Porteus 5.0 RC2 bug reports

Please reproduce your error on a second machine before posting, and check the error by running without saved changes or extra modules (See FAQ No. 13, "How to report a bug"). For unstable Porteus versions (alpha, beta, rc) please use the relevant thread in our "Development" section.
User avatar
Ed_P
Contributor
Contributor
Posts: 5782
Joined: 06 Feb 2013, 22:12
Distribution: 4.0 & 5.0rc2 Cinnamon 64 ISOs
Location: Western NY, USA

Porteus 5.0 RC2 bug reports

Post#166 by Ed_P » 03 Oct 2020, 00:14

ncmprhnsbl wrote:
02 Oct 2020, 23:49
well there's (part of) your problem ;) : deactivate the porteus-scripts-xorg-20200920-noarch.xzm (updated scripts) because this is overwriting the cinnamon update-menu script..
if you want to use porteus-scripts-xorg-20200920-noarch.xzm: extract it; remove /opt/porteus-scripts/xorg/update-menu; and recompress it.
Ok, progress. :good: I'll just rebuild the iso using your 002-xorg-20200920.xzm file. Thank you ncmp. :beer:
Ed

fulalas
DEV Team
DEV Team
Posts: 1605
Joined: 26 Oct 2016, 15:34
Distribution: Porteus
Location: Brazil

Porteus 5.0 RC2 bug reports

Post#167 by fulalas » 03 Oct 2020, 02:07

ncmprhnsbl wrote:
02 Oct 2020, 05:28
i did a little experiment: activated palemoon> doesn't appear in the menu>opened the .desktop file in an editor>did nothing>saved unchanged file>now it appears in the menu..
now i think the 'GTK' addition to Categories doesn't have an effect, simply resaving the .desktop, somehow "wakes up" cinnamon's menu..
Does it happen in any other DE? I tried on Xfce and it works just nice. Can you confirm?
Ed_P wrote:
02 Oct 2020, 23:25
I wonder if fulalas's file attributes change is tied to this.
What you mean?

User avatar
Ed_P
Contributor
Contributor
Posts: 5782
Joined: 06 Feb 2013, 22:12
Distribution: 4.0 & 5.0rc2 Cinnamon 64 ISOs
Location: Western NY, USA

Porteus 5.0 RC2 bug reports

Post#168 by Ed_P » 03 Oct 2020, 03:52

fulalas wrote:
03 Oct 2020, 02:07
What you mean?
There was a series of postings a few weeks back where you noted that moving files from one location to another were getting their permissons changed in rc2, not like they were in rc1. I don't remember the specifics. But ncmp believes he has found the solution to this menu problem.
Ed

donald
Full of knowledge
Full of knowledge
Posts: 1692
Joined: 17 Jun 2013, 13:17
Distribution: Porteus 3.2.2 XFCE 32bit
Location: Germany

Porteus 5.0 RC2 bug reports

Post#169 by donald » 03 Oct 2020, 22:45

XFCE -- Not a bug (?) but aesthetics
Truncated Icon text -- look:
Image

If you hate that, like I do, create:
~/.config/gtk-3.0/gtk.css
Open gtk.css and paste in:

Code: Select all

/* icon text not truncated */
XfdesktopIconView.view {
   -XfdesktopIconView-ellipsize-icon-labels: 0;
}
log out/in ..much better,isn't it?
Image
B)

fulalas
DEV Team
DEV Team
Posts: 1605
Joined: 26 Oct 2016, 15:34
Distribution: Porteus
Location: Brazil

Porteus 5.0 RC2 bug reports

Post#170 by fulalas » 04 Oct 2020, 00:00

@donald, try to put some long filenames in the desktop and you'll this solution won't work well.

donald
Full of knowledge
Full of knowledge
Posts: 1692
Joined: 17 Jun 2013, 13:17
Distribution: Porteus 3.2.2 XFCE 32bit
Location: Germany

Porteus 5.0 RC2 bug reports

Post#171 by donald » 04 Oct 2020, 03:01

@fulalas
Yup, long file names that would not be displayed due to the cut-off text make sense - like some
screenshots, all labeled screensho..lol. :)

However, here is my current gtk.css and screenshot

Code: Select all

/* icon text not truncated */
XfdesktopIconView.view {
   -XfdesktopIconView-ellipsize-icon-labels: 0;
   -XfdesktopIconView-cell-spacing:0;
   -XfdesktopIconView-cell-padding:0;
   -XfdesktopIconView-cell-text-width-proportion: 2.5;
}
Image

This controls the width of the icon text field:
-XfdesktopIconView-cell-text-width-proportion: 2.5;
Less than 2.5 the text field returns to 1 line

You can restart the desktop with:

Code: Select all

xfdesktop -Q && sleep 1 && xfdesktop &
if you want to experiment with various values.

sams
Legendary
Legendary
Posts: 31
Joined: 05 Jan 2011, 18:53
Location: Alaska

Porteus 5.0 RC2 bug reports

Post#172 by sams » 05 Oct 2020, 00:44

Trying 5.0rc2 just now, a few comments:

I had a couple of partitions with similar labels (space, ssdspace) and encountered a problem with the cheatcode "changes=EXIT:LABEL:space/p50/porteus/" ; (space=sda6, ssdspace=sdb4) On boot porteus says this:

setting up directory for changes
sh: sdb4: unknown operand
mkdir: can't create directory /mnt/sda6: File exists
rmdir: /mnt/sda6: Resource busy
sh: sdb4/p50/porteus/: unknown operand
^ couldn't find LABEL:space/p50/porteus/. Correct your from= cheatcode

So it works if I relabel space->bigspace such that the string is not a substring of ssdspace. Anyway it's alarming when it says it's trying to rmdir /mnt/sda6 which is never what I would want, though it failed to do anything unfortunate.

And on another note, my laptop runs hotter on porteus than some other distros because
/sys/class/scsi_host/host*/link_power_management_policy is set to max_performance; laptop runs much cooler when this is set to med_power_with_dipm. Here's a reference:
https://insanity.industries/post/sata-p ... nsumption/

More impressions later, otherwise so far so good.
My best always to all y'all.

User avatar
Ed_P
Contributor
Contributor
Posts: 5782
Joined: 06 Feb 2013, 22:12
Distribution: 4.0 & 5.0rc2 Cinnamon 64 ISOs
Location: Western NY, USA

Porteus 5.0 RC2 bug reports

Post#173 by Ed_P » 05 Oct 2020, 13:20

Hi sams.

I boot from ISOs and have used the changes= cheatcode for a long time. My code looks like this: changes=EXIT:$linux_folder/changes/porteussave.dat where $linus_folder is set to: set linux_folder=/porteus5.0. I don't specify the drive because I search for it with: search -f $iso --set=root which sets root to it. $iso being the name of the ISO I am booting: set iso="/porteus5.0/rc2ISOModules/Porteus-CINNAMON-v5.0rc2-x86_64.iso" On an installed system I've used: search -f /boot/syslinux/vmlinuz --set=root.

I think I've used the changes=EXIT:UUID= option on a USB drive before but I can't find an example to post.
sams wrote:
05 Oct 2020, 00:44
And on another note, my laptop runs hotter on porteus than some other distros because
/sys/class/scsi_host/host*/link_power_management_policy is set to max_performance; laptop runs much cooler when this is set to med_power_with_dipm. Here's a reference:
https://insanity.industries/post/sata-p ... nsumption/
Thank you for this suggestion. But when I checked my files they were both set to med_power_with_dipm.

Added in 1 hour 39 minutes 27 seconds:
This thread on UUIDs and LABELs may be helpful. porteus-v4.0-x86_64.cfg vs porteus.cfg cheatcodes? and having stuff on different partitions
Ed

sams
Legendary
Legendary
Posts: 31
Joined: 05 Jan 2011, 18:53
Location: Alaska

Porteus 5.0 RC2 bug reports

Post#174 by sams » 05 Oct 2020, 18:27

Hi Ed,
I have porteus 5rc2 (with changes) working fine but the bug is that changes=LABEL should only match an exact label; instead code matches "*LABEL*" and spits up if results return multiple matches. I sometimes prefer lables to UUIDs as they are more descriptive and all my filesystem labels are unique. Anyway this is a documented cheatcode that should work properly.

Stock 5rc2 has my 2 SSD drives in max_performance mode, shedding much heat full time. I added a udev rule to my changes folder and they run cool. I will guess you don't have such a udev rule added but perhaps your drive firmware or other kernel interaction has them defaulting to the cooler mode without user changes?

Cheers
-s

User avatar
Ed_P
Contributor
Contributor
Posts: 5782
Joined: 06 Feb 2013, 22:12
Distribution: 4.0 & 5.0rc2 Cinnamon 64 ISOs
Location: Western NY, USA

Porteus 5.0 RC2 bug reports

Post#175 by Ed_P » 05 Oct 2020, 23:21

sams, in that you have the max_performance mode problem and I don't implies the problem is either DE related, I use Cinnamon, or install related in that I boot ISOs directly.
Ed

User avatar
ncmprhnsbl
DEV Team
DEV Team
Posts: 2702
Joined: 20 Mar 2012, 03:42
Distribution: 5.0rc1-64bit all-DE+more
Location: australia
Contact:

Porteus 5.0 RC2 bug reports

Post#176 by ncmprhnsbl » 06 Oct 2020, 04:11

sams wrote:
05 Oct 2020, 00:44
And on another note, my laptop runs hotter on porteus than some other distros because
hi sams, can you try for me: (without the udev rule you added):

Code: Select all

chmod  +x /etc/rc.d/rc.tlp
/etc/rc.d/rc.tlp start
then check /sys/class/scsi_host/host*/link_power_management_policy
i'm guessing since we have TLP included, we should probably start it.. ie. make it executable and add it to rc.services
sams wrote:
05 Oct 2020, 18:27
the bug is that changes=LABEL should only match an exact label; instead code matches "*LABEL*" and spits up if results return multiple matches.
yeah, looks like in initrd.xz> finit > locate function > the simple grep should probly be grep -w (whole word regex)
Forum Rules : https://forum.porteus.org/viewtopic.php?f=35&t=44

sams
Legendary
Legendary
Posts: 31
Joined: 05 Jan 2011, 18:53
Location: Alaska

Porteus 5.0 RC2 bug reports

Post#177 by sams » 06 Oct 2020, 06:47

ncmprhnsbl wrote:
06 Oct 2020, 04:11
chmod +x /etc/rc.d/rc.tlp
Well right you are, that does set the drives to the cool mode. Thank you!
Label grep fix seems important and would be appreciated.

Other quirks I've noticed and I'll try to research a little: I added a "sam" user to my changes, and when I use the cheat "login=sam" it doesn't log in but the "=sam" is the default on the login pane.

I noted that tlp sets my thinkpad battery thresholds which is cool. My own battery threshold setting software doesn't work because it relies on acpi_call which isn't present. lsmod shows thinkpad_acpi which I'm not yet familiar with (just looking now, looks like it was patched to support thresholds through sys interface)

I also note that my default porteus trackpoint settings are way too slow and heavy for trackpoint use. I don't mind poking these to suitable values myself but I wonder if porteus would like to use speedier defaults? I'm currently setting sensitivity to 254 and speed to 242. I'll also note that I don't believe modern thinkpads support sensitivity setting anymore (grrr, actual IBM trackpoints have gone bye bye...)

I appreciate all the help
cheers!

User avatar
ncmprhnsbl
DEV Team
DEV Team
Posts: 2702
Joined: 20 Mar 2012, 03:42
Distribution: 5.0rc1-64bit all-DE+more
Location: australia
Contact:

Porteus 5.0 RC2 bug reports

Post#178 by ncmprhnsbl » 06 Oct 2020, 08:58

sams wrote:
06 Oct 2020, 06:47
Other quirks I've noticed and I'll try to research a little: I added a "sam" user to my changes, and when I use the cheat "login=sam" it doesn't log in but the "=sam" is the default on the login pane.
yep, there's little bug that slipped through in a couple of DEs: in /etc/rc.d/rc.4 :

Code: Select all

login=`egrep -o "^login=[^ ]+*" /etc/bootcmd.cfg | cut -dn -f2`
nologin=`grep -o "^nologin" /etc/bootcmd.cfg`
should be:

Code: Select all

login=`egrep -o "login=[^ ]+*" /etc/bootcmd.cfg | cut -d= -f2`
nologin=`grep -o "nologin" /etc/bootcmd.cfg`
Forum Rules : https://forum.porteus.org/viewtopic.php?f=35&t=44

User avatar
francois
Contributor
Contributor
Posts: 5891
Joined: 28 Dec 2010, 14:25
Distribution: xfce plank porteus nemesis
Location: Le printemps, le printemps, le printemps... ... l'hiver s'essoufle.

Porteus 5.0 RC2 bug reports

Post#179 by francois » 07 Oct 2020, 01:57

Code: Select all

guest@porteus:~$ sudo su
Password: 
root@porteus:/home/guest# updatedb
bash: updatedb: command not found
root@porteus:/home/guest# 
It seems that mlocate is no more part of the small sofware. Is it the case? I use to like that one a lot to find missing libraries equivalents.

Thanks.
Prendre son temps, profiter de celui qui passe.

User avatar
ncmprhnsbl
DEV Team
DEV Team
Posts: 2702
Joined: 20 Mar 2012, 03:42
Distribution: 5.0rc1-64bit all-DE+more
Location: australia
Contact:

Porteus 5.0 RC2 bug reports

Post#180 by ncmprhnsbl » 07 Oct 2020, 09:27

francois wrote:
07 Oct 2020, 01:57
It seems that mlocate is no more part of the small sofware. Is it the case?
looking at 4.0, seems it had slocate(basicly the same thing), not sure if mlocate was omitted purposely.. probly wouldn't hurt to include it...
in the meantime, getpkg mlocate will fetch it from current ..
Forum Rules : https://forum.porteus.org/viewtopic.php?f=35&t=44

Post Reply