 / 25 as a casualty of war instead of patching it
well they'll pretty much have to fix FF 24, because it's ESR, but FF 25 can go to hell as far as they're concerned
especially if this is completely fixed in FF 26
anyone knows where is the changelog in this page mozilla.org/en-US/firefox/27.0a2/auroranotes?
Bob
Bob
I might test when I get home.
> SVG: Added support for transform-origin (see 923193)
hm
23:05
adding support for an SVG directive that we don't use won't, on its own, matter
Bob
Bob
cool, Elm is finally moving out of Nightly
unless they changed some stuff around to implement transform-origin that impacts other code
Bob
Bob
@allquixotic Yea, that just happens to be the only SVG-related change on that list
the bad SVG uses regular transform
but not transform-origin
Bob
Bob
> Update SVG-in-OpenType implementation to new spec draft
That's fixed in 26.0
> Figure out if SVGElement.ownerSVGElement should throw
23:07
our root element is svg
this guy looks like he might be able to help us
> Major SVG rendering improvements around Image tiling and scaling (see 600207)
nah, that's not... but seems like maybe tangentially touch our bug...
@Braiam the activity on that is extremely old
heck was seeing 24 beta D:
Bob
Bob
note to self: sort by last changed
NO NOT ASCENDING
...2004 bug reports
> Bug 936136 - Sluggish scrolling when page has several hundred SVG elements
o.O
> We have been working on an internal proof-of-concept dashboard for Engineering. The page renders several hundred SVG elements on screen. The performance is satisfactory in release (25.0) but demonstrates sluggishness in Nightly (28).
apparently they aren't using the same SVG elements as we are
Bob
Bob
23:18
hey, apparently that was fixed by the introduction of transform-origin
the important thing to remember is that it's not SVG itself which is painfully slow on FF 25 (or any version), but the specific directives in these particular SVG files
I'm leaving work for four days !!!! yesssssss! I'll be back here in about an hour, but I won't be back at work until Monday :d
brb
Bob
Bob
@allquixotic what's that link got to do with it?
O.O
@Bob idk, probably nothing
Bob
Bob
> Bug 817993 - SVG Repainting twice with attribute transform is changed
> Bug 839865 - Stop calling nsSVGUtils::InvalidateBounds for SVG transform changes, and use DLBI instead
the second one has a comment saying it might fix the first
fascinating
Bob
Bob
23:22
> This made a huge difference to bug 884787.
> Bug 884787 - kieranhealy.org/philcites is very slow to load on Linux, and crashes Fennec
@allquixotic I think we have it!
...or not
are test case available?
Bob
Bob
> For me, the fix for bug 884787 makes a huge difference here. The patch for that bug landed in v23 though, and you've filed this against v24, Sam. Was that a mistake, or do you really not see perf that is close to that of Chrome in Nightly/24/23?
> Bug 875175 - Stop continuous reflow and invalidation of entire SVG when an element that is partially/wholly outside the viewport is subject to a transform animation
that... might be related
well, we'll see how it goes on my systems at home when i get home
brb
Bob
Bob
lots of bugs recently for moving from gfxContext to Moz2D
there was actually a performance regression: bugzilla.mozilla.org/show_bug.cgi?id=846181
> There are potentially large performance wins for dynamic SVG here, plus it will help us unify SVG effects (including transforms etc) with non-SVG.
which is also listed as a possible fix for bugzilla.mozilla.org/show_bug.cgi?id=612250
another performance enhancement: bugzilla.mozilla.org/show_bug.cgi?id=869611
and that's it
23:38
@Bob Platform: ARM Android
Bob
Bob
@Braiam Which one? (can't see message replies on mobile)
13 mins ago, by Bob
another OOM crash: https://bugzilla.mozilla.org/show_bug.cgi?id=693484
Bob
Bob
23:55
@Braiam Ah. The listed possible fix should apply to all platforms anyway.

« first day (1209 days earlier)      last day (4107 days later) »