| === bjsnider_ is now known as bjsnider | ||
| nonuby | on 13.04 will there be an seamless upgrade path from b1 to b2 and then to final release? or will it be a pave/format job? | 04:21 |
|---|---|---|
| centrelink | hi | 04:29 |
| centrelink | any unity devs around, I have a quick question | 04:29 |
| === SwedMike is now known as SwedeMike | ||
| zniavre | good morning | 07:37 |
| cwillu_at_work | is there a changelog for 13.04 sitting somewhere? | 07:37 |
| cwillu_at_work | the announce email linked to http://www.ubuntu.com/testing/, but that's still talking about quantal | 07:37 |
| === yofel_ is now known as yofel | ||
| elmargol | Hi I'm wondering if there is a decent nautilus alternative for 13.04 since the version in raring is only a demo version of nautilus used to be | 10:20 |
| bekks | demo version? | 10:20 |
| bekks | elmargol: And you are on 13.04, so you might better ask in #ubuntu+1 | 10:21 |
| smartboyhw | bekks, this is #ubuntu+1 LOL | 10:21 |
| bekks | :D | 10:21 |
| elmargol | lets see if nemo works... | 10:22 |
| * bekks grabs some more coffee | 10:22 | |
| === murthy_ is now known as murthy | ||
| murthy | hello everyone | 10:34 |
| zniavre | good afternoon | 11:29 |
| zniavre | BT icon is not displayed that is known or i got wrong settings ? | 11:29 |
| 52AAAPUHU | zniavre: BT means Bluetooth right ? | 12:11 |
| === snadge_ is now known as snadge | ||
| === murthy is now known as murthy_ | ||
| BluesKaj | Hiyas all | 12:57 |
| alankila | https://bel.fi/alankila/lcd/epiphany-vs-firefox2.png | 13:01 |
| alankila | linux (default) font rendering on right, notice artifacts in the section labeled Browser, corrected rendering on left (perfect). | 13:02 |
| alankila | or well, it depends on what one calls perfect. I think people likely find the rendering too light and I just found out that Skia does this same technique -- I developed it independently, and they also appear to have a hack to make black-on-white text appear heavier | 13:02 |
| alankila | but now I feel really stupid. I mean, I did spend one afternoon going through the formulas and developing this technique, and appear to have arrived at exactly same end result minus that hack to make black text heavier | 13:04 |
| alankila | but I didn't know that Skia did this. | 13:04 |
| rye | alankila: what screen resolution and display size do you have? | 13:10 |
| alankila | 1440x900 I think. It's Apple LCD screen for Macbook Air | 13:12 |
| rye | alankila: so it is 11", right? | 13:13 |
| alankila | a 13" device | 13:13 |
| alankila | iirc dpi worked out to something like 126 | 13:14 |
| rye | alankila: what i am puzzled in your pictures is that the horizontal lines are visibly blurred across adjacent pixel rows, same thing I noticed in the OS X rendering | 13:15 |
| rye | alankila: is it an expected effect? | 13:15 |
| alankila | I disable all hinting in my fontconfig | 13:15 |
| alankila | so yes, it works as requested. | 13:16 |
| alankila | and yes, it is very similar to OS X's approach. | 13:16 |
| alankila | one is not supposed to really look at the individual pixels and demand them to be correct, but rather just appreciate that the average light emitted from each pixel is correct. Still, if desired, it would be possible to turn on vertical grid snapping, aka 'slight' hinting. | 13:16 |
| alankila | this would be similar to AppleFontSmoothing setting 1 | 13:17 |
| alankila | I personally believe that pixel grid should be ignored for rendering, and DPI should be raised sufficiently that the mild blurriness would not be noticeable. | 13:18 |
