blackpage

greetings hineiko That mobo is one of the kind with "integrated everything" (including video). Those are a bit bound for errors. But you could try to press "F1" on the MDK10-boot screen and start the Mandrake installer by typing "linux noapic". That should circumnavigate possible issues with some problematic devices on such cheapo-boards. Keep us informed if that worked.

howdy mjwebb007 According the installation of W98/XP and Linux I'd like to throw in my 2c ... As far as I can see the Windows 98-install is the crux of the whole thing. The first thing to check is if there's a way to install that OS on a SATA drive at all. I haven't installed W98 in years now, but from what I can remember I think it must be at least w98SE to even recognize SATA drives. The first step therefore would be to check if you can install W98(se) at all. For this I'd recommend to place it on the slower drive (Caviar/250GB) which should be partitioned somewhat like that ... Partitions on Caviar-250 1: ~ 32GB (for W98), primary, FAT32 2: ~120GB, primary, NTFS, for all the apps/data that XP needs 3: ~ 80GB, primary, Linux filesystem (ext3 or reiser) as data-vault for your Linux OS 4: ~ rest, primary, FAT32, partition to allow data-exchange between the OS' As I use the Raptor myself I'd definitely make this drive the home for the 2 major OS': XP and Linux. This results in creating at least 3 partitions on the Raptor ... Partitions on Raptor-73GB 1: ~40GB (for XP), primary, NTFS 2: 0.5-1GB (Linux swap), primary 3: rest (Linux root), primary, ReiserFS (or ext3) As you can see this is a lot of partitioning and any decent Linux installation CD should provide the necessary partitioning tool. Installation sequence After partitioning the drives, start out with the Windows98 installation ("windows"-dir goes to 1st part. on Caviar). If you can get W98 running on the SATA drives, just start the installation of XP from within W98. Install XP to the 1st part. on the Raptor. The installer should recognize the W98-installation this way, and automatically configure a "dual-Windows"-boot setup managed by the XP boot-manager still. As the last OS, install Linux. Linux normally sees a Windows-install and provide a triple-boot setup in the end. If all goes well, you should start with LILO/Grub-screen, followed by the XP-boot-menu ("Windows XP" and "Windows 98"), meaning that you have actually 2 daisy-chained boot-managers. Any boot-code, of course, goes into the MBR of the Raptor-drive. That's how the procedure at least here works with multiple W2k- and Linux-installs. I can only hope XP acts likewise.

Code: <?phpinclude_once( "./inc/browser_funcs.php" );include_once( "./inc/local_html_files.php" );$goodTip = "DapperDan's last posting";if ( isMozillaSlow(cWITH_LOCAL_HTML_FILES) ) {__blameGeckoRenderingEngine( true );}else {__blameNetworkSettings( true );__checkOutGoodTip( $goodTip );}?> hope you have chosen your nick wisely cu

gidday Noir Alternatively you could d/l a more recent kernel-source from the 2.6-series and compile a kernel that supports your GB-LAN chip. The chip in question is the regular "Marvel"-Gigabit-LAN controller, which can be found in the "Networking"-section of the kernel-config-application (the LAN-card module is also named "sk98lin", I think). Hope that helps

Howdy vital I fear you've picked probably the most bothersome ADSL-modem for linux one could imagine. According to this document (PDF) there is no linux-support for the PCI version. All the internet has to offer are howtos according the USB-version. Other sources meanwhile indicate it's possible to utilize the "mgmt.o"-management module of the USB version for the PCI-card as well. The common denominator for all these references is: "The Speed Touch PC-card is a p*** in the a** to set up under linux." And according to Alcatel's lousy and nosey driver-download page I can only subscribe to that. So ... no solution to offer, but a tip: Trade the card for some decent external modem. Should only be a couple of EUROs/$$$.

