Havin_it

Getting de-railed from this at the moment by update-source probs, but I did find a MDK RPM for ndiswrapper 0.8 on the web. I'm not highly keen to go installing before I do some checks on the 0.4 I 'supposedly' uninstalled, though. That one was installed from a local RPM as well, and although there were no apparent blips with the uninstall, I went back into 'install' and still found an entry for it in my local source - even though I'd deleted the RPM from that location!

Not wanting to drag this thread off-topic, honest; I just need to get these issues sorted before I can get down to the business in hand. Rest assured, come that time I will be relentless!

Will the installation bork if I don't have the kernel source in the location above(or at all)? There was no dependency message or error when I installed the 0.4 version.

Is it normal for all of them to be updating at the same time? I can't seem to get anything from the lists at Mandrakesoft or easyURPMI today. I was looking because my update_source (Strasbourg Uni, the only online source I've used in the past) wasn't working last night, although I'd successfully installed a couple of months' worth of updates from it earlier yesterday.

I also had a power-outage yesterday - I'm sure that's not good, but I had restarted after the updating so would think my local hdlist would be okay...darn well hope so anyway...

Can anyone offer confirmation about the current levels of availability? I've been trying various UK/EU sites including the PLF one at ftp://mandrakeusers.org with no joy. I've tried Community and Official repos. I've tried doing it through MCC and through bash using easyURPMI-generated commands. ...Nothing.

Edit: found a working mirror eventually (ftp.club-internet.fr) and seems to be synching allright with RPMDrake. Conclusion: Sites at Mandrakesecure are not reliable, nor are those at easyURPMI. Go for the download page of Mandrakelinux main site.

Very little joy there!

The version of ndsiwrapper available from my update-source is only 0.4 so I don't even know if it will work with this card. Once installed it had naff-all usage info with it, so I can't really tell anything.

I've looked at the INSTALL notes for both versions, and I noticed that kernel source is required. Possibly due to it being all installed from RPMs, there's no /lib/modules/<VERSION>/build location on my machine. So if I want to build 0.8 myself, I need to have kernel source to put in there?

I'd heard that people have had nightmares with this software, starting to think I see why...

Gotcha. Thanks for the quick reply! I didn't want to touch anything until I knew.

<props>

Hi, I just pulled a rather huge update, and RPMdrake has thrown up this change in /etc/inittab in the section 'Default runlevel':

(old) 5 - X11

(new) 3 - full multiuser mode

I have no clue! I'm tempted to stick with the old, but could that cause problems with a related bit of updated software?

The RPM this relates to is initscripts-7.06-49.1.100mdk

Hello from Windows (HEY! Stop throwing that mud at me!)

Just thought I'd chime in on this old thread because I have this card myself and am about to try testing it out. Will have to reboot to Linux to setup the hardware, an I'm not on a hotspot right now but will be later.

I have:

10.0 Community (probably needs a rack of updates but will do that first when I get MDK booted up)

Drivers from the card's CD available on shared partition

Latest ndiswrapper from SF (might just re-dl that thru RPMDrake anyway just to be safe)

Speak to you soon. Feel free to chime in with any new info that might be relevant since this thread went dead.

Mmm, apparently that's not a goer because pppd is a binary...

<despair>

Well, this is just ticking me off now. I've chmodded everything, and it still won't work. It seems the command /usr/sbin/pppd call adsl is still the problem. It's doing something, it returns without output, but when I do ifconfig there's still no sign of ppp0. I do it as root and there's no problem!

pppd must be using something else that doesn't have permission set for me, but I wouldn't have a clue what that could be.

I dunno, maybe I'll just stick with doing it as root. All this carry-on hardly seems worth it.

Yes, I think I am down to one version now. It appears it was the first uninstall attempt that went awry, as I went back into 'Remove Programs' again and 1.2 was still there! However this time it was successfully banished, even from the menu.

RPMdrake seems to be pretty temperamental generally. Setting up tyme's repo site in Package Manager was an arduous process - RPMdrake timed out and hung several times before success - and whichever part of the program I use, installing, managing, updating or whatever, my cursor is permanently in 'waiting' mode (although that's not the case as I can continue my activities regardless).

