Re: [gentoo-user] Google Chrome and audio capture
Mick wrote: On Tuesday 23 Jun 2015 11:54:02 Neil Bothwick wrote: On Tue, 23 Jun 2015 05:26:31 -0500, Dale wrote: Take from that what you will. Note, the issues are for chromium and not for Google Chrome, shouldn't make a difference for what you want to know though. Thanks. That was what I was looking for. I guess they did do this then. This may be the first time I checked into a story from that site and it be true. It seems google did sort of sneak some code in there. o_O There is a now a USE flag to specifically enable this. It defaults to disabled but if you previously emerged chromium before the flag as added, you will still have it. Using --newuse will cause a world update to re-emerge chromium, but if you use --changed-use it doesn't, so re-emerge chromium if you want to get rid of this. What is the new USE flag and does it also apply to 43.0.2357.65? https://code.google.com/p/chromium/issues/detail?id=491435 It's mentioned in post 4 first. I don't know if Gentoo is using the same name or not tho. It should be a start at least. Dale :-) :-)
Re: [gentoo-user] One Time Passwords
On 06/24/2015 02:04:57 AM, Rich Freeman wrote: On Tue, Jun 23, 2015 at 8:49 AM, Helmut Jarausch jarau...@skynet.be wrote: Hi, I'd like to log into my Gentoo system from my smartphone. But I don't trust Google (Android's parents). Therefore I need a OTP solution for loggin into my Gentoo system. Can anybody recommend a solution? You'll laugh at the irony, but my /etc/pam.d/sshd: auth include system-remote-login auth required pam_google_authenticator.so accountinclude system-remote-login password include system-remote-login sessioninclude system-remote-login The Google Authenticator PAM module comes from sys-auth/google-authenticator, and accepts OTPs from the Google Authenticator app, or any other app that uses the same algorithm (which is fairly standard I believe). It is FOSS, and doesn't give Google access to anything. That one line is all it takes to block anybody not using an OTP from logging in. To actually set the key for an account there is a utility that will generate a key and give you the seed for your OTP generator. It stores a file in your home directory with the seed, which the PAM module reads. It is very simple to set up, and very effective. Note that public key authentication with sshd normally bypasses PAM and doesn't require the code - I don't know offhand if you can have both. Many thanks, Rich. I wouldn't like to use an OTP generator on my smartphone because Big Brother might watching me when I use this. I feel like the German parliament which has been hacked by a foreign secrete service. Parliamentarians have to resort to classic types of communication now. I'd prefer a solution where I carry the OTPs with me in printed form. Is this possble with the google-authenticator, as well?
Re: [gentoo-user] Google Chrome and audio capture
On Tuesday 23 Jun 2015 11:54:02 Neil Bothwick wrote: On Tue, 23 Jun 2015 05:26:31 -0500, Dale wrote: Take from that what you will. Note, the issues are for chromium and not for Google Chrome, shouldn't make a difference for what you want to know though. Thanks. That was what I was looking for. I guess they did do this then. This may be the first time I checked into a story from that site and it be true. It seems google did sort of sneak some code in there. o_O There is a now a USE flag to specifically enable this. It defaults to disabled but if you previously emerged chromium before the flag as added, you will still have it. Using --newuse will cause a world update to re-emerge chromium, but if you use --changed-use it doesn't, so re-emerge chromium if you want to get rid of this. What is the new USE flag and does it also apply to 43.0.2357.65? -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] purchasing a dell laptop
