Re: [gentoo-user] serial in /sys
On Sat, Dec 1, 2012 at 3:47 AM, Walter Dnes waltd...@waltdnes.org wrote: On Fri, Nov 30, 2012 at 09:58:10AM +, Jorge Almeida wrote On Fri, Nov 30, 2012 at 3:09 AM, Walter Dnes waltd...@waltdnes.org wrote: I thought of keeping data-- (key,value) = (serial_number, custom_name_of_device)-- in a cdb database. I think it is faster than reading a text file, and no parsing needed. It is also scalable (although this is probably irrelevant on a custom-workstation scale...). My (untested) program is about 11k, statically compiled. Come to think of it, I'd want to make it more general. I'd have my script read through a textfile (flat-file database) with 3 columns. 1) Attribute; one of manufacturer, product, or serial; are there any others? I checked that netlink and hotplug export exactly the same variables (well, at least for the devices I tried), but SERIAL is not one of them, one has to dig /sys. But the location of the serial file seems to be consistent, so problem solved. 2) Value to match, i.e. manufacturer string, product name, or serial #. 3) The custom name I want my device to have The problem with that approach is that the code is much less simple, and the program less lean. I don't know what that implies regarding efficiency, but inserting a usb pen with only one partition gives rise to 12 (!) events, and thus the program (and hence mdev) is run 12 times, 10 of which are probably completely useless (but there's no helping that). You could have 3 cdb files, one for each type of key (manufacturer,...) and have the program seek them sequentially. It would still be very fast (faster?), and less error prone. (I know cdb is not the most widely known type of database, but it really is the appropriate one for this kind of usage--read many times, write once in a blue moon--and it is very easy to deal with. I'm using tinycdb (http://www.corpit.ru/mjt/tinycdb.html) because it provides a library interface.) By the way, I don't suppose there is a mailing list to talk about these matters (mdev/ udev-alternative/ udev-fork related)? This is really distro-agnostic stuff... For mdev-related questions, the best place I know of is the busybox mailing list http://lists.busybox.net/mailman/listinfo/busybox mdev is part of busybox. OK, I just feel that mdev-related questions are really somewhat related, for example, with the recent intentions of forking udev (which I hope will be successful). For example, the evdev xorg driver requires udev! Is there a really good reason for this, or is it another way to make udev mandatory? (A bit of paranoia...) I don't know what list would be appropriate to raise this question in. Jorge Almeida
Re: [gentoo-user] serial in /sys
On Sat, Dec 1, 2012 at 4:55 AM, Joshua Murphy poiso...@gmail.com wrote: On Fri, Nov 30, 2012 at 10:47 PM, Walter Dnes waltd...@waltdnes.org wrote: On Fri, Nov 30, 2012 at 09:58:10AM +, Jorge Almeida wrote By the way, I don't suppose there is a mailing list to talk about these matters (mdev/ udev-alternative/ udev-fork related)? This is really distro-agnostic stuff... For mdev-related questions, the best place I know of is the busybox mailing list http://lists.busybox.net/mailman/listinfo/busybox mdev is part of busybox. And, while auto-mounting might be a bit beyond the 'official' scope of it, I've seen a fair bit of recent chatter on the LFS mailing lists on the general topic of device management, alongside a bit of grumbling about the kitchen sink approach being taken with systemd/udev/etc. Might be worth glancing through their archives to see if anything on this particular question crops up, and if nothing else, it shouldn't hurt much if you toss the question towards the lfs-chat list. I'm subscribed to the lfs lists (I'm not sure about lfs-chat, though) for years. I'll have a look. Thanks Jorge Almeida
Re: [gentoo-user] threads use flag for apache and php
Am 01.12.2012 00:13, schrieb William Kenworthy: why? - threads sounds like a good thing, but is it really if its optional? Threads - at least for apache - is optional, because you can compile apache with different MPMs (see documentation). MPM_PREFORK for example does not use threads, it forks itself - as opposed to MPM_WORKER which is based on threads. So if you set threads, apache is compiled with MPM_WORKER, if not, MPM_PREFORK is used. If you are running a threaded apache, and you are choosing to use php as apache module (apache2 use flag) then php with threads is required.