Howdy DapperDan, you can concatenate multiple commands in a script e.g. in this way ... Code: [sup]#!/bin/shbaseDir="~/script_commands"outFile="$baseDir/testfile_$( date +%Y%m%d-%H%M%S ).txt"alias cmdSeq="mkdir -p $baseDir ; cp /var/log/messages $outFile ; tail -n 5 $outFile"cmdSequnset cmdSeqecho "all done"[/sup] The "alias"-thingie here is necessary and it is also often a quite useful method as e.g. a straight invocation of the commands would not work properly ... Code: [sup]NO-GO-EXAMPLE ...cmdSeq="mkdir -p $baseDir ; cp /var/log/messages $outFile ; tail -n 5 $outFile"$cmdSeq[/sup] In the above example you'd get error-msgs about invalid commandline options for "mkdir". A step further: If you want to ensure that the command-sequence only runs through if there are no errors encountered, you should concatenate the commands with ampersands ("&") like this ... Code: [sup]EXAMPLE: ensure proper exec. of previous cmd ...cmdSeq="mkdir -p $baseDir & cp /var/log/messages $outFile & tail -n 5 $outFile"$cmdSeq[/sup] But I'm sure you already knew that one from compiling 2.4-kernels (make & make dep & make xyz & make world_go_round ...". As it goes for a "script corner": I wholehartedly support this idea. And if anyone's interested, I could throw in e.g. an iptables-setup script that utilizes "arrays in bash-scripts" for hosts and services. Neat stuff regarding string handling and "loops" in there hope that helps

evening koppaspider Obviously you have some sort of Live-CD you can boot from. After booting the Live-CD just mount your harddrive with write-privileges and proceed with the usual "reset root password"-procedure (editing /etc/passwd and /etc/shadow). In case read/write-access isn't available via the context menu, just open a console and mount the drive "manually" ("user@box# mount -t ext3 /dev/hda1 /mnt/hda1"; just replace "ext3" with the filesystem the partition that contains the /etc-folder has, and replace "hda1" with the proper handle of the partition) That should do the trick.

howdy ceejay949 Well, there is help for NTFS on the way in the form of the "Captive"-tools. That software utilizes the original MS drivers of a w2k/XP installation and allows read/write access on NTFS partitions. I ran across these during the installation of Kanotix (a debian based linux distro) that already comes with a link to these drivers. So, at the moment, I cannot say much about it as I need to test that. But sources on the internet indicate it's actually working quite sweet. hope that helps

greetings bungalowbob Last things first: I can still remember my cursing when I first tried to get some Linux-distros up and running a few years back. You are right, the learning curve is sometimes a hard one, but you need to know that you will be rewarded with an OS and free apps that surpass anything that has made Billy "The Squid" Gates rich in both, stability and security. So if you think you're at your wits end, just come here, ask your question, and be sure to get help if it's at all possible OO fonts Open the "Tools"-menu and go to "Options -> View". There you should see a field labled "Scale" that probably is set to "100%". Blow this up to whatever size you want your GUI fonts to be. If you are interested in the reasons as to why the fonts are displayed so weird: The combination of a monitor's size and the resolution it's running at results in a so called "display dpi"-value (dots-per-inch). Not always is this dpi-value computed and set properly at installation time. You can check what MDK10.1 thought your dpi-value was by taking a look in the system configurator (Mandrake Control Panel) or by peeking into the various files whre this value is stored (e.g. [PATH_TO_KDE_CONF]/kde3/kdm/kdmrc). You might want to check these dpi-settings. Take a ruler, measure the width of the displayed image on the monitor and devide the value of the horizontal resolution you are using with that number. Example: You are using a resolution of 1024x768, the displayed image on the screen is 13.5 inches wide, so the proper dpi value would be "1024 / 13.5 = 75.85" which rounds to "75dpi". Purpose of the remaining MDK CDs Well, that's simple: Those are stuffed with software-goodies that you can install. Mandrake - as any other distro - comes with an overwhelming number of applications, tools, GUIs and whatnot. You will need those CDs once you want to install some software that you didn't pick during setup. So take good care of the CDs. "halt"-ing the machine That's of course a security thing, and it's generally a good idea to not let users shut down a machine willy-nilly whenever they want. To make things not too hard for you, you can check the internet about how to set up e.g the "/etc/sudoers"-file. This file stores infos about superuser-priviledges that are granted to certain users. You can also set there if this certain user has to enter the root password for a priviledged command. Put your user-name in there (read up o the "sudoers"-file on the net) and you will be able to stop your machine e.g. by typing "sudo halt". hope this helps