On 23/06/2015 18:53, gottl...@nyu.edu wrote: On Sun, Jun 21 2015, Alan McKinnon wrote: On 21/06/2015 21:16, gottl...@nyu.edu wrote: I am seriously thinking of updating my dell 6430s laptop purchased 3 years ago. NYU has an arrangement with dell so that is the only maker I am considering. This machine will be essentially gentoo only (I configure my computers to dual boot some version of windows for ease in dealing with dell support). [...] So on the whole, my experience with higher-end Dell is that hardware is pretty much well-supported across the boards with very few gotchas. The only two exceptions would be wifi cards (cheap to fix) and maybe GPU co-processor (if you are unlucky to get an unsupported cutting edge one and need to wait a bit for Linux support to catch up). I've had similar experiences but very much appreciate the confirmation and your comments. One more thing I can add: From observation, I can say that Dell has 2 or more grades of kit they sell: 1. Cheap shit. You find these in supermarkets and Walmart. It's just as crappy as all the other cheap shit around with the same bargain basement price. 2. Good stuff like the Precision and XPS range. You tend not to find these at Walmart, and have to go to a proper dealer for them, or your workplace has a scheme to provide them. You want category #2. But seeing as you are looking at the i7-5600U and similar, I think you are already in that bracket. Just wanted to clarify that :-) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Google Chrome and audio capture
On Wed, 24 Jun 2015 07:08:01 +0100 Mick wrote: On Tuesday 23 Jun 2015 11:54:02 Neil Bothwick wrote: On Tue, 23 Jun 2015 05:26:31 -0500, Dale wrote: Take from that what you will. Note, the issues are for chromium and not for Google Chrome, shouldn't make a difference for what you want to know though. Thanks. That was what I was looking for. I guess they did do this then. This may be the first time I checked into a story from that site and it be true. It seems google did sort of sneak some code in there. o_O There is a now a USE flag to specifically enable this. It defaults to disabled but if you previously emerged chromium before the flag as added, you will still have it. Using --newuse will cause a world update to re-emerge chromium, but if you use --changed-use it doesn't, so re-emerge chromium if you want to get rid of this. What is the new USE flag and does it also apply to 43.0.2357.65? The flag is USE=hotwording, it applies to 45.0.2431.0 and later versions. Please note that this flag disables autoload of hotwording nacl plugin, so if one had earlier chromium versions installed, one will still have this plugin installed on a system. In order to remove already installed plugin one have to delete the following directory: ~/.config/chromium/Default/Extensions/lccekmodgklaepjeofjdjpbminllajkg See also: https://bugs.gentoo.org/show_bug.cgi?id=552298 Best regards, Andrew Savchenko pgpnWFEhA3sVN.pgp Description: PGP signature
Re: [gentoo-user] repos.conf
2015-06-23 21:12 GMT-06:00 James wirel...@tampabay.rr.com: Take java: * Warnings: * -- * The source of the overlay java seems to have changed. * You currently sync from * * git://git.overlays.gentoo.org/proj/java.git * * while the remote lists report * * 1. git://anongit.gentoo.org/proj/java.git * 2. git+ssh://g...@git.gentoo.org/proj/java.git * 3. https://anongit.gentoo.org/git/proj/java.git * * as correct locations. * Please consider removing and re-adding the overlay. This is because recent changes of the infrastructure providing overlays, explanation it's on the frontpage(gentoo.org) Maybe a few examples of /etc/portage/repos.conf to look at? Maybe this is useful, what I have been doing, is use layman as always for external repos from gentoo overlays( layman -a ...), As far as I know the repos.conf still doesn't handle svn, so I think devs are keeping layman the way it is to not break this, but if I want to change something in a particular external overlay say java, what i do is this A new conf /etc/repos.conf/java.conf [java] priority = 100 location = /home/$MY_USER/$MY_OVERLAYS_DIR/java layman-type = git auto-sync = no And of course clone the repo to that dir and make my changes. Note the gentoo and local is fine. I keep /usr/local/portage for my stuff alone, so I guess I can just use the old layman dir stucture (/var/lib/layman) for my collection of exteranal repos? I find the use of /usr/local/portage less useful than making your own overlay, I have my own overlay in repos.conf pretty much the same as the example before /etc/repos.conf/j-overlay.conf:(This is my box used for development) [j-overlay] priority = 100 location = /home/$MY_USER/$MY_OVERLAYS_DIR/j-overlay layman-type = git auto-sync = no But I distribute it to some containers and some other gentoo installations by a little different config file: [j-overlay] location = /var/lib/portage/repos/j-overlay sync-type = git priority = 100 sync-uri = https://github.com/j-g-/j-overlay.git auto-sync = true And as expected, emerge --sync syncs the portage three(also via git) and changes I commit to my own overlay. I hope you find this useful.
