(old)Puppy Linux Discussion Forum

Hi Berry

I have been trying to find out why the gxset will not save our use my settings. When I ran it in term window I was able to find out why, I can save my setting but when I hit the apply button it exits with a segmentation fault. I hope it is not my system or that it can be fix. If there is any thing else that I can do to help let me know.

BTW Puppylinux 4.0 is greet thanks and keep up the good work.

While working on my previous fix for modem detection, I found commentary on dealing with multiple modems in a PC -- a built-in versus an add-on. Other scenarios include a USB and/or serial modem connected to a PC with a PCI modem. Or even multiple kinds of PCI linmodems.

I have implemented a remedy that maintains a small table that the user or pupdial could change, to ignore use of the unwanted modem(s). In addition, a USB modem takes priority over PCI modems.

The changes are to /etc.rc.d/rc.local0 and rc.modem (including previous fix). However, I discovered the need to correct the condition "[ -e /dev/modem ]" in the service scripts to "[ -h /dev/modem ]". Here is why:
The SmartLink service (zzslmodem) links /dev/modem to /dev/ttySL0, which is itself a link to /dev/pts/0. Under the "sh" shell, the -e option does not consider a link to be a file (contrary to under the command prompt); the -h option always detects /dev/modem as a link to anything. This is appropriate because the scripts are testing merely for the presence of /dev/modem, which is always a link.

In my attachment I include 3 service scripts, but here are the changes needed for all of those scripts: A change that would give pl2303-type USB modems priority would be to change its name to "aapl2303" so it runs first. (The code changes zzslmodem to aaslmodeusb if a SmartLink USB modem is plugged in at bootup, so it runs first, too.)

I am uploading rc.local0 and rc.modem to go into /etc/rc.d. The service scripts belong in zdrv.sfs, currently in /etc/init.d, but wherever Barry is moving them for the beta.

I have tested and refined this over several days with several modems. It seems solid. I hope this can go into the beta, if it's not too late.
Richard

UPDATE 3/17: To summarize my request for service script changes:
1. Change all "-e /dev/modem" occurrences to "-h /dev/modem".
2. In zzslmodem, also add a test (line 28) for the name by which it is invoked, verifying it is just "slmodem" for the PCI version, and change the file name to match. I assume that the only reason for the "zz" is to work around the -e/-h problem triggered by the SL modems. Remove the "zz" in the name test and in the rc.local0 logic.
3. Since pl2303 is for USB modems (exclusively?), change its name to aapl2303 to give it priority. Do this AA-thing for any other USB modem types that I have not discovered.
4. Move the logic that determines the slmodem name to be used, if a USB version is present, out of rc.local0 and into wherever it is that the modem scripts are copied from zdrv.sfs.
5. Also move the logic for non-self-identifying (Thinkpad) Lucent modems out of rc.local0 to wherever it is that the modem scripts are copied from zdrv.sfs.
Thanks, Barry.

I would like to make the requested updates, but I do not know what file to change or where I could find it. I might then be able to verify them. Would someone point me to the modem detection script?

BTW: The file that controls modem selection is /etc/modemoverrides. It will contain commented (#) "ignore" statements that can be uncommented to skip selection of particular serial-modem ports or linmodem types (that have already been detected).
Richard

Here is the attachment that would not go with my original post immediately above.

UPDATE 3/18: I re-uploaded the modem-control attachment with corrections to rc.local and a minor change to rc.modem. Whoever uses non-self-identifying laptop modems or SmartLink modems should replace those files from the attachment, for testing. Barry, please be sure that this version goes to the release, if you plan to use the fix. Thanks.
Richard

Is the copy to RAM memory threshold 256MB? Could it be lowered for this Puppy?

While testing my updated "monitor control" fix, http://www.murga-linux.com/puppy/viewto ... 549#182549, I checked /tmp/bootsysinit.log for error messages. With my Thinkpad i1720 I see the following:

Running PCMCIA cardmgr program...
cardmgr[10973]: watching 2 sockets
Fetching x dri modules (but not loading)...
FATAL: Module neomagic not found.

However /tmp/Xorg.0.log shows that modules glx ande dri were not found, but that neomagic was found and used.