Anyway, I'm now happily working away with the GIMP - first impressions make it a serious rival to Photoshop in my book. The panel-docking improvements definitely eased my transition; it'd be nice to see them all anchored together, but that's a small criticism for what seems a really top endorsement of Linux for me as a newcomer. I shall tell my friends!

(both of them)

That's right - it was part of my initial install, but I used RPMdrake to uninstall it. I'd launched it from shell, and I didn't see any unusual output there during the uninstall.

I should also say that when I launched the 'phantom' 1.2, I just got that 'installing files' monologue and not the 'first run' setup dialogue (where you setup the page file etc.). Everything was clearly still there, as I wasn't prompted for the CD.

Does the fact that I installed it with the OS maybe make it harder to remove?

Thanks everyone - speshly tyme for that first link. I have indeed set that site up as an update source. It does seem that gimp2 is part of the repository now, but it never appeared on my previous update site.

Now I have another problem - I uninstalled gimp1.2, but it's still in my menu! I clicked on it, expecting a Windows-style "That's not there anymore!" message, but it created a bunch of files and loaded up as if it'd never left!

I'm not suffering any apparent problems due to both being present, but I'm aware that it's quite a weighty app. and I only have a 9Gig partition and I'm trying to keep it tight. Any advice how I can remove 1.2 completely?

Thanks all - just what I was after and a lot more.

. >

|

. >

Hi,

Just wondering if anyone knows of a RPM of the GIMP 2.0 that works with Mandrake 10.0CE? Would the Fedora or SuSE RPMs from gimp.org work, if anyone's tried them?

I'm not scared to build from source if necessary (well maybe a bit) but I'd sooner have RPM because I've no clue how to uninstall source packages.

Pardon a thickie!

<hee hee hee>

Nice site, very funny pic. They still just don't look quite as loony as I want though...

umm, dunno about that, sorry. You say you have a router - I take it it's connected to the ethernet port?

If it's plugged in during setup, you might be getting a problem something like the ethernet card waiting to get an IP address through the router, but not getting one for whatever reason.

Think this one's outta my league - passing to the router guys.

Thanks, I will try chmodding adsl as well. BTW, is the 'in your path' thing really such a hacker-beater? If my user was possessed by evil (well, not by me anyway), wouldn't it just take a few educated guesses on system topology to locate the path and use it if permitted?

Just a thought - forgive me if I'm missing something obvious, us noobs usually do!

Hi, I was just wondering where I could get some wallpaper with lots of penguins like the little fellas in this site's top banner? those guys are so cute ^_^

I'm just big on penguins and have a friend who REALLY likes 'em. Yeah, she's a girl. What can I say.

PS - Not to diss your smileys, but I wish they wouldn't keep jumping in when I'm putting my good ol' ASCII art like above. that guy doesn't look anything like my 'caret underscore caret' masterpiece! Can I escape them somehow?

Does it freeze fatally, or just pause for a REALLY long time before returning FAILED?

I had many problems with this one, and still don't know how to do it gracefully. On clean install, my eth0 (Realtek ACsomething) is detected and - yes - apparently pre-configured in summary. Then on booting it doesn't freeze, but initialization returns FAILED. Doubly wierd because it is recognised by ifconfig...

If it's configured for DHCP, it takes about a minute to return OK (or sometimes FAILED depending what hapless details I punch in) - maybe this is what you are getting? Also, there may be some issue with 'network hotplugging' so I'd try disabling that.

If you can still get into eth0 configuration, you might want to try giving it a static IP (if that's workable with anything you're wanting to do with your LAN card later). This gives an instant OK on boot for me, though it might not fix your problem.

10.0 seems to be giving everyone problems with their eth0 so you're not alone. In my case it was preventing my DSL modem from working, I just wished I didn't even have a darn LAN card.

Thanks guys - yes, I see the point about not making myself able to SUDO everything with no password, so I'm reverting to the chmod method. The wierd thing is it doesn't seem to work as I'd hoped.

