Posts: 1,445
skidoo
Joined: 09 Feb 2012
#166
antiX-14-a3-RV_386-full

a few belated observations:

-- a session autostarted clipboard manager utility would be a wecome amenity for"full" version

-- may have been mentioned previously: unlike the other desktops, conky is not displayed for"rox+icewm"

-- unsure which combination of factors triggers it, but during some sessions, the conky blinks/flickers with each (1sec) update.

-- noticeable"slow repainting" when dragging application windows on the desktop (default boot seetings, nouveau driver)
Last edited by skidoo on 23 Oct 2014, 13:52, edited 1 time in total.
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#167
Some of my responses may be inadequate. Perhaps someone else can do a better job.
skidoo wrote:-- a session autostarted clipboard manager utility would be a wecome amenity for"full" version
I don't know if one comes with full or not. If so, you can have it start in all desktops by editing /etc/desktop-session/startup.
-- may have been mentioned previously: unlike the other desktops, conky is not displayed for"rox+icewm"
Is this consistent and repeatable? Did you try increasing the STARTUP_DELAY in /etc/desktop-session/desktop-session.conf?
-- bootmenu F1 help mentions 1600x1280.
I've tried F7 S1600 and *1600 and 1600 (label without S and without asterisk)
and each results in an error message"undefined video mode number: 347" (or 31F)
I also tried adding vga=839 to the bootline and wind p with same (non-fatal) error. It falls back to 800x600 and life goes on...
Basically vga kind of sucks and we may have to move away from it like most others have done a while ago. The vga parameter for 1600x1200 works inside of Virtual Box. One of the main problems with vga is that except for a handful of resolutions, it is not standardized and different manufacturers have different vga codes (or no codes) for their native resolution.

Now that I finally have a machine (a laptop I bought this past summer specifically for testing) that can use a modeset video driver (that handles the console as well as the X display) I think the best solution is to load the video module ASAP and go back to a default with no vga parameter. I don't know if we should include the"Console" (vga=xxx) boot menu or not. If you (and others) find it useful then we should include it even though the vga= mechanism is somewhat broken as you have discovered.

You say below that you have the nouveau driver. If you can run in the default video mode and if we load the video driver ASAP then you won't have to worry about vga= because nouveau will give you framebuffers on the consoles with the maximum resolution available.

Thoughts or suggestions?
-- unsure which combination of factors triggers it, but during some sessions, the conky blinks/flickers with each (1sec) update.
Are you able to fix this by restarting the session? It is an option in the exit menu. You might find a solution doing a Search(conky flashing). Other people have had similar problems.
-- noticeable"slow repainting" when dragging application windows on the desktop (default boot seetings, nouveau driver)
Can you run the 64-bit version on the same hardware? Dolphin_oracle has found that the 3.15 kernel used in the 32-bit versions has a lot of problems. They all cleared up when he used the 64-bit version with the 3.17 kernel. Maybe that would fix most of your display problems too. I wish I had thought of this before I addressed all of your specific problems.

I am now realizing that we should probably stop testing the 32-bit version. We know there are a lot of bad bugs with the 3.15 kernel (although they may be caused by our config). The obvious solution is to move the 32-bit version to the 3.17 kernel. Ironically, the reason anticapitalista went with the 3.15 kernel on 32-bit was because the vga=xxx parameter stopped working correctly with the 3.17 kernel.

Should we drop the vga=xxx boot menu in order to fix all of these other problems in the 32-bit version?

I appreciate all of the testing you do. Which reminds me, I know you like to make your LiveUSBs on a fat32 filesystem. Are you able/willing to add an ext3/4 partition for persistence? You can specify a separate persistence device with pdev=xxx plab=yyy or puuid=zzzz. There are also ways of specifying what type of device. I think"persist=root!,home,usb" will cause us to scan all usb devices for /antiX/rootfs and /antiX/homefs files.
Posts: 325
male
Joined: 04 Nov 2011
#168
BitJam wrote:@male, do plain old /sbin/reboot and /sbin/poweroff work with systemd? Are they deprecated? The systemctl program does not exist on non-systemd systems. I hope we are not going to need to add code to detect if systemd is running in order to poweroff or reboot from within a script.
The question irritated me ...
Commands were executed in the console of the RD-live-iso and they work.

I see in the hard-installation of the Control Center exitantix converts still the old commands in the.

