mimeapps.list problem
hi, When I run a commande like "amixer " a mimeapps.list directory is created, containing: mimeapps.list/pulse mimeapps.list/pulse/a84a11bc8ed348db86feca80371e1e38-runtime -> /tmp/pulse-s6YhteYQrpJA This doesn't occur when running the same command as root, or as an other user. I have a mimeapps.list file in ~/.config, as recommended... Any explanation? best regards, -- Pierre Frenkiel
Re: Debian 9.4 and Aspeed/AST2400
Am Freitag, den 15.06.2018, 16:31 -0700 schrieb Ryan C: > > Hello, > > We are having trouble getting Linux to function properly on SuperMicro > hardware (X10SRI-F + E5-2620v4) which uses the AST2400 for VGA output. We have > tried several "desktop" distributions: Ubuntu, Fedora, Xubuntu, Lubuntu, and > Debian. As well as several versions of Ubuntu. We noticed the issue first > starts with Ubuntu 16.04.2 and is still present in 18.04. In 16.04.1 and below > the video is working fine. Debian is also using a newer kernel which has > problems with this aspeed chip. > > Our issue is that the video output is very very slow at updating. Even the TTY > framebuffer console can be slow at rendering. We open a window or click a menu > and we can see each line of the display being drawn very slowly. Similarly, > the login screens respond very slowly to keyboard input and mouse lags and > jumps around. > > We have narrowed the issue down to specifically to Linux Kernel version 4.9 > and above. Running Kernel version 4.8 or lower has no video issues at all. > > Is this a known issue? Is there a known fix? Maybe we need to roll back the > BIOS? > > There are two possible fixes we found, but they may just be covering up the > issue rather than fixing it: > > 1) set the cpu frequency in Linux to the maximum speed, with "performance" > governor. This helps, but doesn't completely solve the slow video rendering > and tearing. Also, it is not permanent. > > 2) Set the gpu/vga OPROM to "EFI" in the BIOS. This seems to fix the slow > graphics issue on the desktop, but does not help with the slow rendering on > the login screen. > > From what we can tell, the "ast" video driver is being used. Moving the ast > kernel module from one system do another doesn't do anything, but switching > between 4.8 and 4.9 is where everything breaks. > Do you have the same issue in the KVM/HTML5 viewer over IPMI? I use a X11SSM-F witch has also the AST2400. Currently I use the stretch- backports kernel because the EDAC module of kaby lake cpu's is not supported in 4.9 With 4.15 I had some strange wrong color output during boot, the tty color switched to red. This issue was gone gone after the installation of intel-microcode from stretch- backports. Have a nice day, -- Jens Sauer GPG Key: 5C0B0084 Fingerprint: 75F3 6232 1F69 82E8 F5E0 D151 850E 2908 5C0B 0084 signature.asc Description: This is a digitally signed message part
Re: How to diagnose and report a bug
On 2018-06-15, Gene Heskett wrote: >> > >> > Obviously one test to attempt to corner the real culprit in this >> > affair would be to try another editor. How about gedit, for example? >> > Or have you done that already? > > Gedit can be dangerous to your mental health. At least recommend an > editor that does not randomly play 52 pickup with your code when you > save it. gedit looks nice, but I no longer allow it on the premises. > And life is a heck of a lot simpler. > Unsubstantiated, anecdotal, unscientific, cantankerous, off-the-wall, and excessive. The adjectives, in fact, fail me. ;-) As OP navigates in the gtk library world, and the other guy (in the thread I pointed to) successfully used gedit under similar circumstances to the OP's, I just blurted out gedit without further intellectual ado for a simple test of a simple hypothesis. I wasn't asking him or anybody else to change religions or anything. But you do you have to be prudent here. There are so many sensibilities, so many toes, so many ways to go wrong and be chastised in all the innocence of the generosity of your initial impulses. Jesus, what a world. I think I'll retire and watch some of that funny football they play here (Spain-Portugal last night lived up to every and all expectations).
