Re: [Cooker] KDE3 WHAT A MESS
On Thu, 2002-04-04 at 05:29, Joe Simon wrote: Just installed Mandrake's KDE3 packages. This must be the worst yet! 1) No menu's. 2) Installed in /opt rather than /usr (KDE2 is in /usr). The packages are not relocatable. What if someone does not have fs /opt or enough space on / ? 3) ksplash seg faults just right after login. 4) Many other seg faults. Is this just a Mandrake problem or KDE problem? How can kde.org say the following on their web site: On April 3rd 2002, the KDE Project released KDE 3.0, the Third-Generation of the Leading Desktop for Linux/UNIX, Offering Enterprises, Governments, Schools, and Businesses an Outstanding Free and Open Desktop Solution. Can you imagine students in grade 7-9 getting all these seg faults and not having any menu's? I am disappointed! _ Join the worlds largest e-mail service with MSN Hotmail. http://www.hotmail.com Well, /usr is for software integral to the system, /opt is for add-on software according to FHS 2.2. KDE2.2.2 is part of the system, and KDE3 is regarded as an experimental add-on at this time. If you don't have enough room to make an /opt in /, then mkdir -p /usr/opt ln -s /usr/opt /opt It isn't yet ready for newbie time by a long shot, but you know the clamor when we delay these releases. Turn up your messaging to warning level to catch the KDE warnings and you will get a clue to the segfaults--most are in reference to opening sockets. Drop your security level to 2 with msec 2 and the most of the crashes disappear. Perhaps the KDE team was concerned with desktop and not thinking of settable security or perhaps they decided the distros should make the permissions available (which is going to take some distro-centric engineering) Sound may work on your computer without the sound server in KDE3 functioning at all. There is little new in this. The linux kernel was not really designed for sound, and the problems multiply from that. Civileme
Re: [Cooker] please help.... menudrake is killing me
On Thursday 14 March 2002 10:08 pm, SI Reasoning wrote: --- r00t_tty [EMAIL PROTECTED] wrote: i do appologize for that... here is the symptoms.. mdk 8.2rc1 kernel compiled by me.. off of the 2.4.18-5mdk source rpm -qa |grep menu menu-2.1.5-99mdk mkbbmenu-0.2-2mdk menudrake-0.6.5-2mdk fluxbox as the desktop manager test... simply open menudrake on any window manager and attempt to move a menu tree from one to another... i've reinstalled rc1 several times in hopes that it was one of the libs or binaries i installed.. this has been happening since beta1.. i have spoke w/ others that are having issues as well.. i do have a coredump but it's a 2.5mb bz2 file.. too big to send.. i will post it to www though ... http://webpages.acs.ttu.edu/bgrace/core.tbz2 ... thanks in advance for you even looking @ it.. Using cooker I successfully moved the /multimedia/graphics directory to the /configure directory and back again. No issues. I am running this from a thin client from my X term server. This cooker was updated from 8.1 also which has a few problems that my direct cooker install does not have, but not this one. menudrake-0.6.5-2mdk menu-2.1.5-100mdk fluxbox-0.1.7-3mdk Brandon On Wed, 2002-03-13 at 17:02, Frédéric Crozat wrote: Le Wed, 13 Mar 2002 07:42:56 +0100, r00t_tty a écrit : keep getting coredumps when trying to move one menutree to another... How do you want me to try to fix your problem with such a bug report.. You don't give menudrake version neither a reproducable testcase .. -- Frédéric Crozat MandrakeSoft = SI Reasoning [EMAIL PROTECTED] A requirement of creativity is that it contributes to change. Creativity keeps the creator alive. -FRANK HERBERT, unpublished notes __ Do You Yahoo!? Yahoo! Sports - live college hoops coverage http://sports.yahoo.com/ Not able to reproduce your problem here... Tried dual PPro with 160M and a /usr split in RAID0 between a scsi and an ide drive, running RC1 VIA C3 with 256Mb and 40Gb Maxtor with Beta 4 Duron 750 with 128M and 15G Beta 3 Athlon 800 with 512M and RAID0 arrangements on all partitions except /boot Beta 2 K6-2 350 with 64M and SCSI/IDE drives RC1 (updated from 8.1) P2-300 with 192M and 12G Hdd. 8.2b1 Every single one did just fine. What is different about your machine? We really need to know, because no one should be seeing this behavior. Civileme
Re: [Cooker] 8.2 Beta laptop power management