Re: [gentoo-user] One Time Passwords
On Wed, Jun 24, 2015 at 5:29 AM, Helmut Jarausch jarau...@igpm.rwth-aachen.de wrote: I wouldn't like to use an OTP generator on my smartphone because Big Brother might watching me when I use this. I feel like the German parliament which has been hacked by a foreign secrete service. Parliamentarians have to resort to classic types of communication now. I'd prefer a solution where I carry the OTPs with me in printed form. Is this possble with the google-authenticator, as well? Well, the protocol for generating the TOTPs is standard: https://tools.ietf.org/html/rfc6238 I don't have any recommendations, but it seems like there are various dongles out there which generate the codes. There might be others around here that have a bit more experience with this but you could probably get them working. Pre-printed OTPs are a different matter, but they suffer from the obvious vulnerability that they can be copied (which is definitely a big contrast from a hardware OTP generator which is typically hardened against such attacks - having a code is basically proof that you have the device in your possession RIGHT NOW vs having had it in your possession at some time in the past). I used to use skey (which is packaged on Gentoo) for this. Skey uses OTPs which are sequential instead of time-based. You can use it in a challenge/response mode which requires software (such as on your phone), but you can also pre-print a big list of keys and carry them with you. When ssh asks you for key# 100 you look for it on the list and type it in. You can print more keys at any time as they get used up. The Yubikey is getting a lot of attention right now with protocols like U2F (as well as OTP). It is cheap and capable. The downside of the Ubikey is that I believe you can only use it if it is plugged into a USB port, and protocols like U2F are designed more around browsers than various other bits of software that authenticate (like POP, ssh, etc). However, I believe you can plug it into a PC, hit the button, and have it act as a keyboard and type in a TOTP. So, if you're sshing from PCs with USB ports that accept external keyboards it might be an option for you. I've been assuming that you're talking about ssh all along. That tends to work well since ssh clients generally support an interactive login with some kind of challenge/response or such. It breaks down for other protocols that don't have allowance for that like IMAP/POP/etc, unless you ditch the regular password and just pass the OTP as your password (and honestly I'm not sure how great an idea that is). And, of course, unless your mail client keeps the connection open that could get really painful anytime it checks for new mail. All that said, on the list of things I worry about each day, the Google is secretly uploading data from my phone to their servers without telling/asking fear is pretty low on the list. Google authenticator does not sync to any kind of central server, and its source is published. Sure, the OS could be spying on you, but so could your ethernet card or any number of other things. But, if you really want strong security an un-networked hardware token whose seed can be set by the user is probably your best bet. -- Rich
Re: [gentoo-user] necessary use flgas
On Wed, Jun 24, 2015 at 01:13:40PM +0200, Alan McKinnon wrote: On 23/06/2015 15:05, behrouz khosravi wrote: Hello everyone. I really like to have control over my machine as much as possible. In this way I will learn a lot, so I am trying to remove all the default use flags and control them manually. Here's some good advice: Don't do that. See below. Nonsense - do that. If your goal is to learn how stuff works and you're already reasonably familiar with C/C++ so you can debug any strange errors that can happen, have fun. Just don't think you'll get any real work done ;). i.e. it might be good to do this in a virtual machine and still have a stable system for work. I just don't know which global use flags are absolutely necessary to the system to make it snappier or secure. That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. ... Put -march=native in CFLAGS Yes. Also, properly setting CPU_FLAGS_X86 is another thing that can speed up software *if* said software supports any special instruction sets. Most normal desktop software like web browsers, email clients, terminals, editors, etc. probably will not get a whole lot of benefit either way, since most of this software is generally not CPU-bound and is instead network/disk bound. In the mornings I primarily use my desktop for reading email and browsing news with firefox (mostly on sites with minimal JavaScript), and I have yet to see my load averages climb higher than maybe 0.5. Any software that does anything requiring lots of math will get a boost from this type of stuff, though; graphics editing, most things in sci-* categories, audio/video transcoding, etc. Alec P.S. Just realized I don't have -march=native in my CFLAGS. Time to rice - could be getting 1% better performance. ;) P.P.S. Also, on 1% better performance: My professor for the compilers class I took used to (maybe still does) work at Google. Apparently Google sees a 1% increase in performance as *the best thing ever*, because it can save them a bunch of money in infrastructure and power. Apparently Google are the ultimate ricers.