Re: A newer kernel for my genesi smartbook
Hi Sven, Am 14.06.2018 um 21:52 schrieb Sven Joachim: > On 2018-06-13 21:25 +0200, Robert Pommrich wrote: > >> I am a still proud owner of a genesi smartbook. >> >> It is still running wheezy with a kernel in version 2.6.31 with Uboot as >> boot loader. >> >> Someone told me, that the hardware should be supported by now by a >> mainline kernel. > > Unfortunately, it's not - not anymore, that is. Support for the Efika > MX smartbook was added in Linux 2.6.39, and it should work with the > linux-image-mx5 kernel in Wheezy - FSVO work, because the GPU only works > to its potential with proprietary drivers, as on nearly all ARM-based > machines, and I am not sure about other drivers (sound, Wifi…) either. > > The bad news is that support for your machine got ripped out again in > Linux 3.7, as nobody added device tree support for it, and so the Jessie > kernel does not work with it, as mentioned in the release notes[1]. > >> I don't know how to find out, if the hardware is really supported, nor >> how to build a mainline kernel and integrate it with Uboot. > > Install the linux-image-mx5 kernel and the flash-kernel package from > Wheezy, it should know what to do. If the official Wheezy kernel works, > you can upgrade your userland to Jessie (and maybe to Stretch, but note > that systemd in Stretch officially requires Linux 3.12). This would > give you a few extra years. > Thank you very much, I'll follow your instructions and report back. > Good luck, > Sven > > > 1. https://www.debian.org/releases/jessie/armhf/ch02s01.html.en > Robert
Re: [Solved, largely] Re: Latest version of X not starting in buster
On Thu, 14 Jun 2018 11:44 -0400, Cindy-Sue Causey wrote: > On 6/14/18, Siard wrote: > > On Thu, 14 Jun 2018 12:40 +0200, Jörg-Volker Peetz wrote: > > > Siard wrote on 06/13/18 17:28: > > > > > > > > I took a try and upgraded the xserver-*** packages in testing to > > > > version 2:1.20.0 taken from unstable. > > > > It did work; X (started with 'startx') is back, despite the > > > > systemd packages having version 238. > > > > It looks like things will be back to normal after the next > > > > upgrade of X in testing. Anyway, thanks for responding. > > > > > > Which init package do you use? On my systems it's sysvinit-core. > > > > systemd at the moment. sysvinit-core is not installed. > > Just throwing this out there. xinit is mine (on Xfce4). It's always > high on my to-do list immediately after initial debootstrap installs. > > While searching my inbox to see what had come up in the past, > xserver-xorg-legacy had a few mentions related to startx/xinit. One > person, Debian user Jason, experienced success by *removing* it > instead of installing, but I don't know the advisability of going that > route since I've never had to tinker with it myself. :) > > https://lists.debian.org/debian-user/2017/07/msg00640.html > > And user Gareth discovered that their window manager had been removed > during an upgrade: > > https://lists.debian.org/debian-user/2016/08/msg00222.html Jörg-Volker's post prompted me to upgrade xserver-** to 1.20.0 taken from unstable, and everything is fine again. Since upgrading xserver-xorg-core caused a whole bunch of packages with 'xserver' in their name to upgrade with it, I'm somewhat reluctant to downgrade all that to 1.19.6 to sort out what was wrong. Now in some local list, I've heard someone maintain that starting X with startx is very old, untested, deprecated, certainly in a systemd environment. A display manager would control various things concerning your login session in cooperation with systemd. I've never heard of that. Is it true, should one refrain from using startx nowadays?
Re: [Solved, largely] Re: Latest version of X not starting in buster
Siard composed on 2018-06-16 12:06 (UTC+0200): > Now in some local list, I've heard someone maintain that starting X > with startx is very old, untested, deprecated, certainly in a systemd > environment. A display manager would control various things concerning > your login session in cooperation with systemd. > I've never heard of that. Is it true, should one refrain from using > startx nowadays? 'Tis an oft repeated song of openSUSE developers, especially on mailing lists, but see e.g.: https://bugzilla.opensuse.org/show_bug.cgi?id=974660 https://bugzilla.opensuse.org/show_bug.cgi?id=1089620#c1 https://bugs.freedesktop.org/show_bug.cgi?id=88900#c8 https://bugs.freedesktop.org/show_bug.cgi?id=83020#c1 -- "Wisdom is supreme; therefore get wisdom. Whatever else you get, get wisdom." Proverbs 4:7 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/
Re: USB Host-Host cables
