Bug in Porteus 3.1 EXIT cheatcode

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.
rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Bug in Porteus 3.1 EXIT cheatcode

Post#1 by rchase » 21 Jan 2016, 08:00

Having once booted with EXIT:, I now must always include it or on boot Porteus displays the error:

Failed to set new theme

Failed to rename file 'root/.cache/xfce4-notifyd-theme.rc.41PVBY' to '/root/.cache/xfce4-notifyd-theme.rc': g_rename() failed: Input/output error

I carefully reinstalled (for the umpteenth time) after first encountering this error; I now attribute it to the use of EXIT:. Is there a fix?


P.S. I modified the layout using panel preferences, but I have not changed/installed any themes per se.

P.P.S. My laptop hangs on reboot after this error too.

User avatar
Ed_P
Contributor
Contributor
Posts: 8341
Joined: 06 Feb 2013, 22:12
Distribution: Cinnamon 5.01 ISO
Location: Western NY, USA

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#2 by Ed_P » 21 Jan 2016, 20:53

rchase wrote:Having once booted with EXIT:, I now must always include it or on boot Porteus displays the error:
When you used EXIT did you include where the save.dat file was? Something like this:

Code: Select all

changes=EXIT:/porteus3.1/changes/porteussave.dat
Ed

User avatar
brokenman
Site Admin
Site Admin
Posts: 6105
Joined: 27 Dec 2010, 03:50
Distribution: Porteus v4 all desktops
Location: Brazil

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#3 by brokenman » 21 Jan 2016, 21:01

I am not sure you fully grasp the module concept of Porteus. If you include NO changes= cheatcode (a.k.a fresh mode) then porteus will always boot ok. Fresh mode is often used for trouble shooting for this reason. Booting with changes= cheatcode and then not being able to reboot means that somewhere your changes are still being loaded, and that the problem lies within those changes ... or that you have a module in the modules or base folder containing the problem files.
How do i become super user?
Wear your underpants on the outside and put on a cape.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#4 by rchase » 21 Jan 2016, 22:08

I suppose I wasn't explicit enough. I am trying to get Porteus to the point I can try to upgrade the kernel to see whether that will fix the video on my Kangaroo. I installed it to sda2 (an ext4 partition) on the hard drive of my laptop, added exfat support, and kodi -- everything worked. I was using changes=/dev/sda2/porteus, then I changed it to changes=EXIT:/dev/sda2/porteus; Porteus booted fine. I didn't want to try to do make_iso or upgrade the kernel in RAM, so I changed the cheatcode back to /dev/sda2/porteus; now when Porteus boots, it always displays the above message (with a different four characters in place of "PVBY" each time). If I put EXIT: back in changes=, the error message is not displayed. The error does not seem consequential, but I am not sure of that, and it is exasperating to see it every time I boot. As I said, I've been through this before; this time I am certain that the only change which occasions the error was having used EXIT: once.

I want to use Porteus very much more than I am interested in diagnosing it.


P.S. There may be hundreds of millions of others who would feel the same way if they tried it.

User avatar
brokenman
Site Admin
Site Admin
Posts: 6105
Joined: 27 Dec 2010, 03:50
Distribution: Porteus v4 all desktops
Location: Brazil

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#5 by brokenman » 22 Jan 2016, 00:04

Try booting into fresh mode and deleting the folder /mnt/sda2/porteus/changes/root/.cache
Backup the folder first if you feel it is important.
added exfat support, and kodi

I am not sure what this means.
How do i become super user?
Wear your underpants on the outside and put on a cape.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#6 by rchase » 22 Jan 2016, 02:23

brokenman wrote:Try booting into fresh mode and deleting the folder /mnt/sda2/porteus/changes/root/.cache
Backup the folder first if you feel it is important.
added exfat support, and kodi

I am not sure what this means.
I downloaded the modules for exfat and kodi through USM, placed them in /porteus/modules, and they work.

With regard to /mnt/sda2/porteus/changes/root/.cache, I have no feelings one way or the other; my working assumption is that files that are part of the OS are important unless documentation states otherwise. I find your advice very cryptic -- will there be some untoward consequences of deleting it, or not? I had thought you would have definite knowledge.