[gentoo-user] Re: External monitor is stretched 4:3
On 2012-12-01, Grant wrote: I've connected my laptop to a lot of HDTV's and whenever I switch the output to display on both screens, black bars appear on the left and right of my laptop screen so it displays at 4:3, and the HDTV output is 16:9 but looks horizontally stretched. Does anyone know how to keep the output at 16:9 on both screens? - Grant You don't give a lot of information here. Are you using mirrored screen or an extended desktop? Also what is the desktop environment or window manager you use? On thing that might help is to provide the output of xrandr. I'm using xfce4, but I'm not sure if I'm using a mirrored screen or an extended desktop. All I do is plug the laptop into the HDTV with an HDMI cable and hit the keyboard shortcut to switch screens which brings up a little dialog. There is nothing too complex here, if the TV and laptop are showing the same thing, one screen is *mirroring* the other, otherwise, if you see different things in different screens, you're using an extended desktop. I was able to change the resolution from 1024x768 to 1366x768 with xfce4's Display settings, but when I disconnect and reconnect to the HDTV it displays at 1024x768 again. Do you know how to select the output resolution for an external screen permanently? Is this done in xorg.conf? This is, I'd guess, a preferred video mode announced through EDID, where the TV, even if it supports 1366x768, will anounce 1024x768 as preferred. You could do the change with a small xrandr one-liner, and there must be some way to do it through xorg.conf, although I don't know how. In the end, having the output of xrandr (both before and after you change the video modes) would help *a lot*, as it answers most of our questions... -- Nuno Silva (aka njsg) http://njsg.sdf-eu.org/
Re: [gentoo-user] xfig won't compile
- Mail original - De: Alain Didierjean alain.didierj...@free.fr À: gentoo-user@lists.gentoo.org Envoyé: Vendredi 30 Novembre 2012 10:00:27 Objet: Re: [gentoo-user] xfig won't compile - Mail original - De: Peter Weilbacher newss...@weilbacher.org À: gentoo-user@lists.gentoo.org Envoyé: Vendredi 30 Novembre 2012 09:54:07 Objet: Re: [gentoo-user] xfig won't compile On 2012-11-27 12:34, Hinnerk van Bruinehsen wrote: On Tue, Nov 27, 2012 at 12:20:02PM +0100, Alain Didierjean wrote: I cannot emerge xfig. Both versions (amd64 ~amd64) return that informative message: * Messages for package media-gfx/xfig-3.2.5b-r2: [...] So: - known bug ? - tip available ? or should I fill a bug report ? [...] Otherwise you could file a bugreport (where you should attach the build.log nonetheless) There is https://bugs.gentoo.org/show_bug.cgi?id=405475. Peter. Thanks, Peter, I'll check that. As for you, Hinnerk, your method is OK, but as for me, when that type of problem arises, I first check on this list if it is a known problem, eventually with a work around. Then I check for an existing bug report, then I fill abug report if necessary. As a conclusion: the link provided by Peter lead to the description of the bug I had noticed and to a quick and dirty patch allowing me to print a big fat LyX document with a lot of .fig figures, urgently needed by my students. Can I ask for anything more ?
Re: [gentoo-user] emerge --update : how to keep it going?