William Bouterse wrote: On Sat, 2 Feb 2002 20:53:26 + OS [EMAIL PROTECTED] wrote these words: Yes ! Me and now several others have started threads on more than one mailing list stating that the old Linux power management was bad, but the new systems (APIC and APMD beta 3) REALLY suck !!! What really intriques me is how Linux is being positioned as the OS for all platforms, and yet locks completely solid on platforms that require any kind of power management that doesn't involve init 6 !!! I am wanting to deliver a Mandrake 8.2 Linux based laptop system to an important client. This machine will show off Mandrake and Linux to people who have probably not even heard of Linux, or at least certainly not considered using it. I can't deliver it because it locks solid when ever a power management feature is invoked !!! I am seriously faced with having to put a Window$ on to it. This is a real shame, but it is a reality. Owen On Friday 01 Feb 2002 3:05 am, you wrote: I installed 8.2 beta on a Dell Inspiron 8000. When the screen attempted to dim the machine powered down. With a fresh 8.2 install power management still doesn't work. apmd is running but any change in power status (unplug for instance) causes the machine to hand. This worked fine in 8.1 darrell What make and model is the notebook you plan to use ? Is it a Dell as well ? Have you attempted to install LM 8.1? If they have never seen/used linux anyway an older version which works might be an alternate approach !? Don't give up yet !!! William Bouterse Talkeetna, Ak Well, the issue of APIC is complex. Dell at least fixed the BIOS bug that was providing bad bridge mapping errors in the inspiron 4000 above 192M RAM. They still have issues with ACPI and APIC both. It takes special dodges with the hardware to make all that stuff fit in a tiny box, so standard drivers often need modification to work in notebooks. Since the notebook makers install windows in their notebooks, they make their own drivers to compensate for the non-standard solutions they come up with to make the notebooks work in the first place. If they were a little more forthcoming about their dodges, linux drivers might work as well. But this is a situation where we have a conflict. Without a UP install kernel with APIC, many multipe-processor installs won't work. There are also rigs with multiple IDE controllers that need APIC even with the UP platform. And there is a small, intensely working group of engineers trying to keep up with all the kernel changes and make them workable who cannot make unnecessary forks in the kernel. So a simple report about APIC is quite sufficient. Now the problem of still behaving the same way when linux noapic is the kernel message is significant and may suggest other issues as well for the platform. There are other notebooks which do not work correctly with any of our stock kernels at all. Some need linux noauto nousb noapic for example. Some yield to nothing we have tried for 8.0, 8.1 and 8.2beta. Without samples and some tech info from the manufacturer, our chances of making notebooks work out of the box aren't as good as they could be otherwise, Civileme
Re: [Cooker] XFS is needet! [Was: AbiWord]
Alexander Skwar wrote: So sprach »Pixel« am 2002-01-30 um 10:56:20 +0100 : remove package basesystem and this limitation will go away... Is lilo needed at all if I use grub? If not, then both grub and lilo should provide bootloader and basesystem should require bootloader and not lilo and/or grub. Alexander Skwar Ummm, and economics rears its ugly head You see, if you do a --whatprovides query, you should get a unique answer in mdk. Otherwise it is a nightmare when a tarball changes--have to repackage all of them that provide it, and the distro gets _really_ bloated in terms of install CDs. In fact I wrote a tarball sweeper to check that tar.bz2's are not duplicated with a few exceptions in the distro. (Like ALSA in several kernels). Civileme
Re: [Cooker] please disable APIC in kernel