I moved .cache to the Desktop (from the installation on USB) and Porteus then booted from /sda2 without EXIT: in changes= and without error -- thanks. Upon trying to reboot, Porteus hung. I found viewtopic.php?f=117&t=4220 in which Michele13 had the same problem a year ago. Fanthom pronounced it kernel-related and eschewed responsibility for the kernel (employed by the offical release of Porteus) -- he also stated that one solution would be to use 64-bit Porteus, but that would seem to be incorrect, since I am using it and it hangs on reboot. Three months later, Michele13 found a post from 2009 (http://linux.koolsolutions.com/2009/08/ ... -restarts/) suggesting that passing the parameter reboot=acpi to the kernel would likely resolve the problem; she expressly stated that no change to the kernel was required.

Incomprehensibly, the thread was then marked as "SOLVED", even though Michele13 had just reported that Porteus wouldn't reboot and that she had found a solution (for herself) not requiring a different kernel. I have just run through the reboot parameters for 64-bit machines and found that only reboot=efi consistently allows me to reboot my laptop through the Action Buttons. Porteus' aging official release has a bug rebooting; it may be one that can be resolved by using a different kernel than the one in the official release, but Michele13 and I found a means of rebooting without changing the kernel. Regardless of whether there is a problem in the supplied kernel, Porteus has a problem if it won't reboot when users click the Reboot button, and this problem has apparently persisted for a long time without being properly addressed. MS has found the means to consistently reboot computers; I suggest that Porteus should too -- Fanthom's dismissive reply to the effect that Porteus' developers aren't responsible for problems with the kernel completely ignores the fact that the vast majority of users (i.e. end-users) have no idea what a kernel is and are completely reliant on the one developers chose -- as for the airy advice to try another, which?

Please reopen Michele13's bug thread from a year ago -- the bug that prevents Porteus from rebooting has not been resolved, nor is it restricted to x86 machines, nor can it be blamed solely on the kernel (since it can be resolved without changing the kernel). What is the point of all the work on an easy-to-use, GUI-based OS if its developers won't acknowledge so significant a bug or fix it? At the risk of belaboring the point, end-users expect to be able to reboot their computers without research and troubleshooting; I am not an end-user, and I expect the same. Where is this project going? Nemesis? What kernel do you expect to offer average users next? v3.18.25? v3.19? v4.2.8? All have been suggested here as possible alternatives to v3.17. Is there a timetable for the next release? As someone far from an end-user but primarily interested in using rather than endlessly tinkering with Porteus, from the information on the official website and the forum, I cannot divine what I should do to get it to work on a new machine with no video other than to upgrade the kernel -- to something.


P.S. In reference to the bug in EXIT:, will this be corrected in the next release, or is it to be considered "solved" because it no longer troubles me?

User avatar
brokenman
Site Admin
Site Admin
Posts: 6105
Joined: 27 Dec 2010, 03:50
Distribution: Porteus v4 all desktops
Location: Brazil

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#7 by brokenman » 22 Jan 2016, 03:36

When posting a problem please give as much information about the problem as you can. At the very least some system specs would be nice. Perhaps putting your system specs in your signature would be a good idea. I saw in another thread that you are running an old system with a savage graphics card, but now you mention passing an efi parameter to the kernel, so I am confused as to what system you are talking about in this thread. Knowing at least the age of the machine can help a lot in trouble shooting. Right now I have very little insight to work with.

I am a little confused by what you say you pass to the kernel. Is it reboot=acpi or reboot=efi? You say that passing reboot=efi to the kernel allows you to "consistently" reboot. Does this not mean that your problem is resolved? Is there a problem with adding this to the porteus.cfg file in the boot folder and continuing to use Porteus?
I find your advice very cryptic ... I had thought you would have definite knowledge.
Nothing too cryptic about asking you to delete the directory. Simple and straight forward. The definitive answer is that you will be deleting the upper layer of a union file system branch that will give way to the underlying read only branch made up of the base Porteus modules which are know to be in good working order. This will allow you to boot your machine. Is there really any point in adding this technical knowledge to my reply?
Incomprehensibly, the thread was then marked as "SOLVED", even though Michele13 had just reported that Porteus wouldn't reboot and that she had found a solution (for herself) not requiring a different kernel.
The original poster clearly stated that her problem was solved. No point leaving an open thread if the OP will no longer participate because they no longer have the problem.
Fanthom's dismissive reply to the effect that Porteus' developers aren't responsible for problems with the kernel completely ignores the fact that the vast majority of users (i.e. end-users) have no idea what a kernel is and are completely reliant on the one developers chose -- as for the airy advice to try another, which?
If I know Fanthom, then his reply was the best he could offer at the time. If it was within his control I am certain he would have given MORE of his valuable free time, of which he volunteers without any return, to help resolve the issue. At the time of the release the Porteus kernel was known to work on the majority of hardware existing at the time. Newer computers emerge and the kernel no longer supports some hardware. In this case an update of the kernel is required.