Now all we need is, emerge --i-dont-care for when you have 100 packages ready to be updated and one stupid ruby package has conflicting dependencies. I doubt if there is point to runaway from problms. In your case I'd say keep calm and --keep-going :)
Re: [gentoo-user] Re: External monitor is stretched 4:3
On Saturday 01 Dec 2012 09:49:01 Nuno J. Silva wrote: On 2012-12-01, Grant wrote: I've connected my laptop to a lot of HDTV's and whenever I switch the output to display on both screens, black bars appear on the left and right of my laptop screen so it displays at 4:3, and the HDTV output is 16:9 but looks horizontally stretched. Does anyone know how to keep the output at 16:9 on both screens? - Grant You don't give a lot of information here. Are you using mirrored screen or an extended desktop? Also what is the desktop environment or window manager you use? On thing that might help is to provide the output of xrandr. I'm using xfce4, but I'm not sure if I'm using a mirrored screen or an extended desktop. All I do is plug the laptop into the HDTV with an HDMI cable and hit the keyboard shortcut to switch screens which brings up a little dialog. There is nothing too complex here, if the TV and laptop are showing the same thing, one screen is *mirroring* the other, otherwise, if you see different things in different screens, you're using an extended desktop. I was able to change the resolution from 1024x768 to 1366x768 with xfce4's Display settings, but when I disconnect and reconnect to the HDTV it displays at 1024x768 again. Do you know how to select the output resolution for an external screen permanently? Is this done in xorg.conf? This is, I'd guess, a preferred video mode announced through EDID, where the TV, even if it supports 1366x768, will anounce 1024x768 as preferred. You could do the change with a small xrandr one-liner, and there must be some way to do it through xorg.conf, although I don't know how. In the end, having the output of xrandr (both before and after you change the video modes) would help *a lot*, as it answers most of our questions... You can set this up either with xranrd entries in your ~/.xprofile or in your /etc/X11/xorg.conf First run xrandr -q to see what you get from an xterminal and then manually alter the resolution according to your requirements on each screen, e.g. xrandr --output DVI0 --mode 1366x768 --rate 60 xrandr --output LVDS1 --primary xrandr --output TV1 --mode 1920x1080 etc. until you get things as you want them. Look at your /var/log/Xorg.0.log to see what resolution and refresh rate your card can do, although xrandr -q will show this. The refresh rate is not really required (I think it is automatically set to match the screen resolution). You can if you prefer set it up in /etc/X11/xorg.conf: Name your monitors in your Section Device: === Section Device [snip ...] Identifier Card0 Driver radeon BusID PCI:1:0:0 Option monitor-VGA my 2nd monitor Option monitor-LVDS my laptop Option monitor-TV1 my TV EndSection === Then set up the screen resolutions for each monitor: === Section Monitor Identifier my 2nd monitor Option PreferredMode 1024x768 OptionPosition 1024 0 EndSection Section Monitor Identifier my laptop Option PreferredMode 1366x768 OptionLeftOf my 2nd monitor EndSection [snip ...] EndSection === Then set up the default screen: == Section Screen Identifier Screen0 Device Card0 Monitor my laptop [snip ...] == The above is just a guide of course. You can tweak it according to your requirements and see what gives. For more permanent set ups I would tend to use xorg.conf (old habits die hard), but .xprofile may be quicker/easier to try out. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Kmail-1.13.7 subfolders are no longer recognised
Since I updated KDE to 4.9.3 my Kmail 1.13.7 no longer recognises the Sent Draft subfolders I have set up, to save sent/draft messages from different email accounts (both POP IMAP). All sent messages regardless of the account I send them from, end up in the defaul Kmail top level sent-mail folder. Changing the settings in 'Settings/Configure Kmail/Identities/Advanced/Sent- mail folder/' or '/Drafts folder' is not saved no matter what I do. Looking in kmailrc the paths seem correct, but are not recognised by the application when sending messages. Has anyone come across this? -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] intel HD graphics 4000 and viewing DVDs
Am 28.07.2012 10:22, schrieb Florian Philipp: Am 27.07.2012 22:57, schrieb Michael Mol: On Fri, Jul 27, 2012 at 4:39 PM, Florian Philipp li...@binarywings.net wrote: Am 27.07.2012 22:22, schrieb Michael Mol: On Fri, Jul 27, 2012 at 4:10 PM, Allan Gottlieb gottl...