gidday pwandrews What you need is to get some bootloader into the MBR of the C-drive. As to how easy this is, depeneds on a few things. But here are a few options ... 1: Re-install MDK10 on a secondary HDD If you have a spare disk somewhere, then attach it as secondary master, flip in the MDK CDs and install MDK10 anew. Towards the end you will come to the bootloader-setup. Check if MDK detected the Windows-install on your primary disk. If so, complete the MDK-install and let MDK write the bootloader into the MBR of the Windows-drive. This should give you a fully qualified boot menu (including windows) again. !!!ATTENTION!!!: If you ain't got no spare disk, and your Windows-partition is NTFS and not defragged, I would refrain from resizing the XP-partition. Try the boot-disk option in that case instead ... 2: Reactivate the primary MBR with bootdisks To do this you should grab the W2k/XP boot-diskettes which can be found probably here ... Various bootdisk-images/links or directly from ... Tools for the doomed (from the makers of doom In any case you need either 3 (w2k) or 6 (xp) diskettes. Boot from these disks and enter the recovery console. Once there type "fixmbr" to have a shiny new master boot record written to your harddisk. There are other options too, involving Linux-live CDs. Though I don't know what level of difficulty you consider acceptable. In any way ... hope that helps

'lo there Midway through this thread ... Java installation in Fedora/RH ... you will find some convenient tips as to how to get things fixed. Hope that helps

Heya Dapper Dan This is really odd. Firefox is one of the very few apps that never bothered me with crashes, no matter under what Linux or Windows flavor I had it running. I really wonder what's going on on your box. Maybe you want to try to run firefox through the debugger. So in case you have "gdb" installed, open a shell and change to your firefox directory (e.g. "/usr/local/firefox-1.0/"). Once there launch that command ... Code: ./run-mozilla.sh -g ./firefox-bin -d gdb (EDITED - forgot the main part: to launch the app You should end up within the dbg-shell and be greeted by a friendly (gdb)-prompt. Just type "run" here to start firefox. Don't be alarmed about a possible million of warnings and notes that you will get when gdb launches the app. Those are normal and most of them deal with telling you that no debugging symbols have been found and that some GTK/GNOME permissions are wrong (another reason why I don't like GNOME; and on a sidenote: what the smeg is "orbit"? Now do what always makes the application crash. gdb should catch the error and at least provide some info as to WHERE (what lib) it occured and WHAT occured. Hope that helps (keep us updated)

heya matttah Unfortunately I ain't got no decent solution at hand, just this: I ran into a convo-log of Alan Cox (kernel developer) the other day who mentions this error in combination with ACPI, USB-kbds and kernels > 2.6.5. I can't remember the URL right now (will try to find it again) but as far as I recall turning OFF "USB legacy support" in the BIOS and disabling ACPI have been amongst the suggestions. Unfortunately not even these kernel-masterminds mentioned any specific fix that made its way to my ganglias. So maybe you just give it a try with a little trial-and-error testing. On a sidenote: Does the prob occur too when you initiate the rendering process with the mouse? good luck

I see ... then you'd need to incorporate some MaxKeepAliveRequests as described in the manual ... <snip> MaxKeepAliveRequests directive Syntax: MaxKeepAliveRequests number Default: MaxKeepAliveRequests 100 Context: server config Status: core Compatibility: Only available in Apache 1.2 and later. The MaxKeepAliveRequests directive limits the number of requests allowed per connection when KeepAlive is on. If it is set to "0", unlimited requests will be allowed. We recommend that this setting be kept to a high value for maximum server performance. In Apache 1.1, this is controlled through an option to the KeepAlive directive. For example: MaxKeepAliveRequests 500 <snip> (from the Apache 1.3 manual)