| rye | alankila: right, re: not look at the individual pixels. In case of 1336x768 14" screen it makes these dark|light pixels extremely visible. On the other hand, it is a low-res display according to the latest screens | 13:19 |
| alankila | because, really, the fully hinted text looks very ugly and causes kerning problems because apparently the font metrics aren't usually adjusted correctly for the results of hinting. Neither is the bytecode hinting the right thing to do because it doesn't work with microsoft fonts because microsoft rasterizer is so different from freetype's. | 13:19 |
| alankila | and the freetype's autohinter has no idea how to preserve glyph symmetry so the results are artifacting. | 13:19 |
| alankila | it's not the resolution but the DPI that is the problem | 13:20 |
| * rye switched to full hinting setting, ugh | 13:20 | |
| rye | alankila: right, but should I have higher resolution at the same screen size I might have looked at it the different way | 13:21 |
| rye | alankila: how does the rendering with gamma correction and added weight for black on white look if hinting is set to slight? | 13:25 |
| alankila | well the horizontal line blurriness is gone, but otherwise it's more or less the same. | 13:26 |
| alankila | because these lines will generally become adjusted to coincide with the pixel grid | 13:26 |
| * rye actually reads the page | 13:28 | |
| rye | alankila: wow | 13:34 |
| alankila | hm, like what you read? | 13:36 |
| rye | alankila: i do, now I understand what it is about | 13:37 |
| alankila | I guess my communication skills are effective enough to describe the problem at sufficient detail | 13:38 |
| alankila | let me say that this is an uphill battle. | 13:39 |
| alankila | it's a pity that alpha correction as I have implemented it -- despite I can see it in practically all applications on my desktop right now -- is basically unshippable in its current form. | 13:39 |
| alankila | additionally the guys at cairo have a plan for implementing it that is somewhat disruptive. | 13:40 |
| alankila | because they want to add SRGB surface into cairo, and then ask applications to construct SRGB surfaces and then some bunch of magic happens which makes it all work. | 13:42 |
| alankila | but there's details -- once you mix multiple colorspaces in, you need to be careful about interpreting say image data read from files correctly in sRGB sense. Because these things are also in sRGB space. | 13:43 |
| alankila | still, stuff like that can be fixed and if you get it wrong it's pretty noticeable because the images will probably appear either too dark or completely washed out, so it's hard to not notice. | 13:44 |
| alankila | also if you read a color from a file, that color is probably 24-bit sRGB color triplet, or 32-bit ARGB value with RGB in sRGB and A linear. | 13:45 |
| alankila | details, details. | 13:45 |
| rye | I can't help thinking that this is somehow related to cairo drawing on a transparent background causing artifacts... such as infamous https://bugs.launchpad.net/ayatana-design/+bug/723167 but I may be completely off the track | 13:45 |
| ubottu | Launchpad bug 723167 in unity (Ubuntu) "Fuzzy fonts caused by Cairo antialiasing artifacts with Radiance theme." [Medium,Triaged] | 13:45 |
| alankila | it's probably result of component-alpha masking | 13:46 |
| alankila | if you end up with a texture, you lose the component-alpha behavior because 3 alphas must be joined into 1. | 13:46 |