On 06/15/2018 09:06 AM, Gene Heskett wrote: On Friday 15 June 2018 06:37:57 Richard Owlett wrote: On 06/14/2018 08:54 AM, Dan Ritter wrote: On Wed, Jun 13, 2018 at 02:50:51PM -0500, Richard Owlett wrote: and now you can access the other side via ssh and scp and whatever. I've never used either "ssh" or "scp". *THEREFORE* I believe I have a reading assignment . After all, that was essentially what I was asking for :} I just picked them as common tools for logging in across IP networks and moving files around. All of the Internet Protocol is open to you, just as with any other ethernet-equivalent interface. OK. I looked at the man pages for both. The focus on secure communication is, in *MY* case, is an undesired complication. Some preliminary web searches helped me to better describe my setup. The two machines are about a cubit apart. The USB Host-Host cable and related software addresses the physical connection. The second machine has no physical means to access the web. Thus removing the need for "secure" communication. Are you saying this second machine has no rj54 for a network cable? <*ROFL*> All computing objects on site physically have such a connector. Are they connected to functional silicon? ?? ??? ;/ IIUC there are two different types of connecting cables. [One is topological equivalent of null modem]. Which are present? WHO knows ;/ All computing objects have physically accessible know functional USB ports. I did my homework. I discovered that USB-ethernet adapters existed. That gave me a choice between *3* failure points [either converter PLUS cable] and *EXACTLY ONE* failure point [a USB master - USB master cable assembly]. If you wish to chuckle, see thread beginning at [https://lists.debian.org/debian-user/2018/05/msg01073.html]. Then that leaves a serial port, or possibly a usb to serial at the 2nd's location. This is a bit unusual, but between a linux or windows pc, and an old trs-80 color computer, has been done. The software is called drivewire, Can Linux outperform Radio Shack? I am not familiar with drivewire. So I just did a web search yielding interesting hits. I suspect they will fill in some gaps in my background. I took my 1st programming course as an E.E. student in the 60's. Since then my background might be termed eclectic ( more accurately random ;) . and it adds 14 i/o channels good for 115k kilobit speeds. This includes a channel dedicated for printing text, which I wrote a script for to pass it off to cups, a channel for general midi use, and which can be handed off to timidity, and 14 more channels which can be used as text terminals, or disk drives, including the ability to "mount" an image of the coco's file system as if it was another floppy drive, but can be a 130 megabyte file. Just reading the DuckDuckGo results page caused me to realize that at an application level I'll want a client-server relationship [I haven't addressed the issue of which should be which.] My search string was [ +"linux" +host" +"ip network" ]. It gave many attractive links. As a major motivation for this project is educational, can you suggest sources or search terms to survey: common tools for logging in across IP networks and moving files around. Drivewire facilitates all that. Its GPL, written in jave so it should be pretty portable. But for machines other than the coco's, will need substitutes written for its output modules. On the coco's it runs over the bit banger port, so any machine with a bit banging serial port would be a candidate to become a drivewire client. Thank you
Re: NitroShare doesn't seem to work with Buster
On 2018-06-15 11:39 PM, Felmon Davis wrote: On Fri, 15 Jun 2018, Gary Dale wrote: On 2018-06-14 12:19 AM, Felmon Davis wrote: On Wed, 13 Jun 2018, Gary Dale wrote: Or on another subject, does anyone know if mtp will ever work? I don't know; there has been discussion of using: gmtp go-mtpfs jmtpfs I've used jmtpfs (requires) java; I've used one of the others but forget which, whichever, it worked alright. I use sshdroid on the phone to and sshfs on the laptop now. I believe krusader is relevant in kde environments but no direct familiarity. f Thanks. It took some doing but I seem to have sshfs working now. To add the phone to /etc/fstab, I'm using root@192.168.1.11:/ /home/garydale/GarysPhone fuse.sshfs _netdev,port=,user,idmap=user,transform_symlinks,identityfile=/home/garydale/.ssh/id_rsa,default_permissions,uid=1000,gid=1000 0 0 Which is similar to what I found by googling sshfs for /etc/fstab settings. Unfortunately it lets me mount the phone but I have to be root to umount it. Any suggestion on fixing that? does fusermount -u mount-point work? f. Thanks. It did.
Re: Disable/skip/ignore !! Configure the package manager dialog in preseed
On Fri, Jun 15, 2018 at 12:02 PM, Ravi Roy wrote: > Hi, > > [!!] Configure the package manager > >Cannot access repository > The repository on xx.xx.xx.xx could not be accessed, so its updates will > not be made available to you at this time. You should investigate it later. > Commented out entries for xx.xx.xx.xx have been add /etc/apt/sources.list > > > > > Does somebody have faced this issue? i did googling but did not get a > workaround/fix for this > > Just want to mention here for reference if somebody face this issue - it can be resolved by adding the following entry into preseed. d-i apt-setup/services-select multiselect Thank you.
Re: [Solved, largely] Re: Latest version of X not starting in buster
On Sat 16 Jun 2018 at 12:06:24 +0200, Siard wrote: > Now in some local list, I've heard someone maintain that starting X > with startx is very old, untested, deprecated, certainly in a systemd > environment. A display manager would control various things concerning > your login session in cooperation with systemd. > I've never heard of that. Is it true, should one refrain from using > startx nowadays? Old? Indeed. Prehistoric even. :) That's its charm. Untested? Not at all; many test it all the time Deprecated? By whom? And why? Refrain from using it? It works well on Debian. -- Brian.