howdy dje007 Yes, you can tell Apache exactly how it should handle requests and how many processes there sould be to serve requests. In particular you will find the settings plus the respective descriptions in either the file "httpd.conf" (Apache 1.3.xx), or "httpd2.conf" (Apache 2.0.xx) which are loccated under "/etc/httpd/conf" (for most apache versions). The settings to look for would be ... StartServers MinSpareServers MaxSpareServers MaxClients MaxRequestsPerChild (and maybe some other) As it goes for connections per IP handled by multiple apaches on one machine ... You'd need to simply start Apache twice with different conf-files to get different connection settings for different interfaces. In brief a setup like this would look like this ... Apache 1: installed in "/usr/local/apache1" is started/stopped via "/etc/init.d/apache1" (this startup script also tells the server which config file to use, e.g. "/etc/httpd/conf1/httpd.conf" and uses the above conf-file which contains ... Code: Listen 111.111.111.111:80 This tells the first server to only listen on this specific IP. Also, set the spare-servers/requests-settings as mentioned above to your liking. Apache 2: installed in "/usr/local/apache2" is started/stopped via "/etc/init.d/apache2" which tells the 2nd server to use e.g. "/etc/httpd/conf2/httpd.conf" which contains ... Code: Listen 222.222.222.222:80 As you can see, running multiple Apaches for the sake of individual request-handling can get a tad complex, especially with Apache 2-series server, which use a flock of config files that you'd need to seperate or keep unique according to your needs. Also the startup scripts need to reference the different servers. But in general: What you ask is in fact possible (I suppose hope that helps (edited due to hasty typing

@ egorgry Hey, thanx a million, that's a superlative hint! If it indeed manages it to make file I/O more smooth in GIMP e.g. then I'll owe you big time. Thanx again

I can work with everything, though not for long if GUIs show those little flaws that drive me nuts, as GNOME does by always starting at root-folder with no option to store fav-folders, never showing hidden files without me doing tricks and so on. I hope they change that some day soon, cause GNOME certainly has potential - especially for Linux-beginners. That's why my motto is: KDE or bust! Having worked for a long time with Macs I just _need_ KDE with "acqua-graphite" as style and Baghira-Brushed as window-decos to achieve my favorite Linux-Mac crossover look. As it goes for the Mac-feel: I never liked that anyway, so I'm glad I can utilize a 3-button mouse without having the Icons inflate to the size of Texas

Evening dudes/dudettes After a good experience with FC3 last week on an old P3-box with ASUS P3B-F board, I had our boys from the tech dept. install FC3 as tryout on newer hardware. The problem The installation on the new machines went superlative, everything was partitioned and detected nicely, but ... on the first boot Fedora hangs on always the same spot: It activates the swap partition ok, then stands for about 2 mins, displaying the boot-messages, then the screen goes grey, the mouse pointer stays and nothing else happens. The pointer is sorta blue-ish, so this might indicate background activity, dunno? We even let the machine stand still for 15 mins, but FC3 just keeps displaying a grey screen and a mouse pointer. On the harddisks or the DVD-drives there seems to be no activity, also the LEDs on the NICs show no networking activity whatsoever. Interestingly enough this is definitely not some sort of kernel panic. When you press the power-button, the boot-console appears again and notifies you that the system is going down now. The installation that I did last week still works nicely. Here some specs of the machine that works ... Machine specs FC3 working on that hardware: P3/850MHz, ASUS P3B-F mobo, 768MB RAM, Gladiac 2MX gfx-card, 2x3Com-NICs, 2x60gig Seagate Barracuda IDE disks, 1xPLEXTOR IDE DVD FC3 failing on this hardware: Athlon XP3000+, ASUS A6V600-X VIA KT400-board, 1 gig GEIL PC3200 dual channel RAM, onboard LAN, 1 x 3Com 100MBit NIC, Geforce 4MX, 2xSAMSUNG 120gig IDE HDDs, 1xPlextor DVD. I'd like to note that -out of curiosity- we installed a flock of distros on one of the newer machines. MDK10, Debian Woody, Yoper, Suse 9.2. All these distros booted the machines without the slightest problem (and "no", we don't have too much time, we just have apprentices for the tedious OS-install tasks I'd also like to mention that on all machines the following onoard components have been disabled: USB controllers, serial ports, parallel port, sound, gameport, SATA controller and SATA RAID controller-bios. Maybe someone can shed some bits of light on this? Unfortunately FC3 is still "young" so there is not much to find on the net. Many thanx in advance