We usually follow a slackware release schedule and as you may or may not know slackware has not released anything since the last porteus release. At present we (the Porteus community) are debating whether to move the base from slackware to a manjaro rolling release base.
Please reopen Michele13's bug thread from a year ago -- the bug that prevents Porteus from rebooting has not been resolved, nor is it restricted to x86 machines, nor can it be blamed solely on the kernel (since it can be resolved without changing the kernel).
Sorry, I'm not going to open an old closed thread that was clearly marked as solved based on the reply of the OP. You allude to that in the quote above when you say "since it can be resolved without changing the kernel". You have started a new thread (this one) where the problem can be investigated further if you so choose.
What is the point of all the work on an easy-to-use, GUI-based OS if its developers won't acknowledge so significant a bug or fix it?
Bug acknowledged. As I stated above, fixing it requires releasing a new porteus. We are at a crossroad at the moment so a new slackware based Porteus may be some time. I am not sure if you intended to be abrasive, but I find this comment rather gross and somewhat uninformed.
Where is this project going? Nemesis?
We already have a beta release of Nemesis available which I suggest you try to see if the problem is in fact resolved with an updated kernel and packages.
What kernel do you expect to offer average users next?
4.4 LTS (as of today's date)
Is there a timetable for the next release?
At present I can't give a timetable. I am the sole developer of Porteus and until I decide which direction to take I don't want to put any dates out there. I am working hard everyday testing the beta version and also keeping the slackware version updated in the wings. All of this is done in my free time (read: 10pm-2am) so I would not expect a new release this month.
I cannot divine what I should do to get it to work on a new machine with no video other than to upgrade the kernel -- to something.
I suggest you try the nemesis release to see if the updated kernel and package base resolves your problem. If so then you can continue testing it and offering feedback until:
a) It matures and becomes our default
b) An updated slackware base is released.

Failing both of these options I would suggest finding a rolling release distro that can support your hardware and computing goals. If tinkering under the hood is not your thing then might I suggest Ubuntu?
MS has found the means to consistently reboot computers; I suggest that Porteus should too
Another somewhat abrasive comment. Well since you went there, that hasn't been my experience at all. Upgrading to Windows 10 for me was a prolonged and tortuous experience that resulted in a non-booting machine that now runs Porteus, which I have found to "consistently" reboot on 99% of our users' computers. You and Michele13 are the only users I know of that have had this problem. I would say those are pretty good odds. It's all relative.
How do i become super user?
Wear your underpants on the outside and put on a cape.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#8 by rchase » 22 Jan 2016, 07:29

Thanks for your reply. I do not mean to be abrasive, but I am exasperated at my slow progess in getting a Porteus version where I want it -- on several machines (references to which I have tried to keep straight in these posts). I did not realize that you were the sole developer; I will bear that in mind. I understand that no one involved is being paid for their labors and that developing an entire OS is a huge project. I appreciate your efforts and those of your collaborators. With my rather limited understanding of the lay of the land of Linux distributions, it seems to me that Porteus should be one of the leading ones, but that it still has too many failure modes -- not that Ubuntu is much better, if it is. I came here because I am unimpressed with Ubuntu; it seems to me that you are at least within striking distance of its acceptability and useability by end-users (and obviously ahead of it and others in efficiency, which, as someone who first learned to program on a Honeywell H1642, I prize greatly), and I hope that you prioritize making Porteus (yet) more user-friendly because it seems so close to becoming a generally preferable OS.

You have resolved the problem with EXIT: on my old Toshiba Satellite A135-S4656 (an early 64-bit machine) -- what I found cryptic was the suggestion to back up the .cache folder if I felt it was important; I am no Linux expert and not in a position to judge the importance of OS folders -- you are! I was afraid that problems such as this and with hanging on reboot were just being churned in the forums and not necessarily coming to the attention of the developer. I tried the various options for the reboot parameter for 64-bit machines: cold, acpi, pci, force, efi. I tried successive reboots with each of these options, altering the cheatcodes at boot, issuing a reboot= command from the terminal, or by modifying porteus.cfg -- the laptop did reboot a couple of times using acpi and one of the other options, but when I attempted a subsequent reboot, Porteus hung. I found that reboot=efi seems to work consistently. I consider my problems with this machine resolved.