In the scope of (having done the chmods on the programs in question) trying to run the commands as myself, what's now happening is the commands appear to return without error but are not actually succeeding - probably why the non-sudo script I wrote doesn't seem to succeed.

The commands are:

/usr/sbin/modem_run -m -f /usr/share/speedtouch/mgmt.o

/usr/sbin/pppd call adsl

sbin/ifconfig

I'm still able to run these as root (and without the paths), but as myself, the ifconfig returns only eth0 and lo, not ppp0 as intended. modem_run appears to work, as the modem's LEDs cycle in the normal way, so I'm assuming pppd is the problem. Do I need to chmod the peer file for adsl as well? As it's basically a config file I didn't think it would need to be executable.

Woah ... mad props for the Partition Magic then. Thanks for the all-clear guys. Like I say I'd sooner not have to guddle with partitions at all, but I figured use each OS's native tools to work with the partitions that it will use. And I'm so glad I have my two funky desktops now!

***Just waitin 4tha next thunderbolt***

PS - sorry for double-posting. Could I be any more new?

Hi,

I hope this question is valid as it's half-Windoze-related, and arguably more software- than hardware-related; but here goes.

I used Partition Magic to shrink my Win partition and setup a dual boot with Mandrake 10.0 CE, used DiskDrake to setup all the other partitions. Since then I've reinstalled Mandrake a couple of times, including the bootloader (which is in the MBR).

Everything has been working quite smoothly in both OSes, but I started Partition Magic the other day to be greeted with the following message:

Powerquest Partition Magic has detected an error 116 on the partition starting at sector 112518 on disk 1.

The starting LBA value is 112518 and the CHS value is 1783845.
The LBA and CHS values must be equal.
Powerquest Partition Magic has verified that the LBA value is correct and can fix the CHS value.

Would you like to fix this error?

Well, I have no clue what any of it means. I probably should just choose 'yes', but I'm paranoid that this might stuff up the Linux install somehow, and like I say I'm not suffering any ill-effects from this problem at present. (I only opened Partition Magic again out of curiosity - I wasn't planning any further meddling.)

Does anyone know if it's (a) safe and (B) important to fix this? Thanks for any advice.

Thanks guys. Just been checking out sudo (or more specifically visudo) and it's driving me frickin' nuts already.

I think the syntax for the powers i want should be

havinit ALL=NOPASSWD: (root) ALL

But apparently not! So I guess this is now a plea for some sudoers knowledge. The manpages are not really great on this subject! I'm currently doing these edits in Kwrite because I couldn't make head or tail of visudo's control syntax.

(I need the NOPASSWD bit so I can hopefully use these commands in a script)

Well, I guess the title says it all. I need to use shell commands to start my modem, so I'm thinking of writing a script to automate the three commands this involves. It's my first attempt at such a thing in Linux.

So when I have it written, and sitting on my desktop, is there a way I can configure it so that it just runs when I double-click it?

Also, the commands are privileged, so can the script also perform 'su' automatically by some command that includes the root password? Or is there a safer, permission-based way of setting-up the script so that this is possible?

Any ideas or pointers would be ... well, great, ya know.

Edit: moved from Software by spinynorman

Given how many peeps are having nightmares with this modem (DSL generally infact), a copy should also go to MandrakeSoft for the 'Hi, I'm new here, why doesn't anything work page'

Well I mucked it up again and have done 2 more clean installs in the last 24 hours. I just couldn't resist diddling with eth0, and I think that's what did it (waaaaaah!) But I'm up again now (Yaaay!)

More observations: Don't do ANY net configuration while installing. The system summary I get on install says Network: lan and I didn't touch it at this stage (might be important). As long as the modem is plugged in when you're setup and ready to configure it, that should be enough.

I noticed from watching the 'verbose' startup that eth0 initialisation FAILED. That doesn't matter - in fact it's the only circumstance in which I've made DSL work.

If you're lucky enough to have a floppy drive (unlike me) have the mgmt.o or whatever file ready on that, and copy into /usr/share/speedtouch - I had to copy it out of my windoze partition manually because the option in Control Centre to retrieve the file from there DOES NOT WORK. EVER. If you haven't run the config process yet, you'll need to create the directory.