@nyu.edu wrote: I am getting a new laptop. (likely dell 6430). The two graphics options are intel HD 4000 and nvidia NVS 5200M. Dell is as expected suggesting the 5200M. I do not need 3D or fast response. Dell hinted that DVDs might not play with the intel HD 4000. This seems weird to me as the 4000 is supposed to be a big improvement over the 3000 and I can't believe dell or others would have sold laptops that can't play dvds Any comments or experiences? My Duron 750MHz was able to decode DVDs in realtime. After that, all you're doing is blitting (or using xv) the frames to the screen. I would be absolutely shocked if the Intel HD 4000 GPU couldn't handle that basic of a 2D acceleration function. Now, DVDs use MPEG2. Blu-Ray uses h.264, which is a much harder beast to decode in realtime. It's possible the HD 4000 GPU can't handle hardware decode of h.264, but I don't know. I've never looked into it. (Software decode of 1080p h.264 on my Phenom 9650 worked somewhat, but highly active scenes would cause frame drops.) I've experienced issues playing DVDs on fullscreen with the OSS radeon driver. Therefore I'm cautious of assumptions that something works simply because the input is easy to decode. Upscaling to large displays with high resolutions can be an issue. I'm not saying the Intel driver cannot handle it. I'm just saying you should try it or look for reports. How high is 'high' resolution? I was upscaling to 1600x1200 using an a Radeon 9600; that card would now be almost ten years old. A bit later, I did the same on a 2.4GHz Pentium 4 using an i845-based Intel graphics card. Here's the line from lspci, as run in May of 2007: 00:02.0 VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01) Hardware scaling a 2D image is one of the most trivial hardware-accelerated options GPUs perform. If someone had difficulties upscaling a 480p (roughly what DVDs are) to 1080p at 24 or 33fps, I would be very highly suspicious of a software misconfiguration. That kind of scaling should even be comfortably doable in software on any modern x86-derived processor. (With the plausible exclusion of the Atom CPU) 1920x1080, on-board Radeon HD 4250. I haven't diagnosed it further (except of playing around with mplayer2 options) as it was easier to use the closed source driver. Regards, Florian Philipp I realize this thread is pretty stale but since I talked bullshit and just now realized it, I want to correct myself: Since updating the kernel to 3.5 forced me to update the X server beyond 1.11 which in turn forced me to update ati-drivers to a version that no longer supported my Radeon HD 4250, I had to look into my issues with the open source driver. It turns out, my problems had two reasons: - I didn't enable KMS and DRM for radeon in the kernel - I didn't have x11-drivers/radeon-ucode installed Both resulted in a fully functioning X server that - could run glxgears just fine - could (with some tuning) render videos in DVD quality with opengl output - was too slow for videos in any higher resolution Regards, Florian Philipp signature.asc Description: OpenPGP digital signature
Re: [gentoo-user] Kmail-1.13.7 subfolders are no longer recognised
On Saturday 01 December 2012 14:22:09 Mick wrote: Since I updated KDE to 4.9.3 my Kmail 1.13.7 no longer recognises the Sent Draft subfolders I have set up, to save sent/draft messages from different email accounts (both POP IMAP). All sent messages regardless of the account I send them from, end up in the defaul Kmail top level sent-mail folder. Changing the settings in 'Settings/Configure Kmail/Identities/Advanced/Sent- mail folder/' or '/Drafts folder' is not saved no matter what I do. Looking in kmailrc the paths seem correct, but are not recognised by the application when sending messages. Has anyone come across this? Not exactly, but I do find that my filter to send messages from this list to their own folder sometimes works and sometimes doesn't. Then I read each one in the in-box and hit ctrlj and the filter does work. It seems that filtering is broken in the current version, which is 4.9.3 here. How do you have 1.13.7 still around? -- Rgds Peter
Re: [gentoo-user] Kmail-1.13.7 subfolders are no longer recognised
On Saturday 01 Dec 2012 16:31:15 Peter Humphrey wrote: On Saturday 01 December 2012 14:22:09 Mick wrote: Since I updated KDE to 4.9.3 my Kmail 1.13.7 no longer recognises the Sent Draft subfolders I have set up, to save sent/draft messages from different email accounts (both POP IMAP). All sent messages regardless of the account I send them from, end up in the defaul Kmail top level sent-mail folder. Changing the settings in 'Settings/Configure Kmail/Identities/Advanced/Sent- mail folder/' or '/Drafts folder' is not saved no matter what I do. Looking in kmailrc the paths seem correct, but are not recognised by the application when sending messages. Has anyone come across this? Not exactly, but I do find that my filter to send messages from this list to their own folder sometimes works and sometimes doesn't. Then I read each one in the in-box and hit ctrlj and the filter does work. It seems that filtering is broken in the current version, which is 4.9.3 here. How do you have 1.13.7 still around? Because kmail-1.13.7 is still the last stable version in portage and when I tried kmail-2 on another box I could never get it to work properly with sqlite3. Kmail-2 really screwed up my messages (creating duplicates and being unable to delete them, not showing IMAP4 folders, or their content, etc.) so I steered away from trying to upgrade for good, or at least until kmail-1 is no longer supported. I did try to learn how to use mutt, but I found that it was getting in the way big time when I had to use s/mime and gnupg for signing/encrypting messages - trying to remember what the serial number of a certificate is for each recipient is a bit too much even for a console application. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] emerge --update : how to keep it going?
