openbox on older computer ?

[Texstar]

I think that driver is too old to build with our existing kernel.

[T6]

and the video card

[Yankee]

and the video card

Well I'm thinking perhaps an entirely new video card.
Seen some for $20.

[T6]

together with a new mainboard, cpu and ram?   

[Taco.22]

hello,

thank you for your reply djohnston.

i have tried also the vesa driver and just to be sure i have installed it this morning again. i notice that loading webpages is going slower than with a nv setting and if i am scrolling down it is much more wobbier.

i also had a dozen freezes, so the frustrating is going on....................

i have download the nvidia driver from the website you suggested. but how do i go from here ?

i have to exit the X server first to have the driver installed............since i am new to this all , i do not know the commands for this,perhaps someone can type the process and can give it a try.

i know the command to run the nvidiadriver for installation.

regards,

Willem

I recently hit a similar issue with OB on a 2.4ghz P4 with an FX5200 card.  It also had the shutdown/reboot issue.  It became a nightmare to use - unexpected crashes.  Installing the nvidia driver just broke X.  It has worked for me before but not this time.  In the end I downgraded the kernel to 2.6.37.4-pclos1.bfs.  Then the browsers started doing the "aw snap" thing, so I downgraded the flash plugin.  So far all has been fine.  Even the shutdown problem disappeared.

It may be that the latest kernel is too much for the older hardware.  I know I have had all sorts of issues with 2.6.38... in the past.  It may also just be certain machines that don't get on with it.  Curiously, I have a 2.6ghz P4 running OB Bonsai with the latest kernel and have had no problems.  Work that one out!!  Anyway, hope this helps a little.    

[djohnston]

It may be that the latest kernel is to much for the older hardware.  I know I have had all sorts of issues with 2.6.38... in the past.  It may also just be certain machines that don't get on with it.  

That is curious. 2.6.38.8-pclos3.bfs is what is running on my 1999 Dell PentiumIII. Haven't had a problem with it.

[Taco.22]

It may be that the latest kernel is to much for the older hardware.  I know I have had all sorts of issues with 2.6.38... in the past.  It may also just be certain machines that don't get on with it.   

That is curious. 2.6.38.8-pclos3.bfs is what is running on my 1999 Dell PentiumIII. Haven't had a problem with it.

A little while back when Melodie was revamping OB we hit some issues that seemed to centre on the 2.6.38 kernel.  However it wasn't universal, and there seemed to be no rhyme or reason.  It just seemed to affect some 'puters, and mostly not others.  I just seem to have hardware that's problematic .  In fact the previous install of OB wouldn't even accept 2.6.38.  One suggestion was hyperthreading in some P4s, but it seems to affect P3s as well!

http://www.pclinuxos.com/forum/index.php/topic,93164.msg782125.html#msg782125 and http://www.pclinuxos.com/forum/index.php/topic,93031.msg782017.html#msg782017 are earlier posts on a similar topic.

[djohnston]

A little while back when Melodie was revamping OB we hit some issues that seemed to centre on the 2.6.38 kernel.

Yeah, I was part of that thread. I experienced the reboot issue on the same PentiumIII I am referring to here. I believe it was 2.6.38.8-pclos1 rather than 2.6.38.8-pclos3. In any case, I really don't believe it was the kernel at fault. And I don't want to open that can of worms again.

[melodie]

A little while back when Melodie was revamping OB we hit some issues that seemed to centre on the 2.6.38 kernel.

Yeah, I was part of that thread. I experienced the reboot issue on the same PentiumIII I am referring to here. I believe it was 2.6.38.8-pclos1 rather than 2.6.38.8-pclos3. In any case, I really don't believe it was the kernel at fault. And I don't want to open that can of worms again.

Hi,
We can open it and close it at same time. I found 2 methods to solve that one : GDM has a custom.conf file, which is written in when you personalize the commands with the "Edit commands" button. To do that, you can edit at login, or with "dbus-launch gdmsetup" command (as user, then you are prompted for the password). Right down in tab "general" there is a button "Edit commands", and the first option is to edit shutdown commands. There is a long line with several commands. Adding this at the beginning does help on some machines : /sbin/halt;

Then the file /etc/X11/gdm/custom.conf contains this line:
HaltCommand=/sbin/halt;/usr/bin/poweroff;/sbin/poweroff;/sbin/shutdown -h now;/usr/sbin/shutdown -h now

and the other solution is to add "acpi=off" at the end of the kernel line in the menu.lst. One or the other, or both fixed it in my P4, and I intend to deliver the next Openbox versions with the additional /sbin/halt; in the gdm setup.