Run through the connection setup from Mandrake Control Centre - you might want to launch it from a shell so you can track any errors (there'll probably be some), like so:

[foo @ bar root]$ su

<enter root password>

[root @ bar root]# drakconf <this'll launch the Control Centre>

Do the configuration through Hardware > Hardware > Alcatel speedtouch modem > Run config tool, keep watching the shell output (mine always complains about not having a floppy - well DUH!)

Nothing very complex in the config process, the pacages you'll need are installed as required, one less obvious thing to do is click 'Advanced' on the ISP settings page and enter your country's VTI / VCI numbers. If your ISP is really nice they might have explained what the other advanced setting next to them is - I haven't the faintest idea myself.

For the radio-buttons that follow, I'd suggest: NO to start at boot, NO to restart Network, and YES to try connection now (so you can see the shell output when it FAILS).

If it doesn't FAIL and you are now connected, leave this page immediately. You sicken me.

Jot down the error output you've had from the shell, then try the shell commands modem_run followed be pppd call adsl (as detailed above). It didn't work for me at this stage, but I think I now know why - read on....

Restart your computer. You need a break. You could hit Esc to watch the Verbose outputif you like.

When you logbackon, go and have a look at the file /etc/peers/adsl (I did this because I tried the shell commands and they still didn't succeed). You may find that the VTI / VCI numbers are not the ones you entered! I entered 0,38 and found this file to contain 0,56 !

You need to be root to modify this, so open a shell there and open it with 'kwrite adsl' - or whatever textpad you prefer. Using Kwrite dumps several pages of rather worrying output to the shell, and seems to leave a phantom backup of the file in the directory, but don't worry too much about that.

Once that's done, if you're me, the shell commands for connection may just work.

If not, you too are a victim of the Redmond-Telstra-CIA-Colonel Sanders zionist alien reptile conspiracy! Give up now! They're watching your house right now! Draw the curtains and fire up your Windows screensaver - quick!

PS - John, if you want to give it a try - http://home.planet.nl/~andje000/mgmt.o

(this one works for me - on modem_run I get the proper LED cycle same as on Windows startup)

Hi, I signed up here just so I can thank Arthur for getting me over this problem. I don't think my problem was as complicated as John's, but I was nonetheless tearing my hair out over it for nearly a week.

I won't lie - my comportment and indeed personal grooming suffered.

The way MCC restarts the network (ifdown/ifup eth0) had me convinced that this was the root of the problem. I tried every ethernet config you can imagine, with no difference to the DSL hookup results. I was pretty close to self-destruction (and by god I'd have taken the darn computer with me, ha ha haar).

Tried Arthur's shell-command tips, modem_run followed by pppd call adsl, and lo and behold - ppp0 UP and indeed RUNNING! Thank you Arthur, from the heart of my bottom.

Like a reformed alcoholic who speaks at AA meetings, I now feel the need to help others, towards the day when all our Speedtouches work in harmony and we may Konquer the Kosmos together...

Umm sorry, I do tend to go off on one. Anyway, John, i don't know what is at the root of your problem, but here are some observations:

1) Many people say the 2.6.3 kernel is to blame. UNTRUE!

1) My setup has a user module, not a kernel module (I think)

1) I see you are using the firmware alcaudsl.sys - I used this initially, but it did not work for me.

1) How quickly does the modem_run command return? If it's instantly, the firmware should be loaded. If not, try rephrasing it thus:

[root @ localhost root] modem_run -m -v 2 -f /path/to/alcaudsl.sys

(The extra -v switch is for verbose output, so you can see any errors this throws up.)

1) If there seem to be errors, or even if there aren't and the command still returns slow, try hunting for some alternatives. The one I used is called mgmt.o (this is for a silver 330 model), I can't remember the URL for it but do a google and you should find it.

1) Don't be downhearted about the line !/bin/bash not working. It's not actually a user command, but rather something that's used in shell scripts as a first line (saying 'please run these commands through the bash shell'). You don't need it when you're already in the shell.

I've said enough, but hope some of this helps. I'll be back!