I also want to install Porteus to my Kangaroo (with an Atom X5-Z8500; I posted about it in Newbie Questions and aus9 is assisting me -- http://forum.porteus.org/viewtopic.php?f=81&t=5393), but found that the video did not work; xorg won't start. Aus9 referred me to a forum page on kernel upgrades (viewtopic.php?f=75&t=4788&hilit=neko, but it offers no explanation as to the distinctions between them (v3.18.25 and v4.2.8) or why one might be preferable. In Michelle13's thread you mentioned upgrading to v3.19 and now you reference v4.4 LTS. I think that now that I am confident in the installation on my laptop, I can use it to remaster the Porteus 3.1 iso and apply the upgrade tools to it. I have not compiled from source but I am willing to try. I downloaded Nemesis 3.4 and 3.5 and booted each from USB on my Kangaroo; both booted to the desktop, but my graphical cursor was invisible. What do you suggest would be my best course of action to get some version of Porteus working on my Kangaroo (which I suppose to be typical of many new low power computers): upgrade Porteus 3.1's kernel to v3.18.25 or to v4.2.8, try to compile another from source, try to fix the graphical cursor in Nemesis 3.4 or 3.5 (how?), or something else?

The third computer on which I want to run Porteus is an old Celeron M with an S3 ProSavageDDR; I think I have reached an impasse, but I will follow up with aus9 and Bogomips in Newbie Questions (http://forum.porteus.org/viewtopic.php?f=81&t=5403).

Thanks again; I trust you have given and will continue to give due consideration to Porteus' potential for end-users -- it seems so great to me. May you find time and assistants enough to make Porteus a truly universal OS.


P.S. I've got to go here: comparing Windows 10 and Porteus might be taken as complimentary too -- please take it that way. If Porteus came to be considered as useable as that other OS, why should not Porteus overtake it? Manufacturers should be seeking you out and offering a lot of money for an OS that could beat out Windows and Chrome for functionality on cheaper hardware. Do licensing requirements for Porteus' components preclude this?

User avatar
brokenman
Site Admin
Site Admin
Posts: 6105
Joined: 27 Dec 2010, 03:50
Distribution: Porteus v4 all desktops
Location: Brazil

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#9 by brokenman » 22 Jan 2016, 09:19

I can upload an updated Porteus kernel for you tonight. I am running it on my test machine with Porteus and it seems to function ok so far. This will require some remastering of the ISO which I can walk you though without a problem. If this works then it would be your best solution. If not then Nemesis is your best bet and resolving the mouse problem. I will leave it in the hands of aus9 and Bogomips, two excellent troubleshooters that help a lot with the evolution of Porteus.

If you have keyboard functionality on your machine without a mouse, then try opening a terminal (ALT + F2) xfce4-terminal and killing and restarting the psmouse module.
modprobe -r psmouse && modprobe psmouse
Also be aware that v3.4 Nemesis uses libxinput while v3.5 uses evdev. This means that you can tweak the input device with xinput and synclient repsectively (for synaptics touchpads). I will look a little more at the commands that may help when I get home from work.

Lastly, thanks for your compliments. Hopefully we can make Porteus more user friendly, and up to date in the coming months. Our community is tightly knit and very friendly so don`t be afraid to ask for help. While we would like to be as user friendly as windows (which I personally use and admire greatly) we ARE running linux. Perhaps one day linux will rule the desktop, but I think we are a long way off yet.
How do i become super user?
Wear your underpants on the outside and put on a cape.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#10 by rchase » 22 Jan 2016, 11:27

Thanks; where should I look for this upload? I just tried booting Nemesis 3.5 again on my Kangaroo. In response to Alt + F2, Nemesis opened a "Run" dialog box, not the terminal. I am not having a problem with the mouse per se (so far as I can tell, it is working and right and left clicking appropriately); the graphical cursor is simply invisible. I managed to open the terminal because its launcher highlights when the cursor is over it. Issuing "modprobe - r psmouse && modprobe psmouse" returned "modprobe: FATAL: Module psmouse is builtin." I'll try Nemesis 3.4 tomorrow in case it behaves differently and look forward to your test build -- Porteus which might/does work on the Atom, I gather.