Re: [gentoo-user] necessary use flgas
Here's some good advice: Don't do that. See below. Oops! I have done it and I am happy so far ! That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. I think you misunderstood me! for example adding CPU specific flags is a good idea right? I meant something like that. For example is it wise to enable opengl flag globally ? is it helpful to do so? What do you recommend ? DO NOT SET USE=-* As I said before I have done it and I totally recommend it to anyone interested to get a better understanding of user land. Pick a profile that suits what you want to use the computer for. You have a desktop? Pick a suitable desktop profile. Don't pick a KDE one unless oyu use KDE for instance (all that does is set some KDE flags (like semantic-desktop or baloo or whatever they call it now) and force some KDE packages to be merged. It doesn't change the underlying way things work. desktop profiles are very big for my taste. In fact I have been using KDE for about a year on the default (basic) profile. I have compiled the KDE with KDE profile and I have witnessed the differences with my own eyes. I very much doubt you can increase security by picking some USE flags. There is no USE=open-me-up-to-the-world or USE=rock-solid-nsa-proff-tight USE flags :-) So what security features do you need or want? Figure that out and then set the system up to provide that. You will get what you want. Well I know there is no USE flag like that! I am not that stupid but I remember that I have read somewhere(unfortunately I dont remember where) that disabling some use flags will degrade the security of system.
RE: [gentoo-user] necessary use flgas
behrouz khosravi wrote: Hello everyone. I really like to have control over my machine as much as possible. In this way I will learn a lot, so I am trying to remove all the default use flags and control them manually. I just don't know which global use flags are absolutely necessary to the system to make it snappier or secure. What do you recommend ? Thanks Oh, USE-Flags are so boing :( They are documented well (quse -D $FLAG), you can even look into specific ebuilds what they actually do. If you want to learn Gentoo the hard way I suggest to emerge -C python. or emerge -C glibc (yeah, glibc is gnome, and I hate gnome! Whoops, it was glib, not glibc... errr) Or install certain packages by hand (make install) into /usr/local and forget about it. (Before someone asks: Those are issues people actually had in the forums and it's really hard to track them down or fix them, you will learn quite a bit about your system ;)...) To be more serious: * Set a minimal basic profile (as already suggested) * Tune your USE-Flags in make.conf. media-related flags (mp3, flac) should be harmless, if you touch flags that get used in core packages (e.g. in the toolchain) double (or triple) check if you don't do evil things. * fine tune USE-Flags on a per-package-base via /etc/portage/package.use
Re: [gentoo-user] necessary use flgas
On 23/06/2015 15:05, behrouz khosravi wrote: Hello everyone. I really like to have control over my machine as much as possible. In this way I will learn a lot, so I am trying to remove all the default use flags and control them manually. Here's some good advice: Don't do that. See below. I just don't know which global use flags are absolutely necessary to the system to make it snappier or secure. That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. Long ago (like 2004) there were a bunch of really stupid fanboy Gentoo users[1] without any clue at all who thought they could make their system race along at faster-than-light speed by tweaking flags. What they actually did was 1. usually slow it down and 2. break it completely with insane compiler flags (like -O9[2]) The single most significant thing you can do to your system to avoid it being slow (note, I did NOT say make it fast) is to select an appropriate CPU type for the compiler to build with. All other optimizations tend to be insignificant compared to just this one thing. Put -march=native in CFLAGS What do you recommend ? DO NOT SET USE=-* This is only useful for people who want a profile that Gentoo does not provide or assemble a system in a way that Gentoo isn't built for. Or people who really know what they are doing and why. You are nowhere near this category. Pick a profile that suits what you want to use the computer for. You have a desktop? Pick a suitable desktop profile. Don't pick a KDE one unless oyu use KDE for instance (all that does is set some KDE flags (like semantic-desktop or baloo or whatever they call it now) and force some KDE packages to be merged. It doesn't change the underlying way things work. Then look at what you have. You never print? Don't install cups or set it's flag. etc, etc, etc I very much doubt you can increase security by picking some USE flags. There is no USE=open-me-up-to-the-world or USE=rock-solid-nsa-proff-tight USE flags :-) So what security features do you need or want? Figure that out and then set the system up to provide that. You will get what you want. It's a lot like starting a restaurant, and wondering what must go on the menu. First state what kind of restaurant, then the menu is easy. If you have a fancy French place, you don't sell pizza and don't need the oven. Got a small cozy like bistro place? Then you DO need a pizza oven. See? You can't answer the menu question till you know what kind of place. Thanks [1] Lucky for us, they all moved on to other distros. Most folks left behind in Gentoo now understand their systems well [2] Which doesn't actually exist -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] necessary use flgas