Am Samstag, 1. Dezember 2012, 15:02:10 schrieb 2sb7...@gmail.com: Now all we need is, emerge --i-dont-care for when you have 100 packages ready to be updated and one stupid ruby package has conflicting dependencies. I doubt if there is point to runaway from problms. In your case I'd say keep calm and --keep-going :) --keep-going does not help you, if the emerge does not start because of missing dep/slot conflict/blocking/masking whatever... -- #163933
Re: [gentoo-user] intel HD graphics 4000 and viewing DVDs
On Sat, Dec 01 2012, Florian Philipp wrote: Am 28.07.2012 10:22, schrieb Florian Philipp: Am 27.07.2012 22:57, schrieb Michael Mol: On Fri, Jul 27, 2012 at 4:39 PM, Florian Philipp li...@binarywings.net wrote: Am 27.07.2012 22:22, schrieb Michael Mol: On Fri, Jul 27, 2012 at 4:10 PM, Allan Gottlieb gottl...@nyu.edu wrote: I am getting a new laptop. (likely dell 6430). The two graphics options are intel HD 4000 and nvidia NVS 5200M. Dell is as expected suggesting the 5200M. I do not need 3D or fast response. Dell hinted that DVDs might not play with the intel HD 4000. This seems weird to me as the 4000 is supposed to be a big improvement over the 3000 and I can't believe dell or others would have sold laptops that can't play dvds Any comments or experiences? My Duron 750MHz was able to decode DVDs in realtime. After that, all you're doing is blitting (or using xv) the frames to the screen. I would be absolutely shocked if the Intel HD 4000 GPU couldn't handle that basic of a 2D acceleration function. Now, DVDs use MPEG2. Blu-Ray uses h.264, which is a much harder beast to decode in realtime. It's possible the HD 4000 GPU can't handle hardware decode of h.264, but I don't know. I've never looked into it. (Software decode of 1080p h.264 on my Phenom 9650 worked somewhat, but highly active scenes would cause frame drops.) I've experienced issues playing DVDs on fullscreen with the OSS radeon driver. Therefore I'm cautious of assumptions that something works simply because the input is easy to decode. Upscaling to large displays with high resolutions can be an issue. I'm not saying the Intel driver cannot handle it. I'm just saying you should try it or look for reports. How high is 'high' resolution? I was upscaling to 1600x1200 using an a Radeon 9600; that card would now be almost ten years old. A bit later, I did the same on a 2.4GHz Pentium 4 using an i845-based Intel graphics card. Here's the line from lspci, as run in May of 2007: 00:02.0 VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01) Hardware scaling a 2D image is one of the most trivial hardware-accelerated options GPUs perform. If someone had difficulties upscaling a 480p (roughly what DVDs are) to 1080p at 24 or 33fps, I would be very highly suspicious of a software misconfiguration. That kind of scaling should even be comfortably doable in software on any modern x86-derived processor. (With the plausible exclusion of the Atom CPU) 1920x1080, on-board Radeon HD 4250. I haven't diagnosed it further (except of playing around with mplayer2 options) as it was easier to use the closed source driver. Regards, Florian Philipp I realize this thread is pretty stale but since I talked bullshit and just now realized it, I want to correct myself: Since updating the kernel to 3.5 forced me to update the X server beyond 1.11 which in turn forced me to update ati-drivers to a version that no longer supported my Radeon HD 4250, I had to look into my issues with the open source driver. It turns out, my problems had two reasons: - I didn't enable KMS and DRM for radeon in the kernel - I didn't have x11-drivers/radeon-ucode installed Both resulted in a fully functioning X server that - could run glxgears just fine - could (with some tuning) render videos in DVD quality with opengl output - was too slow for videos in any higher resolution Regards, Florian Philipp Thanks for the response. I should say that I have indeed purchased the laptop with intel graphics and it works fine with DVDs. allan
Re: [gentoo-user] emerge --update : how to keep it going?