SI Reasoning wrote: --- Neuromancer [EMAIL PROTECTED] wrote: Thus spake SI Reasoning about Life, the Universe, and Everything: | I disagree, | first, this is a major problem as it completely | freezes the box and it has to be turned off manually. Then turn it off manually. To do that would require re-compiling ther kernel, something that most newbies should not be forced to do. I have tried recompiling this kernel (and I have done this many times before) but even that is broken at this point. Regardless, this is still broken. I would be surprised if this was only a Dell issue. I don't think it is smart to release something that has such drastic outstanding issues as part of the standard kernel. I can't imagine that its benefits outweigh having laptops freeze. Understand that this has not been an issue in the past since this was not enabled in the standard kernel. Now that it has been enabled IT IS AN ISSUE. If this was just cooker it would not be such a big deal, I expect things to get tested here, break, get fixed, get broken again. We are talking about a final release kernel for 8.2 though, and I think that it should err on the side of stability. | second, when you release a distribution it needs to be | able to run on as many boxes as possible without major | crashes, if you can turn off an item and it creates | more stable boxes, then you are less likely to lose | people, especially newbies Yell at Dell. Not Mandrake. There is no yelling going on, Dell's reputation is not at stake here but Mandrake's is. This is something Mandrake has control over. | thirdly, this is a feature that more experienced users | will take advantage of and they are more than | qualified to rebuild their kernel. Mandrake is geared | towards user friendlyness and having your system crash | so easily is not user friendly. So far, it appears that there is only one person complaining about this. And to the wrong folks, it seems. I am not the only person who has brought this up here, but I am talking to the right people. This is Mandrake's standard kernel and there is something that can be done in the short run for the 8.2 release. After 8.2 is out, it can be re-enabled and we can try and figure it out if APIC is such a great thing. I really have no idea what it does besides freeze my laptop. | Because the Fn key is right next to the Ctrl key, | there has been quite a few times that I am trying to | change desktops and end up freezing the system. Eh? On the dell laptop the Fn key is right besides the Ctrl key, Fn + F1 = Bios Setup( or freeze with APIC), FN + Esc = Suspend, Fn + F3 = battery display... all of these freeze when hit and it is not that uncommon for a finger to slip to the Fn instead of Ctrl and create a hard freeze. | Finally, this feature has recently been turned on in | the kernel, prior kernels, esp in the 2.4.16 and prior | did not have this turned on. It is not an improvement | when the stock kernel creates more problems as it | progresses forward. Let the maintainers of APIC | resolve some of these issues and make this feature | more stable across many different types of boxes | before it becomes part of the default kernel. I'm pretty sure that's what they did. Wasn't it labeled as experimental until recently? If it's no longer labeled as experimental it means they have found and squashed all KNOWN bugs. You may have found another. Call Dell. -Russ -- If the auto industry were like the computer industry, a car would now cost $5, would get 5000 miles to the gallon, and at random times would explode, killing all its passengers. = SI Reasoning [EMAIL PROTECTED] There never was a good war or a bad peace. Benjamin Franklin __ Do You Yahoo!? Send your FREE holiday greetings online! http://greetings.yahoo.com Do you have any idea how many people submitted bug reports bexcause APIC was off and their machines would not boot the installer? Almost everyone with multiprocessors and more than 2 IDE channels. Obviously we are trying to see what the effects are each way, and there might be an alternate install image, but the problem is that people never look for those. Civileme
Re: [Cooker] please disable APIC in kernel
SI Reasoning wrote: --- tester [EMAIL PROTECTED] wrote: Do you have any idea how many people submitted bug reports bexcause APIC was off and their machines would not boot the installer? Almost everyone with multiprocessors and more than 2 IDE channels. Obviously we are trying to see what the effects are each way, and there might be an alternate install image, but the problem is that people never look for those. At least we know where the problems are. If there are no other major issues then it sounds like it should be enabled for the smp kernel. Hoever, more than 2 ide's is not as standard as a laptop and I think it should be disabled in the standard kernel. This looks like something that may be best