It work both variants in the RD. __{{emoticon}}__
Posts: 1,445
skidoo
Joined: 09 Feb 2012
#169
re: autostarted clipboard manager
Mentioned as an"observation", not a request. While testing, I try to maintain a"what would a new user do/think?" point of view.
Where (Ubuntu? WinXP?) are new users currently coming from?
If clipboard functionality is missing, or seems broken,"hassling with it, troubleshooting it" might be a dealbreaker for the user.

re: conky missing from (only) one desktop session flavor
Absent from only"rox+icewm" (a noticeable inconsistency) and, yes, is still absent after restarting a session and/or using"alternate desktops" changer.
So, yes, it's consistently repeatable and... a new user likely has no concept of"mumbletygook . . . /etc/desktop-session/desktop-session.conf"

re: bootmenu F1 help mentions 1600x1280
Tested using a (borrowed) larger, higher res monitor. Realized when I posted this might be an isolated quirk, yet might serve to reaffirm any similar reports.


In reading what I've just typed, and mentally taking a step back, I'm wondering whether my testing regimen / mindset is misguided (and/or has become moot).
Locally, I've been pandering"Linux" to WinXP emigrants... now, among my acquaintances, few"potential new users" remain.
Folks here are not interested in (enthusiastic about) emigrating to a new O/S. Instead they're migrating away from"desktop PC" computing, period.
-=-
In any event, I do think continuing to support"breathing new life into older PCs" (wrestling VGA/VESA issues, etc) is a noble goal.
Can you run the 64-bit version on the same hardware?
. . .
we should probably stop testing the 32-bit version. We know there are a lot of bad bugs with the 3.15 kernel
. . .
Are you able/willing to add an ext3/4 partition for persistence? You can specify a separate persistence device . . . scan all usb devices for /antiX/rootfs and /antiX/homefs files.
1) Yes.
2) Amen.
3) To what end? Is this a workaround, an alternative to updating kernel from within a live session?
Last edited by skidoo on 24 Oct 2014, 23:21, edited 1 time in total.
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#170
skidoo wrote:3) To what end? Is this a workaround, an alternative to updating kernel from within a live session?
Creating homefs and rootfs would be magically faster in ext3/4. That's all.
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#171
skidoo wrote:-- a session autostarted clipboard manager utility would be a wecome amenity for"full" version
Full already has the following clipboard programs:

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://manpages.ubuntu.com/manpages/trusty/man1/xclip-copyfile.1.html"
    linktext was:"xclip-copyfile, xclip-cutfile, xclip-pastefile"
    ====================================

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://ftp.x.org/pub/X11R6.7.0/doc/xclipboard.1.html"
    linktext was:"xclipboard"
    ====================================

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://www.linuxinsider.com/story/72853.html"
    linktext was:"clipit"
    ====================================
I found them by doing find / -name"*clip*" so there may be others without 'clip' in the name.

Do any of these do what you want? If so can you tell me how it should be called in the startup file? I'm also interested in the resources used but I can probably measure that if you tell me how to start it.

Do you think we should add a button in the control centre for this?
Posts: 1,062
Dave
Joined: 20 Jan 2010
#172
BitJam wrote:
skidoo wrote:-- a session autostarted clipboard manager utility would be a wecome amenity for"full" version
Full already has the following clipboard programs:

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://manpages.ubuntu.com/manpages/trusty/man1/xclip-copyfile.1.html"
    linktext was:"xclip-copyfile, xclip-cutfile, xclip-pastefile"
    ====================================

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://ftp.x.org/pub/X11R6.7.0/doc/xclipboard.1.html"
    linktext was:"xclipboard"
    ====================================

  • ========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
    url was:"http://www.linuxinsider.com/story/72853.html"
    linktext was:"clipit"
    ====================================
I found them by doing find / -name"*clip*" so there may be others without 'clip' in the name.

Do any of these do what you want? If so can you tell me how it should be called in the startup file? I'm also interested in the resources used but I can probably measure that if you tell me how to start it.

Do you think we should add a button in the control centre for this?
This should be addressed by enabling xdg autostart in desktop-session.conf but we need to figure out how to enable / disable this via the antix=mlxd grub line cheat to coincide with the"d" portion of the cheat for dbus. If it is not enabled and disabled via that cheat then we either need to remove the"d" portion and allow dbus to start and have xdg autostart enabled or leave dbus disabled and xdg autostart disabled. Otherwise there will be a bunch of dbus errors on startup particularly with wicd.
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#173
Unless things have changed recently, you need dbus enabled in order to start slim and X-windows. If you look at the isolinux.cfg file on the iso, you will see that only the command line menu entry disables dbus. Therefore anything that runs within X-windows is free to assume that dbus is enabled.
Posts: 1,445
skidoo
Joined: 09 Feb 2012
#174
A button in ControlCentre doesn't address the"it should just work" expectation.
Without the clipboard manager running, user copies (e.g.) from a file in leafpad, closes the file, opens another file and attempts to paste. No joy.