Volker Armin Hemmann volkerar...@googlemail.com writes: --keep-going does not help you, if the emerge does not start because of missing dep/slot conflict/blocking/masking whatever... Though it would be nice if there was some flag, probably mainly of use with either ' -u @world' or --resume, to tell portage to get on and merge what it can and leave any masked packages or those which would generate blockers or conflicts.
Re: [gentoo-user] emerge --update : how to keep it going?
On Sat, Dec 1, 2012 at 11:58 AM, Graham Murray gra...@gmurray.org.uk wrote: Volker Armin Hemmann volkerar...@googlemail.com writes: --keep-going does not help you, if the emerge does not start because of missing dep/slot conflict/blocking/masking whatever... Though it would be nice if there was some flag, probably mainly of use with either ' -u @world' or --resume, to tell portage to get on and merge what it can and leave any masked packages or those which would generate blockers or conflicts. emerge -fDuN @world will both check/resolve dependencies as well as insure that all files required to do the build are downloaded. I typically run this command prior to emerging anything as I dislike coming back and hour later and finding the emerge didn't finish because a file either couldn't be downloaded or its manifest didn't check out. Just an alternative way to do things. HTH, Mark
Re: [gentoo-user] emerge --update : how to keep it going?
Am Samstag, 1. Dezember 2012, 12:19:17 schrieb Mark Knecht: On Sat, Dec 1, 2012 at 11:58 AM, Graham Murray gra...@gmurray.org.uk wrote: Volker Armin Hemmann volkerar...@googlemail.com writes: --keep-going does not help you, if the emerge does not start because of missing dep/slot conflict/blocking/masking whatever... Though it would be nice if there was some flag, probably mainly of use with either ' -u @world' or --resume, to tell portage to get on and merge what it can and leave any masked packages or those which would generate blockers or conflicts. emerge -fDuN @world will both check/resolve dependencies as well as insure that all files required to do the build are downloaded. I typically run this command prior to emerging anything as I dislike coming back and hour later and finding the emerge didn't finish because a file either couldn't be downloaded or its manifest didn't check out. Just an alternative way to do things. HTH, Mark that doesn't solve the problem I mentioned above. -- #163933
[gentoo-user] kernel 3.2-3.5 upgrade unusable: keyboard borked
I'm trying to upgrade from a 3.2 kernel to 3.5.7, but the 3.5.7 kernel is unusable because it always puts the keyboard into a mode where it maps the numeric keypad to the right-hand home position (J-1, K-2, L-3, U-4, etc.). After sshing into the machine and booting back into 3.2, everything is fine again. There must have been a new kernel setting that I missed when I did a make oldconfig which defaults to an unusable settings. I haven't been able to come up with a Google search that provides anything remotely relevent. Does anybody recognize this problem? -- Grant Edwards grant.b.edwardsYow! I represent a at sardine!! gmail.com
Re: [gentoo-user] kernel 3.2-3.5 upgrade unusable: keyboard borked
On Sunday 02 Dec 2012 3:05:21 Grant Edwards wrote: I'm trying to upgrade from a 3.2 kernel to 3.5.7, but the 3.5.7 kernel is unusable because it always puts the keyboard into a mode where it maps the numeric keypad to the right-hand home position (J-1, K-2, L-3, U-4, etc.). After sshing into the machine and booting back into 3.2, everything is fine again. There must have been a new kernel setting that I missed when I did a make oldconfig which defaults to an unusable settings. I haven't been able to come up with a Google search that provides anything remotely relevent. Does anybody recognize this problem? Is this a laptop? with no num pad? On my laptop the numpad is mapped to the keys like you described, so when Num Lock is toggled those keys function as the num pad. -- - Yohan Pereira The difference between a Miracle and a Fact is exactly the difference between a mermaid and a seal. -- Mark Twain