setup as a module at some point. It _is_ enabled in the SMP kernel but it is also a UP kernel to install with and it strill fails. As a matter of fact our failure rate on laptops is very high, APIC or not. That is because there are a lot of dodges employed by laptop makers and compatibility with standards is restored by software they write for windows. If you have ever tried to install windows on any laptop without the restore disk or restore CD, you know well of what I speak. And the particular one you speak of is best served by an alternate install kernel and an alternate running kernel because there is a huge BIOS bug as well. Try to run kernel 2.4 from anyone with 512M of RAM and you will see bad bridge mapping errors. The APIC error is also a BIOS bug--have seen it before on an inspiron that was trying everything to run kernel 2.4 with 512M. He ended up running kernel 2.2 and writing to Dell. Civileme
Re: [Cooker] 8.2 forlorn hope
Ron Stodden wrote: Both 8.0 and 8.1 on a full expert and full network install completely neglect to install the nfs-clients RPM, thus disabling any attempt at nfs-based networking. Well, nfs is broken so readily by a Reiser partition at either end that a default install is a bad idea. It _can_m be installed from individual package selection rather easily. 6.1 completely fails to install wheel logic for a PS/2 mouse with a wheel. 8.0 could be persuaded to by post-installing imwheel, but there is no imwheel in the 8.1 distribution. Strange, my wheel mice work fine. In fact X used to kill the wheel stuff that install set up (in 8.0) and people had to curse and fix it. 3-button emulation and imwheel are not good playmates. 8.1 will not run on one of my machines my machine with a Soltek motherboard. All that happens is a long loop involving the ASCII penguin. Interesting... What steps did you take to remedy the situation? Alternate install images? linux noauto? devfs=nomount? Specifying memory size? linux nobiospnp? Turn off ACPI? Turn _on_ PnP? With kernel 2.4 strange things happen with some BIOSes that aren't standard. Kernel 2.4 was made to take things from the BIOS and the first thing I noticed is that several BIOSes send a bad memory map or bad device map to the kernel for certain sizes of RAM. This is not going to get fixed without a newer kernel and in some cases newer BIOSes. Mandrake should mot countenance any thought of 8.2 until 8.1 is finished. The above are all killers. You must work for someone who has too much money on his hands. Economic realities dictate otherwise. Civileme
Re: [Cooker] why no 'install everything' option?
Dan Mack wrote: I've been running cooker on 3 of my systems for a few months now and I find myself missing an option that Redhat has had for a while and that is the 'install everything' checkbox in the install tool. Even after checking every single box on the select software scrren of Mandrake, things like 'unzip' aren't even installed. This gets a little tedious, especially if you have a big server with plenty of space and you don't want to traverse 20 submenus just to get all the stuff you need to use the system effectively. Thoughts, comments? First of all, MAndrake has many more packages than RH. Second, some packages conflict with each other and cannot be installed side-by side (Glide libraries and Zope servers, for example). And please don't tell Mandrake to pick. You know no matter what they pick, it will be starting a holy war. Last I looked, recommended install didn't install emacs which is my favorite IDE, and second favorite desktop... But during install (individual package selection) there is a little symbol at the bottom which looks like two blue arrows chasing each other -- click that -- Presto! A flat list comes up so you can pick your favorite packages in collating sequence. Civileme
Re: [Cooker] Disk Drake Winxp partition
Roger wrote: Just to advise you guyz, winxp's uses a new thang for it's root partition (mbr or something...i forget the specifics but i have some docs on it incase you don't know what i'm talking about). If you try to resize winxp's partition (the partition containing it's root filesyste), you can damage this partition beyond repair. Partition Magice =6.0 is able to resize winxp's root partition ok. Other partitioner's are also becoming winxp complaint. My question is, can diskdrake resize the fat32 winxp partition without loosing this data? If it's not winxp complaint, there probabely should be either a strong warning not to resize a winxp partition (as i saw known when i selected the winxp and resize option (only one stating backup data) or a resizing for a detected winxp root filesystem should probabely be deployed. I have been testing resizing of winXP FAT32 partitions--no problem yet. Please do report any problems you encounter, because we haven't found any. Yes LILO with 8.1 even seems top work fine with WinXP dual-boot. Civileme