P.S. I retract what I said regarding having resolved the power problems on my Satellite; after having rebooted successfully many times using the Action Buttons and with "reboot=efi" in porteus.cfg, it stopped doing so. I tried to suspend it and it hung; thereafter, reboot fails as well. I will try to use one of the newer kernels and/or Nemesis and see what happens. On the assumption that this due to this kernel's interaction with this hardware, it shouldn't be that uncommon a situation (only cropping up on Michelle13's and my computers). Do you have an estimate of the number of current users of Porteus? The number of unique downloading IPs serves as an upper bound, but I suppose you may have a better number than that.
Last edited by rchase on 23 Jan 2016, 01:04, edited 2 times in total.

User avatar
Ed_P
Contributor
Contributor
Posts: 8341
Joined: 06 Feb 2013, 22:12
Distribution: Cinnamon 5.01 ISO
Location: Western NY, USA

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#11 by Ed_P » 22 Jan 2016, 18:43

rchase wrote: I downloaded Nemesis 3.4 and 3.5 and booted each from USB on my Kangaroo; both booted to the desktop, but my graphical cursor was invisible.
Try this suggestion from kalo86. I had the same problem and this worked for me.

http://forum.porteus.org/viewtopic.php? ... =15#p40780
Ed

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#12 by rchase » 22 Jan 2016, 23:48

Ed_P wrote:
rchase wrote: I downloaded Nemesis 3.4 and 3.5 and booted each from USB on my Kangaroo; both booted to the desktop, but my graphical cursor was invisible.
Try this suggestion from kalo86. I had the same problem and this worked for me.

http://forum.porteus.org/viewtopic.php? ... =15#p40780
Thanks -- I will.

User avatar
brokenman
Site Admin
Site Admin
Posts: 6105
Joined: 27 Dec 2010, 03:50
Distribution: Porteus v4 all desktops
Location: Brazil

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#13 by brokenman » 23 Jan 2016, 00:17

I've uploaded the updated 64bit kernel and 000-kernel.xzm module here. It is kernel 4.4 LTS. Let me know if you need the 32bit kernel and I will compile it tomorrow.

I would also recommend downloading the latest xf86-input-synaptics package from slackware-current
http://packages.slackware.com/

If you need help remastering your ISO to include the above files please let us know.
How do i become super user?
Wear your underpants on the outside and put on a cape.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#14 by rchase » 23 Jan 2016, 01:25

brokenman wrote:I've uploaded the updated 64bit kernel and 000-kernel.xzm module here. It is kernel 4.4 LTS. Let me know if you need the 32bit kernel and I will compile it tomorrow.

I would also recommend downloading the latest xf86-input-synaptics package from slackware-current
http://packages.slackware.com/

If you need help remastering your ISO to include the above files please let us know.
Excellent! Thanks; I'll try this directly. If the 32-bit kernel should be suitable for older machines, sure; I'll take it too.
Besides the old Celeron M with S3 ProSavageDDR I'm trying to debug with aus9 et alia at viewtopic.php?f=81&t=5403 (mass, or at least personal confusion at this point), I have an even older eMachines T1220 with a 1.2GHz Celeron and 384MB of SRAM which runs 3.1 xfce quite nicely from a frugal installation; I'd throw 4.4 on that even if the savage machine never works. I know someone else who has only a machine of similar vintage and I'd like to install the best 32-bit Porteus I can on it too.

Several xf86-input packages are installed, but xf86-input-synaptics is listed as a downgrade in USM -- I am using my Satellite laptop to prepare the image to install to my Kangaroo; the Satellite has a touchpad -- the Kangaroo, a mouse.

I do need help with one aspect of the iso: the make_iso script did not include an efi folder in the iso; how can I make the iso UEFI-installable? I gather from one of Francois' posts that grub2 can be installed from Porteus -- is that an alternative means of getting Porteus to boot on an UEFI machine since grub2 supports it? The Kangaroo only supports UEFI booting and has a GPT-partitioned eMMC, which is where I'd like to put Porteus.
Last edited by rchase on 23 Jan 2016, 01:46, edited 2 times in total.

rchase
Shogun
Shogun
Posts: 282
Joined: 10 Jan 2016, 17:26
Distribution: Porteus 5.0 rc3 xfce
Location: Denver, Colorado

Re: Bug in Porteus 3.1 EXIT cheatcode

Post#15 by rchase » 23 Jan 2016, 01:33

(extra post in case you read my original reply without subsequent emendations; I do need to know how to prepare an UEFI-bootable iso (having upgraded its kernel) before I can test it on my Kangaroo -- I will upgrade the iso and see how it runs on the Satellite while I wait your reply).

Post Reply