Placing a clipit & line in each startup file (e.g. /etc/skel/.fluxbox/startup )
might make it more easily found/removed (vs xdg/autostart) for anyone interested in doing so.

How much ram overhead does clipit process add?
Less than the amount consumed by root window wallpaper image buffer.
Less than the amount consumed by conky.
Posts: 94
james c
Joined: 23 Apr 2014
#175
Interesting adventure installing AntiX-14-a3-RD-x64 on this old dual core.

Booted up live, correct 1440x900 resolution good internet and sound. No problem. Installed, put grub in the root of the partition (I chainload everything from good old reliable Legacy Grub) and rebooted. Logged in,got to the desktop and the computer immediately froze up with a totally distorted multi-color display. Hard poweroff. Repeated for good measure.Same result.

Added the good old nomodeset cheatcode and successfully booted to a working,but incorrect, 1024x768 desktop.Good old nouveau. __{{emoticon}}__

Guess tomorrow I'd better install the Nvidia driver.(It's 3 am, too late tonight)

Code: Select all

james@antix1:~
$ inxi -F
System:    Host: antix1 Kernel: 3.17.0-antix.1-amd64-smp x86_64 (64 bit) 
           Desktop: IceWM 1.3.8 
           Distro: antiX-14-a3-RD_x64-full Killah P 11 October 2014 
Machine:   Mobo: ASRock model: N68-S UCC
           Bios: American Megatrends v: P1.70 date: 03/03/2011
CPU:       Dual core AMD Athlon 64 X2 5200+ (-MCP-) cache: 1024 KB 
           Clock Speeds: 1: 2712 MHz 2: 2712 MHz
Graphics:  Card: NVIDIA C61 [GeForce 7025 / nForce 630a]
           Display Server: X.Org 1.16.1 drivers: fbdev,nouveau (unloaded: vesa)
           Resolution: 1024x768@76.00hz
           GLX Renderer: Gallium 0.4 on llvmpipe (LLVM 3.4, 128 bits)
           GLX Version: 3.0 Mesa 10.2.6
Audio:     Card NVIDIA MCP61 High Definition Audio driver: snd_hda_intel 
           Sound: ALSA v: k3.17.0-antix.1-amd64-smp
Network:   Card: NVIDIA MCP61 Ethernet driver: forcedeth
           IF: eth0 state: up speed: 100 Mbps duplex: full 
           mac: 00:25:22:61:55:fc
Drives:    HDD Total Size: 1000.2GB (0.3% used)
           ID-1: /dev/sda model: WDC_WD10EZEX size: 1000.2GB
Partition: ID-1: / size: 49G used: 2.5G (6%) fs: ext4 dev: /dev/sda9 
Sensors:   System Temperatures: cpu: 31.0C mobo: N/A 
           Fan Speeds (in rpm): cpu: N/A 
Info:      Processes: 119 Uptime: 10 min Memory: 375.6/3704.9MB 
           Client: Shell (bash) inxi: 2.1.28 
james@antix1:~
My intentions are to test as is for a while and then try to purge the evil SystemD from the system.
Posts: 94
james c
Joined: 23 Apr 2014
#176
Sleep is overrated. Much better with the correct Nvidia driver.Used smxi...no problem.