| alankila | and then you need to make a theory about what values must be picked for the color components that were actually transparent in component-alpha mode. | 13:46 |
| alankila | oh that's not it. | 13:48 |
| rye | alankila: aha, this explains why drawing on existing background worked... thank you! | 13:48 |
| alankila | I have no idea what the person is talking about. Or it might be it. I don't know, sorry. | 13:48 |
| rye | alankila: your explanation fits perfectly | 13:48 |
| alankila | but yeah anyway component-alpha is required for correct LCD rendering. | 13:48 |
| alankila | it's a stupid name but I guess it's derived from the notion that the alpha channel -- or mask, as it's often called -- is itself ARGB texture, and when in CA mode the A component is ignored. | 13:50 |
| alankila | so the alpha comes from the color components, hence component-alpha. | 13:50 |
| alankila | we'd really benefit from AAARGB textures, I guess. :-/ | 13:51 |
| alankila | and to add to this, there are the physical modeling loonies who think they should be allowed to model the 400-800 nm light intensity spectrum in their textures, because this allows perfect modeling of transmission through multicolored glass and so on | 13:52 |
| BluesKaj | yup. lots of visible UV in that spectrum , not good for you depending on the intensity of course | 14:05 |
| === murthy_ is now known as murthy | ||
| dholbach | https://wiki.ubuntu.com/UbuntuSDKDays Day 2 kicked off half an hour ago - hope to see you all there! | 15:37 |
| fak3r | booting yesterday's daily-live iso amd64-mac on my MBP booted to a grub> command only | 16:03 |
| Sonikk | so, is the 13.04 stable enough for daily use? and is it a nightmare to install Catalyst 13.1 to it? | 16:04 |
| fak3r | what is the diff between amd64 and the amd64-mac builds? drivers for Macs spec or ? | 16:05 |
| Sonikk | fak3r: sounds like that | 16:05 |
| fak3r | Sonikk: so should I try a amd64 bulid on the MBP instead, or can I boot from the grub> command on the amd64-mac? | 16:07 |
| Sonikk | fak3r: im sorry i can not help you. im very new with ubuntu | 16:08 |
| fak3r | Sonikk: np | 16:08 |
| fak3r | Sonikk: I'm planning on using 13.04 as my desk, with release a few weeks ago, I'm sure it'll be fine. Unsure about Catalyst. | 16:09 |
| fak3r | Sonikk: s/few weeks ago/few weeks TO GO/ | 16:10 |
| Sonikk | fak3r: yeah i think also it should be stable. but the catalyst-thing is a big question mark :) | 16:10 |
| Sonikk | i think i will wait until 13.04 is ready | 16:11 |
| djjeff|bt5 | if I do nvidia-xconfig | 17:36 |
| djjeff|bt5 | /usr/bin/X will seg fault | 17:36 |
| djjeff|bt5 | if I do nvidia-xconfig --no-allow-glx-with-composite | 17:37 |
| djjeff|bt5 | /usr/bin/X will start | 17:37 |
| djjeff|bt5 | BUT all my OpenGL applications WONT WORK!!!!! | 17:37 |
| djjeff|bt5 | im sorry but for NVIDIA users 13.04 is FAR from ready | 17:38 |
| bazhang | djjeff|bt5, its not final, you should plan on there being issues | 17:40 |
| === bipolar_ is now known as bipolar | ||
| djjeff|bt5 | so I need to wait for final before I can use OpenGL based apps ? | 17:42 |
| BluesKaj | djjeff|bt5, I beg to differ , dump that xorg.conf file and try again , my level entry nvidia card works fine on OpenGL , DR and 3D | 17:48 |
| djjeff|bt5 | I have done rm -rfv /etc/X11/xorg.conf | 17:50 |
| djjeff|bt5 | several times | 17:50 |
| djjeff|bt5 | I have tried nvidia 304,310,316 | 17:51 |
| djjeff|bt5 | all seg fault when starting X | 17:51 |
| BluesKaj | djjeff|bt5, look for nvidia.xconf | 17:51 |
| djjeff|bt5 | even glxinfo seg faults | 17:51 |
| BluesKaj | what about nouveau ? | 17:52 |
