Page 5 of 6

Re: Porteus-v3.2.1 bug reports

Posted: 13 Dec 2016, 02:13
by fulalas
It seems that xorg module has some useless/undesirable folders/files:

/usr/share/themes
/usr/lib64/xmms
/usr/bin/mplayer*
/etc/mplayer*

/var/spool
/var/cache
/var/run
/var/lib

/home
/home/guest/.mplayer*
/home/guest/.config/gnome-mplayer*

/root
/root/.config/gnome-mplayer*

*this should come inside 003

Re: Porteus-v3.2.1 bug reports

Posted: 13 Dec 2016, 22:51
by Ed_P
Off Topic but 3.2.1 related.

It would be nice if the savedatspaceck.sh script or something similar got added to the final release for those who use save.dat files.

http://forum.porteus.org/viewtopic.php? ... 614#p47636

Re: Porteus-v3.2.1 bug reports

Posted: 14 Dec 2016, 23:51
by fulalas
brokenman, although cheatcode 'kmap=br' maps my keyboard layout perfectly, it seems that it breaks some key shortcuts inside LXQt, like for example: ALT+SHIFT+TAB (to cycle backwards through opened apps). I'm wondering why exactly :)

Re: Porteus-v3.2.1 bug reports

Posted: 15 Dec 2016, 02:09
by brokenman
@Ed_P
I have added the script to my tree.

@fulalas
I am unable to support LXQt due to a lack of time. I do remember LXQt has always suffered problems with the language applet on the taskbar too. It is one of a few reasons I decided not to go with it.

Re: Porteus-v3.2.1 bug reports

Posted: 16 Dec 2016, 16:24
by tome
If wireless connection is lost nm-applet icon in tray is not refreshed and shows connected signal still. In porteus v3.1 nm-applet works correctly.

Re: Porteus-v3.2.1 bug reports

Posted: 16 Dec 2016, 17:58
by brokenman
Which desktop? Cinnamon seems OK. My connection drops frequently and applet updates immediately.

Re: Porteus-v3.2.1 bug reports

Posted: 16 Dec 2016, 18:11
by Jack
brokenman wrote:Which desktop? Cinnamon seems OK. My connection drops frequently and applet updates immediately.
I lost my connection yesterday and had to reboot to get it back. Then when I woke up the connection was gone again. I was in Firefox both times when it happen. At lease I could save all my info. I'm using Mate 64bit.

Re: Porteus-v3.2.1 bug reports

Posted: 17 Dec 2016, 12:35
by moricedu69
Hi, while I was on porteus 3.1, everything worked on my computer, I could read HD Video fluent and play on 3D games
but since I am on porteus 3.2.1, I have problem with video because it is extremly slow with only 2 images per second. Wholes componant of my PC are more than classic since CPU and graphic card are from Intel.
This problem occure even with fresh install, in always fresh and baseonly, no root, no change, nomagic cheatcode.
3D acceleration work fine (tested on supertuxkart) but booth video and scrolling any page are slow with 2 images per seconds.

Can you help me to have back fluent video and 2D on my computer please.

lspci send me back http://pastebin.com/7xDrAL72

lspci -k send me back
http://pastebin.com/hq5EqEF8
Problem occure even with awaysfresh.

cheers

Re: Porteus-v3.2.1 bug reports

Posted: 17 Dec 2016, 16:07
by Bogomips
@ moricedu69

Which desktop are you trying?

Re: Porteus-v3.2.1 bug reports

Posted: 17 Dec 2016, 19:00
by brokenman
Your system uses the newer amdgpu kernel driver. While it is newer, I have found it to be quite slow for some tasks.

Re: Porteus-v3.2.1 bug reports

Posted: 20 Dec 2016, 05:20
by fulalas
Not exactly a bug, more an inconsistence. On LXQt module thread we've found out that modules created by context menu are larger than the ones created by dir2xzm script. So I dig a little bit and found that dir2xzm script sets the mksquashfs block size to 256kb (which is the fastest to decompress according to this test), but context menu script doesn't set any block size, so it's running on its default which is 128kb. So my suggestion is to change the context menu script to include -b 256K parameter. Sounds fine for you? :)

Also, files inside usb stick /porteus/optional are loaded during the boot regardless their extension ('base' and 'modules' folder only load xzm files). Is this intentional?

Re: Porteus-v3.2.1 bug reports

Posted: 20 Dec 2016, 07:20
by ncmprhnsbl
fulalas wrote: Also, files inside usb stick /porteus/optional are loaded during the boot regardless their extension ('base' and 'modules' folder only load xzm files). Is this intentional?
can't say i've noticed that...
when you say 'loaded' what do you mean? loop mounted?
what, other than a .xzm can be loop mounted? a file? some other squashfs?
nothing from /optional should be loaded at boot without an explicit cheatcode, like load= or extramod=

Re: Porteus-v3.2.1 bug reports

Posted: 20 Dec 2016, 15:58
by fulalas
ncmprhnsbl wrote:can't say i've noticed that...
when you say 'loaded' what do you mean? loop mounted?
what, other than a .xzm can be loop mounted? a file? some other squashfs?
Sorry, I wasn't clear. During tests I usually put a _ in the end of xzm file modules so they won't be loaded during boot. This works when dealing with 'base' and 'modules' folders, but with 'optional' folder a .xzm_ file is loaded anyway. I didn't try to put a non-xzm file inside it and check what happens.
ncmprhnsbl wrote:nothing from /optional should be loaded at boot without an explicit cheatcode, like load= or extramod=
That's not the problem. I'm using vga_detect cheatcode, so Nvidia driver module inside 'optional' folder is loaded, but, as I said, also .xzm_ files :P

Re: Porteus-v3.2.1 bug reports

Posted: 20 Dec 2016, 19:10
by Ed_P
fulalas wrote:During tests I usually put a _ in the end of xzm file modules so they won't be loaded during boot.
Try the approach I use, put a "y" between the "xz". As in "xyzm". It's possible whatever is looking for ".xzm" finds it with your ".xzm_" name.

Re: Porteus-v3.2.1 bug reports

Posted: 20 Dec 2016, 21:47
by Bogomips
^

Code: Select all

guest@porteus:~$ ls tst
a.xzm  b.xzm_  c.xyzm
guest@porteus:~$ find tst/*.xzm
tst/a.xzm
Gives desired result. 8)

Hiowever for vga_detec different kettle of fish

Code: Select all

	if [ $NV ]; then
	    echo $i"nvidia-$NV.xx driver will be activated -"
	    echo $i"if present in $PTH/optional folder"
	    find $PTH/optional -name "nvidia-$NV*" 2>/dev/null >>/tmp/modules
	else
	    echo $i"latest nvidia driver will be activated -"
	    echo $i"if present in $PTH/optional folder"
	    find $PTH/optional -name "nvidia-*" 2>/dev/null | egrep -v '96.43.|173.14.|304.' >>/tmp/modules
	fi
Glancing quickly, looks like only nvidia- modules would get loaded irregardless of xzm_.