Else, I hope Muppet will find a good solution for this issue.

Regards,
Mélodie

[Taco.22]

Yes, playing with worms can get messy .  It's just funny how this problem randomly seems to pop up.  Putting the "/sbin/halt;" in the gdm setup didn't solve my restart problem, but downgrading the kernel did, although I don't know if the addition to the code is relevant now.  I suppose I could edit it out and see if it makes a difference.  The other issue was random crashes, similar to Muppet.  Again, a kernel downgrade seemed to fix that.  I just wish I new why!!  In the meantime the other P4 is puttering along happily, although it does run Bonsai, and I don't know if there are differences between the OBs that may be relevant.    

[EDIT - apparently the difference between 2.6.38...1 and 3 is this - http://www.pclinuxos.com/forum/index.php/topic,94300.msg793383.html#msg793383]

[Muppet]

Hello,

well, Muppet is getting very frustrated..............steam coming out of his ears......................

have tried several options with the nvidia driver. i got it installed. then xorg.conf file edit driver nvidia to it ect.

every time i reboot i become the message that the Xserver can not load...............

i think i give in on this one. everybody has helped me enough with advise en considerable effort of time.

perhaps another card is the solution to overcome this as the solution....i don't know yet

Willem

[melodie]

Muppet,

When I met with issues looking like this one I used to sort the 800 last lines from  "/var/log/messages" " and the dmesg and paste them to pastebin. Usually it appeared that there was an irq conflict on the kernel level.

ie as root:

Code: [Select]

# tail -n 800 /var/log/messages
then you could pastebin it.

At the end of the kernel line of the menu.lst file you could now try  the noirqpoll option.

We still have a few older kernel versions, you also could try them:
$ apt-cache search kernel-2.6.33
kernel-2.6.33.7-pclos6 - i686 Linux kernel
kernel-2.6.33.7-pclos6.bfs - i686 linux kernel with the BFS scheduler
(...)
$

$ apt-cache search kernel-2.6.31
kernel-2.6.31.12-pclos1 - i686 Linux kernel
$

(and headers for each, of course).

[Taco.22]

Hi Muppet,

Give a different kernel a go before you totally give up.  Get rid of the proprietary nvidia driver and set up X through PPC with the nvidia driver from Synaptic.  If you can't log in to a GUI run XFdrake from the console.  It should actually be offered as an option if X doesn't launch.

As Mélodie pointed out, there are a number of alternative kernels on offer in Synaptic.  Install one and set it as default in boot options in PPC, or just choose one at boot up.  DON'T UNINSTALL ANY!!  Did an install/uninstall at the same time once and totally borked my system .  Don't worry about using an "old" kernel.  Some distros use the latest, some stay a bit off the pace.  I think Debian still runs 2.6.32 or 2.6.33.           

I have a TNT2 card in a P4 which works fine, so don't be in too much of a hurry to toss it.  Good luck.

Taco.22

[Muppet]

hello,

thank you for the kind motivition support. Mupput went bananas yesterday. since i have little knowledge of this kind of OS i have done many many reinstalls. for some reason if i am receiving the message that Xserver won't start after changing driver i can get it on again.....

well i tought why am i not asking a question in the nvidia forum. to be sure that i am having the xorg.conf edit in the way it should be.

first i am going to do that and will come back here to let know how this went............

regards,

Willem

[Muppet]

Hi,

in the mean time (have not yet received any reactions on the nvidia forum) i have done the advise from Melodie to add 'noirqpoll' to the end of the kernelline in the menu.lst file. have to see how this goes......

also i have done the command for the /var/log/messages and this can been seen here;

http://pastebin.com/hd1U6yaw

i have done today a new install, installed the nv driver and add noirqpoll in the bootmenu

perhaps there is a clue in the var/log/messages

regards,

Willem