Code: Select all

james@antix1:~
$ inxi -F
System:    Host: antix1 Kernel: 3.17.0-antix.1-amd64-smp x86_64 (64 bit)
           Desktop: IceWM 1.3.8
           Distro: antiX-14-a3-RD_x64-full Killah P 11 October 2014
Machine:   Mobo: ASRock model: N68-S UCC
           Bios: American Megatrends v: P1.70 date: 03/03/2011
CPU:       Dual core AMD Athlon 64 X2 5200+ (-MCP-) cache: 1024 KB 
           clock speeds: max: 2712 MHz 1: 2712 MHz 2: 2712 MHz
Graphics:  Card: NVIDIA C61 [GeForce 7025 / nForce 630a]
           Display Server: X.Org 1.16.1 driver: nvidia
           Resolution: 1440x900@59.89hz
           GLX Renderer: GeForce 7025 / nForce 630a/integrated/SSE2
           GLX Version: 2.1.2 NVIDIA 304.123
Audio:     Card NVIDIA MCP61 High Definition Audio driver: snd_hda_intel
           Sound: ALSA v: k3.17.0-antix.1-amd64-smp
Network:   Card: NVIDIA MCP61 Ethernet driver: forcedeth
           IF: eth0 state: up speed: 100 Mbps duplex: full
           mac: 00:25:22:61:55:fc
Drives:    HDD Total Size: 1000.2GB (0.3% used)
           ID-1: /dev/sda model: WDC_WD10EZEX size: 1000.2GB
Partition: ID-1: / size: 49G used: 2.9G (7%) fs: ext4 dev: /dev/sda9
Sensors:   System Temperatures: cpu: 33.0C mobo: N/A
           Fan Speeds (in rpm): cpu: N/A
Info:      Processes: 121 Uptime: 1:05 Memory: 454.1/3704.9MB
           Client: Shell (bash) inxi: 2.2.15 
james@antix1:~
Now to get my swap detected.
Posts: 1,062
Dave
Joined: 20 Jan 2010
#177
BitJam wrote:Unless things have changed recently, you need dbus enabled in order to start slim and X-windows. If you look at the isolinux.cfg file on the iso, you will see that only the command line menu entry disables dbus. Therefore anything that runs within X-windows is free to assume that dbus is enabled.
Okay then we have another bug somewhere unless this is only specific to my machine. I start live, default options, let it boot, attempt to run wicd and it asks for the root password and says that it can not connect with dbus and find the interfaces. I have always had this problem live and with previous versions it was remedied by removing d from antix = mlxd. Maybe there is a different remedy now because slim calls desktop - session rather than the old xsession string?
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#178
Dave wrote:I have always had this problem live and with previous versions it was remedied by removing d from antix = mlxd. Maybe there is a different remedy now because slim calls desktop - session rather than the old xsession string?
What is the output of"cat /proc/cmdline"?

As I said before, every bootloader main menu entry (except one) has"antiX=MLX". If you start with the"Command Line Install" entry then, yes, you will need to start dbus manually. Likewise, if you are using your own bootloader config that has"antiX=MLXD" then you will have to start dbus manually.

I wonder if you are able to start X just fine without dbus running or if there is still a problem even when"antiX=MLX" is used.

Anyway, here is the code from live-disable-services that parses the antiX= cheat code:

Code: Select all

CMDLINE=${CMDLINE:-$(cat /proc/cmdline)}
for param in $CMDLINE; do
    case"$param" in
        antiX=*|aX=*) antiX_param=${param#*=}; MUST_RUN=true ;;
                lean)        LEAN=true;        MUST_RUN=true ;;
                mean)        MEAN=true;        MUST_RUN=true ;;
            Xtralean)   XTRA_LEAN=true;        MUST_RUN=true ;;
              nodbus)     NO_DBUS=true;        MUST_RUN=true ;;
                db++)   DEBUGGING=true                       ;;
    esac
done

if ["$antiX_param" ]; then
    
    # If the ## expression matches then the resulting string length is zero
    ["${antiX_param##*[lL]*}" ] ||      LEAN=true
    ["${antiX_param##*[mM]*}" ] ||      MEAN=true
    ["${antiX_param##*[xX]*}" ] || XTRA_LEAN=true
    ["${antiX_param##*[dD]*}" ] ||   NO_DBUS=true
fi
Of course, you would only need the dbus parts.

ISTM that if dbus is needed then maybe desktop-session (or slim?) should just start it regardless of the boot parameters. I don't fully understand what is going on so use my advice with care.
Posts: 2,238
dolphin_oracle
Joined: 16 Dec 2007
#179
Okay then we have another bug somewhere unless this is only specific to my machine. I start live, default options, let it boot, attempt to run wicd and it asks for the root password and says that it can not connect with dbus and find the interfaces.
after i supplied the root password the first time, wicd has not asked for it again, as long as I'm running w/ persistence enabled of course. and I don't get the dbus error after the first time either. (64bit antix a3)
Posts: 1,308
BitJam
Joined: 31 Aug 2009
#180
skidoo wrote:Placing a clipit & line in each startup file (e.g. /etc/skel/.fluxbox/startup )
might make it more easily found/removed (vs xdg/autostart) for anyone interested in doing so.

How much ram overhead does clipit process add?
Less than the amount consumed by root window wallpaper image buffer.
Less than the amount consumed by conky.
Can your provide a short paragraph explaining how to use clipit for our"what's new" page? If we are going to add it by default then people should be told about it and be told how to take advantage of it.