(II) LoadModule: "glx"
(WW) Warning, couldn't open module glx
(II) UnloadModule: "glx"
(EE) Failed to load module "glx" (module does not exist, 0)
(II) LoadModule: "dri"
(WW) Warning, couldn't open module dri
(II) UnloadModule: "dri"
(EE) Failed to load module "dri" (module does not exist, 0)
(II) LoadModule: "neomagic"
(II) Loading /usr/X11R7/lib/xorg/modules/drivers//neomagic_drv.so
(II) Module neomagic: vendor="X.Org Foundation"
compiled for 1.3.0, module version = 1.1.1
Module class: X.Org Video Driver
ABI class: X.Org Video Driver, version 1.2
(II) LoadModule: "mouse"
(II) Loading /usr/X11R7/lib/xorg/modules/input//mouse_drv.so
(II) Module mouse: vendor="X.Org Foundation"
compiled for 1.3.0, module version = 1.2.3
Module class: X.Org XInput Driver
ABI class: X.Org XInput driver, version 0.7
(II) LoadModule: "kbd"
(II) Loading /usr/X11R7/lib/xorg/modules/input//kbd_drv.so
(II) Module kbd: vendor="X.Org Foundation"
compiled for 1.3.0, module version = 1.2.2
Module class: X.Org XInput Driver
ABI class: X.Org XInput driver, version 0.7
(II) NEOMAGIC: Driver for Neomagic chipsets: neo2070, neo2090, neo2093,
neo2097, neo2160, neo2200, neo2230, neo2360, neo2380
(II) Primary Device is: PCI 00:05:0
(--) Chipset neo2200 found
(II) resource ranges after xf86ClaimFixedResources() call:

So, what I thought was a failure to find a file that is actually available, is apparently only a problem of the wrong name being placed in the "not found" message. If this is a Puppy error, it would be easy to fix. But if it is an Xorg error, we just have to live with it, or report it. I am attaching those logs and xerrs.log to provide all the evidence.
Richard

Which might be a problem when downloads disappear on Newbies, but is a great way to clear up a hung system from a full pup_save.2fs file.

Xine does NOT have the hangs that Gxine has in Puppy 3.01. And looks great. When I asked Xine to run in full screen mode, then Xine disappears. Might be fixable with the controls inside Xine. I did not look at all the internal controls to Xine, so it is probably not very important.

An upgrade of my pup_save.2fs from Puppy 3.01 to Puppy 4 did not go so good. Some of the Icons do not show color differentiation. Like the Speaker Icon on the Task Bar does not show where the mouse can grab onto it, and in fact, seems like I could not. However since my pup_save.2fs file is also an upgrade from Puppy 2.xx, seems like a rebuild would be in order.

Edit: So I created a fresh pup-save file to play with Dingo.

I installed Firefox 2 Dot Pet from the Puppy 4 files. Somewhere I discovered I have Flash for both Mozilla and Firefox. Maybe Flash crept in with Mozilla? I tried to install the Dot Pup for Java from de website. When I try to start program, it disappears.

I tried to download Java version from Slackware 12, which I think is jre-6-u-i386-1.tgz, and is large.

Ha Ha . Before I tried to install it I thought to clean up the space in pup_save by rebooting computer. As Mozilla download manager put the file in /tmp, bye bye download.

I like the quickness of Dingo. Maybe it could be easier to change the back ground to something different.

All in all, Dingo looks to be great.

purple_ghost wrote:I like the quickness of Dingo. Maybe it could be easier to change the back ground to something different.

Menu -> Desktop -> Wallpaper setter

Does fotox have a print function? If so I can't find it. Or do I have to save the edited image and print it with another app?

Gerry

this is on an acer 5610 laptop with builtin Intel 3945 wireless attaching to an unsecured network (no wep or wpa). also this laptop has a builtin ethernet adaptor. Puppy booted off a cd with a pup_save on hda3 which is a vfat (XP) partition .

Have found that alpha7 doesn't always store wireless networking config between boots even when I "save configuration" -- and it is inconsistent about whether it detects the wireless networking hardware or not too (sometimes will, sometimes won't).

alpha7 sometimes sees the ethernet as eth0 and the wireless as eth1, sometimes the other way about, sometimes only sees the wired adaptor

when it connects it works fine

paul

Barry sir,
I would find the ability to use a web cam with Yahoo
I/M most useful, i have not downloaded dingo 4 a7 as i am awaiting the beta.
Michael.

Barry:

..configure option for Ayttm for Yahoo Webcam support, but it needs a package named 'jasper'. Would this feature be useful to you?

Yes, very much.

Hi all

I have tried to write a small program to handle some of the xset settings. I have not tried to write anything in about 10 years, have only edit some php. I did not write this from scratch but have got some of the xset setting as easy as point and click. I plan on getting more working as soon as I learn some things that I know but have forgot, also if somebody know how to execute a program from inside puppybasic let me know. This is based on gtk and puppybasic. It need very little to run, just put it in your my-application folder and change the setting to execute, then in a term window or gexec command line type Pxset.wxbs and it should run.

Please any feedback would be nice.

Attached is my completed modem-control fix as described in my previous posts here. This version replaces the others and adds the capability to logically disable ("ignore") particular modems if multiple modems exist in or are attached to a PC. I have included difference files for each updated file.