On Wed, Jun 24, 2015 at 01:13:40PM +0200, Alan McKinnon wrote: On 23/06/2015 15:05, behrouz khosravi wrote: Hello everyone. I really like to have control over my machine as much as possible. In this way I will learn a lot, so I am trying to remove all the default use flags and control them manually. Here's some good advice: Don't do that. See below. Nonsense - do that. If your goal is to learn how stuff works and you're already reasonably familiar with C/C++ so you can debug any strange errors that can happen, have fun. Just don't think you'll get any real work done ;). i.e. it might be good to do this in a virtual machine and still have a stable system for work. I just don't know which global use flags are absolutely necessary to the system to make it snappier or secure. That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. ... Put -march=native in CFLAGS Yes. Also, properly setting CPU_FLAGS_X86 is another thing that can speed up software *if* said software supports any special instruction sets. Most normal desktop software like web browsers, email clients, terminals, editors, etc. probably will not get a whole lot of benefit either way, since most of this software is generally not CPU-bound and is instead network/disk bound. In the mornings I primarily use my desktop for reading email and browsing news with firefox (mostly on sites with minimal JavaScript), and I have yet to see my load averages climb higher than maybe 0.5. Any software that does anything requiring lots of math will get a boost from this type of stuff, though; graphics editing, most things in sci-* categories, audio/video transcoding, etc. Alec P.S. Just realized I don't have -march=native in my CFLAGS. Time to rice - could be getting 1% better performance. ;) P.P.S. Also, on 1% better performance: My professor for the compilers class I took used to (maybe still does) work at Google. Apparently Google sees a 1% increase in performance as *the best thing ever*, because it can save them a bunch of money in infrastructure and power. Apparently Google are the ultimate ricers.
[gentoo-user] lvmetad Errors
I have been getting some lvmetad related errors, some of which can certainly be found on Google, but I haven't had luck fixing them so far. I used to get the following error in /var/log/rc.log /run/lvm/lvmetad.socket: connect failed: No such file or directory I am not sure why this was happening, but I read that one fix was to disable lvmetad entirely. And so I set use_lvmetad = 0 in /etc/lvm/lvm.conf. However, now I get error messages of the following form (e.g. when running grub2-mkconfig): WARNING: lvmetad is runing but disabled. Restart lvmetad before enabling it! I ran /etc/init.d/lvmetad needsme and got lvm-monitor lvm I don't understand why lvm is starting lvmetad even when I have disabled it. lvmetad is certainly not in any runlevel itself and I just don't know where to go from here. I would appreciate some advice on how to fix either problem (the socket connect issue when lvmetad is enabled, or the fact that lvmetad still starts when it's disabled...) Thanks, Alex
[gentoo-user] Sorry for the spam
List, Sorry for the spam. I recently switched to mutt and am still learning what all of the various flags in the index mean. My bad. Alec
Re: [gentoo-user] necessary use flgas
To be more serious: * Set a minimal basic profile (as already suggested) * Tune your USE-Flags in make.conf. media-related flags (mp3, flac) should be harmless, if you touch flags that get used in core packages (e.g. in the toolchain) double (or triple) check if you don't do evil things. * fine tune USE-Flags on a per-package-base via /etc/portage/package.use thanks. well what you mentioned was my set up until a week ago. I was on the default profile. Afterwards I installed the other stuff let the portage take care of missing use flags, and to be honest it was the first time in my short linux life that I didnt hate KDE!(I hate the deign choices of GNOME, and I thinks KDE have a good design with bad implementation!) In my opinion it was way better that the KDE profile. I have moved to i3wm and USE=-* and it was not that hard. My concern was the use flags that are better to be enabled globally like bzip2.