| djjeff|bt5 | I keep seeing this crap http://pastebin.com/raw.php?i=y4rL8Z9h | 17:53 |
| djjeff|bt5 | I have a gdb and backtrace from glxinfo http://pastie.org/pastes/6512486/text | 17:57 |
| djjeff|bt5 | 0x00007ffff7ddb68a in _dl_start_user () from /lib64/ld-linux-x86-64.so.2 | 17:58 |
| djjeff|bt5 | looks fishy | 17:58 |
| djjeff|bt5 | lrwxrwxrwx 1 root root 32 Feb 8 23:25 ld-linux-x86-64.so.2 -> /lib/x86_64-linux-gnu/ld-2.17.so | 17:59 |
| djjeff|bt5 | glxinfo does not seg fault on my laptop running 10.10 | 18:02 |
| djjeff|bt5 | but does on my desktop running Xubuntu 13.04 | 18:02 |
| djjeff|bt5 | how do I set mesa as the OpenGL in xorg.conf | 18:15 |
| djjeff|bt5 | /usr/lib/x86_64-linux-gnu/mesa/libGL.so.1 | 18:16 |
| BluesKaj | djjeff|bt5, xorg.conf isn't required on most setups and the default nouveau driver usually works fine on most apps | 18:50 |
| BluesKaj | including OpenGL requirements | 18:51 |
| djjeff|bt5 | this is really pissing me off now...... | 18:51 |
| djjeff|bt5 | I uninstall nvidia completely from my system | 18:51 |
| djjeff|bt5 | using apt-get purge nvidia-* | 18:51 |
| djjeff|bt5 | rm -rfv /etc/X11/xorg.conf | 18:52 |
| djjeff|bt5 | rebooted | 18:52 |
| djjeff|bt5 | X wont start | 18:52 |
| djjeff|bt5 | because I have no xorg.conf | 18:52 |
| djjeff|bt5 | lucky me I can ssh from my laptop | 18:53 |
| djjeff|bt5 | cause all my desktop has is a black screen with a blinky cursor | 18:53 |
| djjeff|bt5 | oooooooooooooo pretty cursor | 18:53 |
| bazhang | djjeff|bt5, lay off the enter key | 18:53 |
| * djjeff|bt5 pets the cursor with his hand | 18:53 | |
| djjeff|bt5 | ohhhhhhh facepalm I blacklisted nouveau | 18:56 |
| === murthy is now known as murthy_ | ||
| BluesKaj | then look in /etc/modprobe,d/blacklist .conf , comment out the blacklist and install xserver-xorg-video-nouveau | 19:01 |
| BluesKaj | if it needs to be | 19:01 |
| djjeff|bt5 | I dont think the driver is the issue | 19:24 |
| djjeff|bt5 | cause i've tried several drivers | 19:24 |
| djjeff|bt5 | 304,310,313 | 19:24 |
| djjeff|bt5 | it was working 2 days ago | 19:24 |
| djjeff|bt5 | something updated | 19:24 |
| djjeff|bt5 | and since then Xorg seg faults | 19:24 |
| bjsnider | djjeff|bt5, the nvidia packages blacklist nouveau for you | 19:41 |
| bjsnider | and then remove the blacklist if they're removed | 19:41 |
| === keith_ is now known as victor9098 | ||
| djjeff|bt5 | time to fall asleep waiting for | 21:24 |
| djjeff|bt5 | linux-headers-3.8.0-13_3.8.0-13.22_all.deb | 21:24 |
| bazhang | djjeff|bt5, are you asking about 13.04 in #ubuntu ? | 22:07 |
| djjeff|bt5 | meh | 22:08 |
| bazhang | ? | 22:08 |
| MoPac | Hello all. I would very much appreciate some help with diagnosing a hibernation problem I'm having in Raring. System does black on hibernate command, does not appear to write much of anything to disk, and then enters what looks like a kernel panic mode, with two lock lights blinking in sync as they do during a panic. Not sure where to start in terms of pinning down the issue. | 22:23 |
| test | yofel: online? | 23:34 |
| yofel | test: yes, but half asleep? | 23:34 |
| test | yofel: I had to reinstall kubuntu raring and KDE / xserver keeps crashing again after login | 23:35 |
| test | well, it loads the desktpp fine... I can see pics and stuff but then crashes.... kde-workspace-randr is installed | 23:35 |
| yofel | hm, ~/.xsession-errors might tell something | 23:35 |
| yofel | but I'm off to bed and can't help right now, sorry | 23:36 |
| test | ok | 23:36 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