The "ignore" statements reside in /etc/modemoverrides, which is built as modems are encountered. To activate an "ignore", uncomment it, by removing its initial # character. The file is to be edited manually, at least until it can be managed by pupdial.

In addition, there is new support for two Thinkpad modems not previously recognized by Puppy. A line for each was added to MODULECONFIG, containing the module name, vendor ID, device ID, and description. I used the old modem wizard to try various drivers and find the tty-type. I used the WinXP device manager to find the vendor and device IDs in the driver details tab. Anyone with an unrecognized laptop modem might try that technique and report any new modems to add to MODULECONFIG, by providing that information and the Device Manager's description.

Some details:
1. All of the files in /etc/init.d of the attachment go to their corresponding zdrv directory. Please note that two of the service scripts are renamed to use an upper case initial letter: Cdcacm and Pl2303. This is to make them run before the scripts for PCI hardware, so if a USB modem is plugged in, it will be used instead of a PCI modem.

2. The firmware.dep.2.6.21.7 file goes to zdrv, or its three minor changes can made manually.
EDIT 3/27/08: The changes are: 3. For testing purposes, the /etc/init.d files can go to the regular /etc/init.d directory.

This has all of the improvements I could think of, so hope this can go into the beta and final. Thanks for your help on this, Barry.

EDIT: Attachment now includes the difference listings.
Richard

Mr. Barry
Not only are ALL ports open on your home (personal) system but you have left everybody running Puppy WIDE open.
Do I need to go underground to show you how bad you have screwed up?

Murga, I have your picture!
END OF STORY!!!!!! you will be added to a hosts file. Not just mine!

I think this has been a problem for a long time, so I investigated the cause. The script, rc.modules, uses lspci to obtain the hardware IDs present, then loads the drivers for that hardware. However, lspci does not see the PCMCIA hardware until the cardbus/PCMCIA driver (e.g., yenta-socket) is loaded. So the cards are never detected during bootup.

I have not coded a solution, but post this now in case Barry prefers to implement the fix himself. I expect that the lspci must be done first to find any cardbus driver, then again once that driver is loaded.

I strongly hope that the fix can go into 4final, but preferably into the beta. It will improve puppy's reputation of reliability. I will work on a solution until I hear that it is no longer needed.

EDIT 3/26/08: I have coded the fix and have attached it. It includes my previous update to the same module (but does not require the rest of that fix be installed). To test it, extract rc.modules to /etc/rc.d, and reboot.
Richard

Hello,

Dont know if this bug is already reported in.

- Cant connect to a WPA 1 wireless network, using PSK passphrase.

This did not work in Alpha 1 or in Alpha 2.

- However using Puppy 2.17 it works like a clock and connects without anyproblem.

My access point is a Proxim Ap-2500M. Standard product when building large wireless networks (mesh).

Best regards,
Johan

Not directly a bug report but a request to add this support in order to better control the synaptics driver, f.e. to disable completely the touch pad.

EDIT - Ignore - that doesn't seem to be what fixed it after all...

divisionmd wrote:Hello,

Dont know if this bug is already reported in.

- Cant connect to a WPA 1 wireless network, using PSK passphrase.

This did not work in Alpha 1 or in Alpha 2.

- However using Puppy 2.17 it works like a clock and connects without anyproblem.

My access point is a Proxim Ap-2500M. Standard product when building large wireless networks (mesh).

Best regards,
Johan

Yes this IS already known but it is hardware dependent, and is why Barry is abandonning the latest kernel which has "retired" certain Modules and replaced them. For example IPW3945 was replaced by IPL3945 which has WEP but not WPA encryption.

I don't know about your specific hardware (forum member Tempestuous will be able to help with that if you get really stuck), but you could try one of the "Retro kernel" versions of Dingo and see if that cures your problem.

I believe that the first Beta is due over the weekend and will be on the retro kernel, so you could try that out when it is launched.

Good luck.

What level of hardware is Puppy compiled for now?
I copied my main (non-Dingo) setup onto an older PC with a 350Mhz Pentium 2 processor. This setup has the Dingo Mplayer and ffmpeg packages installed. Mplayer crashes with I hope xine/gxine in Dingo doesn't have the same problem!
Also, I could have sworn I saw a page on puppylinux.com where Barry explained how he compiled Mplayer with shared ffmpeg, but I can't for the life of me find it now. Maybe it was on the forum and was someone else, but all I can find now is one little part on the puppy-from-scratch page. Does anyone know where to find this information?... because I could never find it anywhere else. If I remember correctly, using shared ffmpeg isn't "supported" anymore, but it's worked perfectly for everything on my other machine for months now.