Re: [gentoo-user] necessary use flgas
On Wed, Jun 24, 2015 at 01:13:40PM +0200, Alan McKinnon wrote: On 23/06/2015 15:05, behrouz khosravi wrote: Hello everyone. I really like to have control over my machine as much as possible. In this way I will learn a lot, so I am trying to remove all the default use flags and control them manually. Here's some good advice: Don't do that. See below. Nonsense - do that. If your goal is to learn how stuff works and you're already reasonably familiar with C/C++ so you can debug any strange errors that can happen, have fun. Just don't think you'll get any real work done ;). i.e. it might be good to do this in a virtual machine and still have a stable system for work. I just don't know which global use flags are absolutely necessary to the system to make it snappier or secure. That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. ... Put -march=native in CFLAGS Yes. Also, properly setting CPU_FLAGS_X86 is another thing that can speed up software *if* said software supports any special instruction sets. Most normal desktop software like web browsers, email clients, terminals, editors, etc. probably will not get a whole lot of benefit either way, since most of this software is generally not CPU-bound and is instead network/disk bound. In the mornings I primarily use my desktop for reading email and browsing news with firefox (mostly on sites with minimal JavaScript), and I have yet to see my load averages climb higher than maybe 0.5. Any software that does anything requiring lots of math will get a boost from this type of stuff, though; graphics editing, most things in sci-* categories, audio/video transcoding, etc. Alec P.S. Just realized I don't have -march=native in my CFLAGS. Time to rice - could be getting 1% better performance. ;) P.P.S. Also, on 1% better performance: My professor for the compilers class I took used to (maybe still does) work at Google. Apparently Google sees a 1% increase in performance as *the best thing ever*, because it can save them a bunch of money in infrastructure and power. Apparently Google are the ultimate ricers.
Re: [gentoo-user] Best way for good video performance in virtual machines
What will the Qt application be doing? Any of those setups should be sufficient for a typical GUI program. Highest performance would probably be passing a discrete card to the guest... not particularly the smartest move, but it would account for every usecase. Barring that, #2 should have the least amount of overhead, as X11 does the drawing on the system running the server (so host's GPU instead of guest's virtual GPU).
[gentoo-user] Best way for good video performance in virtual machines
Hi out there, assume the following situation: I do have a minimalistic hypervisor running a minimalistic virtual machine (qemu with kvm, qxl and spice). Both systems, hypervisor and VM are able to run X servers and apllications. The graphics card of the hypervisor is connected to a monitor. This monitor only has to show one single standalone QT application which is running on the virtual machine. What will give me more video performance? 1. Running a minimalistic X server on the virtual machine showing the application AND running a minimalistic X server on the hypervisor running remote-viewer or 2. Running a minimalistic X server on the hypervisor and using X11-Forwarding via bridged ethernet Or maybe even someone knows a better solution? Thank you folks! Cheers Ralf
Re: [gentoo-user] Best way for good video performance in virtual machines
Ralf ralf+gen...@ramses-pyramidenbau.de wrote: Hi out there, assume the following situation: I do have a minimalistic hypervisor running a minimalistic virtual machine (qemu with kvm, qxl and spice). Both systems, hypervisor and VM are able to run X servers and apllications. The graphics card of the hypervisor is connected to a monitor. This monitor only has to show one single standalone QT application which is running on the virtual machine. What will give me more video performance? 