'lo there Got the prob fixed with FC3 hanging after enabling swap partition. FC3 has disqualified a bit due to this behavior, but in case someone else encounters the same probs, here's the thread that helped me to solved the issue here ... Fedora Core boot probs revisited cu

thanx for the reply, matttah I'm not too sure about the particular keystrokes you are suggesting, but we did of course try to jump to different consoles - which did not work. In fact the systems weren't responsive at all, not even to a harsh CTRL-ALT-DEL combo In the meantime, I have investigated the matter further and found something on some old Fedora forums that looks exactly like my prob ... "Enabling swap space [ OK ]" -> halt From what I've read now it is related to the sendmail daemon blocking the system from finishing the boot sequence. I won't bash FC3 now before I've verified that, but if that is the cause then it is just as stupid as any BIOS that tells me "No Keyboard found ... press F1 to continue" I will continue to peek into that matter and keep you people informed cu

greetings zenarcher You are on the right track already. The file you need is .. <your_java_dir>/jre/plugins/i386/ns7/libjavaplugin_oji.so As crazykillerman already suggested you can install the plugin by creating a symbolic link in the firefox-plugins folder in your home dir or you can install it into the main plugins folder of firefox. This directory is just where you installed firefox, e.g.: /usr/local/firefox/plugins. Keep in mind that you _MUST_ use symbolic links (copying the file to the plugins folder will only give you trouble). Also note that if you link the java-plugin to the main plugins folder, everybody will have java-enabled by default. If you create the symlink in your home directory (~/.mozilla/plugins I think) only you will haveJava enabled. hope that helps

@taeuler Blackpage, just out of curiousity, what had you doing 3 stage 1 installs a year? The answer is quite simple: 8 servers and 18 workstations, and my super-geeky attitude to always furnish these "darlings" with the very latest and (if possible) the fastest and/or most stable OS ) It all started with one old P3/400, and out of curiosity I made a stage-1 install that took around 3-4 days (including KDE, KDevelop and what not else). The performance-result was just amazing (for P3/400), and that lured me into the idea of setting up more Gentoo systems. In the aftermath I set up an AMD XP3000 box as intranet/fileserver, and one of my workstations (after all: I wanted such a fast system for myself too . All these Gentoo-setups are still running, or are valid OS' in multi-boot PCs, and if I had the time, I would even do a stage-0 install, cause, like you, I also happen to find the things one learns during a Gentoo-installation, very helpful for other admin-tasks. And btw: I lurve Gary Larson, so it's a nice thing to see that Larry the Cow has escaped the "wildlife preserved"

greetings jarves Yes, there is of course disk quota support in linux, as this is in fact a feature its Unix heritage. Disk quotas in Linux can be set in very fine granulated way and include all the fancy stuff like user and group-specific setting, as well as "hard/soft quota" or "grace periods". Setting up quotas requires 2 main steps ... 1: enabling quota support in the kernel You'l find the checkbox for quotas under the "File systems"-section when you do a "make menuconfig" or "make xconfig" 2: installing the Quota-Tools Which can be found here ... Quota-Tools on Sourceforge Compile and install these. Those are the preliminaries ... now it's on to configuring disk quota. The complexity of this task depends on how specific you want the quotas to be. A good guide as to how to configure quotas can be found here ... Quota-howto on redhat.com hope this helps

Howdy folks I have been messin' round the last days with Yoper, and althought this distro truly rocks, there is one thing I can't get fixed: menu transparency. No matter how often I disable GUI-effects or the menu transparency itself, it get's miraculously enabled again. I have now "bypassed" the transparency effect by "patching" the theme file and using a PNG as menu-background, still though: any ideas anybody as to what might cause this strange persistency of "enabled transparency"? Info: YOS v2, KDE 3.3, XFree (I suppose XFree 4.4.0) thanx in advance

@danleff Thanx a lot dude, I only found the cutesie little "Reference"-link in your posting today. Indeed this thread answered the thing: if you operate as root you are stuck with transparent menus. Normal users can disable the effect. As during the setup of a machine I usually work as root, I didn't even think about this possibility. Thanx again for providing this link!