Re: [Solved, largely] Re: Latest version of X not starting in buster
On Sat, 16 Jun 2018 20:19:15 +0100 Brian wrote: > On Sat 16 Jun 2018 at 12:06:24 +0200, Siard wrote: > > > Now in some local list, I've heard someone maintain that starting X > > with startx is very old, untested, deprecated, certainly in a > > systemd environment. A display manager would control various things > > concerning your login session in cooperation with systemd. > > I've never heard of that. Is it true, should one refrain from using > > startx nowadays? > > Old? Indeed. Prehistoric even. :) That's its charm. > > Untested? Not at all; many test it all the time > > Deprecated? By whom? And why? > > Refrain from using it? It works well on Debian. +1 I use startx manually (my Stretch system by default boots to a terminal), but don't use a desktop or environment, just Openbox as a window manager and a single LXPanel. Plus, I use sysvinit as the init. Although the systemd libraries are still there for dependencies. Easier than trying to have a systemd-free system. No problems so far. B
Re: [Solved, largely] Re: Latest version of X not starting in buster
On Sat, 16 Jun 2018 20:19:15 +0100 Brian sent: > On Sat 16 Jun 2018 at 12:06:24 +0200, Siard wrote: > > > Now in some local list, I've heard someone maintain that starting X > > with startx is very old, untested, deprecated, certainly in a > > systemd environment. A display manager would control various things > > concerning your login session in cooperation with systemd. > > I've never heard of that. Is it true, should one refrain from using > > startx nowadays? > > Old? Indeed. Prehistoric even. :) That's its charm. > > Untested? Not at all; many test it all the time > > Deprecated? By whom? And why? > > Refrain from using it? It works well on Debian. > After contemplation, my reply is: I use only startx with verbosity. So I can see if the boot lags or stalls or stops and where, on all Debian installs. Have used an alternative GUI start system [display manager] many years ago, for a few days, just to see what it did. I suppose they still have a twirling thingy on the monitor while the machine boots? Way back in the mists of time, when windows was still used, changed windows to start verbose as well. However, it was hardly verbose. Very secretive that O/S. Startx works a treat. May it be long available. But may there also be choice, for those who wish to go another way. Be well, Charlie -- Registered Linux User:- 329524 *** If anything in nature strikes you as ugly, you are not appreciating its diversity. ..anon *** Debian GNU/Linux - Magic indeed. -
Re: [Solved, largely] Re: Latest version of X not starting in buster
On Sun, 2018-06-17 at 08:41 +1000, Charlie S wrote: > > > After contemplation, my reply is: > I suppose they still have a > twirling > thingy on the monitor while the machine boots? > > Be well, > Charlie > Sorry, there is no longer enough time for one of those to be drawn ;) systemd-analyze Startup finished in 2.382s (kernel) + 6.090s (userspace) = 8.473s graphical.target reached after 6.081s in userspace
Re: How to diagnose and report a bug
On 16/06/18 01:30, Thomas Schmitt wrote: > Hi, > >> In pcmanfm's address bar, [...] >> [...] >> Any recomendations on how to diagnose and report this bug? > > Since there is a package with name "pcmanfm", there is also a tracker page > for it > https://tracker.debian.org/pkg/pcmanfm > > Registered maintainer contact is the mailing list > pkg-lxde-maintain...@lists.alioth.debian.org > There seems to be life > > https://alioth-lists.debian.net/pipermail/pkg-lxde-maintainers/2018-June/thread.html I wouldn't trust that either it will be around long, or that anyone useful is reading it - it was supposed to be turned off on 14 April: https://alioth-lists.debian.net/pipermail/pkg-lxde-maintainers/2018-April/001964.html links to: https://alioth-lists.debian.net/pipermail/pkg-lxde-maintainers/attachments/20180402/4b96edbc/attachment.ksh > So you could ask there for advise, or you could file a bug for package > "pcmanfm" by following the description at > https://www.debian.org/Bugs/Reporting > This bug report will appear on the mailing list. I'd choose the bug report. I've cc'd the uploaders, with a view to getting the maintainer email set correctly. Richard signature.asc Description: OpenPGP digital signature
mdraid on SSD - use scterc?
Hi all, I've got a new server with 2 mdraid RAID 1 arrays - one on spinning disks, the other on SSDs. The spinners are set by default to use ERC; the SSDs are not, but appear to support it. Should I enable it on the SSDs? Or are they not prone to failure in such a way that it would be useful? Thanks, Richard signature.asc Description: OpenPGP digital signature