1. Running a minimalistic X server on the virtual machine showing the application AND running a minimalistic X server on the hypervisor running remote-viewer or 2. Running a minimalistic X server on the hypervisor and using X11-Forwarding via bridged ethernet Or maybe even someone knows a better solution? Thank you folks! Cheers Ralf On my system, qemu-kvm has the best performance with X11 on guest. I've also tested vnc and spice with qxl video driver on guest, but X11 was faster. I can imagine that this depends on the hardware and screen resolution. So maybe on your system an other setup is better. My host GPU is a Radeon Ultimate R7 250, resolution is 3840x2160@60Hz, driver is xf86-video-ati. X11 compositing is deactivated on host as well as on guest. As video card on guest I use vmware (-vga vmware). Screen resolution on guest is 2386x1770. That seems to be the highest possible screen size that is available with this settings. I also tested other guest video drivers (std and cirrus) but vmware was the fastest and also the one with the highest resolution. Before I bought an UHD monitor, I used sdl as qemu display (-display sdl) in borderless fullscreen mode (-no-frame -full-screen). I used the same screen resolution on host and guest (1920x1200) and qemu was running on one of my virtual XFCE desktops. This was very handy. But as I've written before, it seems not possible for me to run qemu with UHD screen size. So fullscreen mode isn't possible any longer and now qemu runs in a window on my desktop (-display gtk). But with none of my setups it was/is possible to run a fullscreen video on the guest vm. Performance is much to slow for that. I've spent some time to make it faster, but I failed. I would be happy for any hints about better video performance with qemu. -- Regards wabe
[gentoo-user] Re: necessary use flgas
On Wed, 24 Jun 2015 07:50:07 -0400 Alec Ten Harmsel a...@alectenharmsel.com wrote: On Wed, Jun 24, 2015 at 01:13:40PM +0200, Alan McKinnon wrote: On 23/06/2015 15:05, behrouz khosravi wrote: I am trying to remove all the default use flags and control them manually. Here's some good advice: Don't do that. See below. Nonsense - do that. If your goal is to learn how stuff works and you're already reasonably familiar with C/C++ so you can debug any strange errors that can happen, have fun. Just don't think you'll get any real work done ;). I don't advocate completely overriding the profile defaults, but I do it myself and have been for years. In terms of getting work done, I estimate it costs me roughly five minutes per month, but with a pretty large standard deviation -- most months it's zero. It helps a lot to keep an eye on the -dev group for any discussion of possibly changing defaults; when a default changes, it almost always means you're going to have to change some flag settings either globally or for a few packages. Be particularly vigilant about changes which might leave you up a creek without a paddle, e.g. @system stuff or networking stuff. It also will help to learn which posters here override profile settings with USE=-*. There are a few of us (sorry, I only have a vague mental list), and the threads we start are likely to have the same issues you may hit. And whenever you ask for help about *anything*, whether you think it has to do with USE flags or not, mention at the start you're overriding the defaults. You'll get more little lectures about not doing that, but it's much better than having a guru who's trying to help find out a dozen posts downthread that he's been on a wild goose chase. Also, properly setting CPU_FLAGS_X86 is another thing that can speed up software *if* said software supports any special instruction sets. And thank $DIETY for app-portage/cpuinfo2cpuflags. Well, I guess it's better to thank mgorny.
Re: [gentoo-user] necessary use flgas
2015-06-24 6:23 GMT-06:00 behrouz khosravi bz.khosr...@gmail.com: Here's some good advice: Don't do that. See below. Oops! I have done it and I am happy so far ! That's a bit of a nonsensical line of thought, as what you think you want doesn't really exist. I think you misunderstood me! for example adding CPU specific flags is a good idea right? I meant something like that. For example is it wise to enable opengl flag globally ? is it helpful to do so? What do you recommend ? DO NOT SET USE=-* As I said before I have done it and I totally recommend it to anyone interested to get a better understanding of user land. I don't see the point of using USE=-* for learning, if you want to really learn, create and overlay and make your own profile, read the developer documentation about it, and do a proper thing, not some clunky mess in /etc/portage. You could very well evaluate the basic profile, and part from there or modify it as you see fit, but making a mess that you wont be able to port easily in case you actually make something you might want to keep and reproduce somewhere else, is not getting a better understanding.
Re: [gentoo-user] necessary use flgas
On Wed, Jun 24, 2015 at 04:53:18PM +0430, behrouz khosravi wrote What do you recommend ? DO NOT SET USE=-* As I said before I have done it and I totally recommend it to anyone interested to get a better understanding of user land. The point with USE=-* is to create your own base profile. E.g. when I upgraded my old Dell Core2 Duo from 32 to 64 bit Gentoo, I dropped USE=-*. I went from... USE_BASE=-* a52 aac bzip2 cxx fortran ncurses netifrc nptl nptlonly nsplugin offensive openssl posix readline ssl threads vim-syntax zlib X dga dri exif ffmpeg flac classic gif intel jpeg mng mp3 mpeg ogg opengl png rtmp theora tiff truetype vorbis xcomposite webm x264 xpm xv xvid xvmc ...to... USE=X apng bindist ffmpeg jpeg png truetype xorg -acl -berkdb -chatzilla -cracklib -crypt -gallium -gdbm -gmp-autoupdate -gstreamer -iconv -introspection -ipc -iptables -ipv6 -libav -llvm -nls -openmp -pam -roaming -sendmail -tcpd -udev -unicode So I went from -* plus add a lot of USE flags to default/linux/amd64/13.0/no-multilib and negating a lot of USE flags. Six of one, half a dozen of the other. -- Walter Dnes waltd...@waltdnes.org I don't run desktop environments; I run useful applications
Re: [gentoo-user] purchasing a dell laptop
On 24/06/2015 19:35, gottl...@nyu.edu wrote: On Wed, Jun 24 2015, Alan McKinnon wrote: One more thing I can add: From observation, I can say that Dell has 2 or more grades of kit they sell: 1. Cheap shit. You find these in supermarkets and Walmart. It's just as crappy as all the other cheap shit around with the same bargain basement price. They call this inspiron 2. Good stuff like the Precision and XPS range. You tend not to find these at Walmart, and have to go to a proper dealer for them, or your workplace has a scheme to provide them. These tend to be big or specialized or super-graphics. Latitude is the line for normal business. Yes, that's the ones :-) -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] purchasing a dell laptop
On Wed, Jun 24 2015, Alan McKinnon wrote: One more thing I can add: From observation, I can say that Dell has 2 or more grades of kit they sell: 1. Cheap shit. You find these in supermarkets and Walmart. It's just as crappy as all the other cheap shit around with the same bargain basement price. They call this inspiron 2. Good stuff like the Precision and XPS range. You tend not to find these at Walmart, and have to go to a proper dealer for them, or your workplace has a scheme to provide them. These tend to be big or specialized or super-graphics. Latitude is the line for normal business. allan
[gentoo-user] Re: repos.conf
Jc García jyo.garcia at gmail.com writes: I hope you find this useful. Yes I did. Sure sounds like an excellent topic for one of our devs to post to planet.gentoo.org about an example of how diversified configurations for our current migratory status on source codes from a wide variety of places could be set up. Surely those leading us on this journey have crossed these bridges? I tend to ignore such needs as long as possible, but with cluster code development, I'm reading about or fixing things from the kernel(s) to functional programming (which until recently I strongly avoided) and every thing possible in between. Some days, I think I'm going crazy..(but actually I'm OK with that idea, really I am). Then I read about something like nitrous.io [1] and I realize that I have a far greater grip on the multifaceted aspects of diverse coding than some of the clowns that are getting rich. Maybe, just maybe, we need to get some 'go daddy girls' to help us get organized ? I guess any documented formal or structured approach, that would allow one to focus on just one problem (piece of code) for a few days/weeks is strictly out of the question.. Please keep the ideas and schema coming, as I only want to solve this code migration organization problem once. Maybe I should just file a (bgo) bug about when do we git nitrous.io on gentoo? (reminds me of the dentist :: a_hole). But he does have an extremely attractive hygienist! James [1] https://pro.nitrous.io/