Re: [Cooker] KDE2 vs KDE3 in clean install
Vincent Meyer, MD wrote: Hi, I did a clean install last night on two of my machines, installing both KDE2 and KDE3. On both machines startx brings up KDE3. Startkde tries to bring up KDE2, but bombs because the x server isn't up yet. On one machine I have the KDE login manager, and there are selections for KDE, sawfish, default. The only KDE entry brings up KDE3. contents of /etc/X11/wmsession.d has 01KDE which is KDE2, 08Sawfish, and 11KDE, which is KDE3. Fixes and suggestions? Changelogs I thought over last few weeks reflected changes to allow them to co-exist peacefully V. What happens if you install Xtart and come up in console login and run it? Civileme
Re: [Cooker] Abit KR7A-RAID
Eugenio Diaz wrote: Well, I finaly finished building my new machine (lots of tie wraps, cable routing, etc.), an Abit KR7A-RAID with dual ATA133 60GB maxtors on a stripe. But I just found that Linux do not support the HPT372 on board raid controller :) Double Damn the assholes at High Point!!! I bought this board with the impression that High Point was a Linux friendly company, since when I was doing my light pre-shopping research I saw a big prominent Linux Logo on their page!!! Can Mandrake put pressure with Abit on High Point to release specs, or whatever is needed for the drivers to be merged into the kernel, and not those stupid binary drivers for old Linux distros ... Do you guys know if it is posible to take the RH7.1 drivers they provide, and cram them (some how disabling the versioning of symbols) into an initrd to make a custom Cooker install discs? BTW, this message was copied to [EMAIL PROTECTED], so if you reply, do a reply all, so that they can see how stupid they are ... my immediate reaction was to return the board, but since I invested so much time building the box nicely, I am just going to disable the raid controller, and install on the normal ATA100 channels. THANKS HIGH POINT = Eugenio Diaz, BSEE/BSCE Linux Engineer [EMAIL PROTECTED] __ Do You Yahoo!? Send FREE video emails in Yahoo! Mail! http://promo.yahoo.com/videomail/ Ummm actually, Just make your windows RAID cover less than the whole disks, then use linux software RAID on the rest. The HPT controller, like the Promise FasTrak and the CMD and a couple others are in fact software RAIDs (just a BIOS extension) and expressly designed to add RAID to windows, at a surcharge of $30 US or more for their fake hardware RAID controllers. There is support in the kernel to READ the RAID sectors on the controller if you are dual-booting, but writing is a different story, since linux software RAID has equal or better performance and beaucoup flexibility (like RAID 4 and 5 as well as 0 and 1, and arbitrary physical extents on arbitrary physical volumes making logical volumes). If you really want RAID drivers, there is a chance that the stuff at this site will compile for you http://people.redhat.com/arjanv/pdcraid/ And the Duke of URL was also taken in by the fake hardware, writing an article to show how to make a RH driver work for Other distros--just search his site. I think that one is Promise specific. You _really_ know it is software when there have to be different driver versions for UP and SMP. There is of course the possibility of returning the hardware and finding something else. There are two real RAIDs for IDE, one of them OS-Transparent, and you can find them at linux-ide.org. This link mught also be of interest: http://www.linux-ide.org/endorsements.html ARCO makes the pure hardware RAID controller for IDE, and Promise has the SuperTrak, but I have not heard anything from any applications save one member of the Alaska Linux Users Group who reported success with Mandrake but didn't give details. Civileme
Re: [Cooker] System collapsing after a few hours!
Sergio Korlowsky wrote: I have sent this info twice, hope this time I can obtain a least a 'simple response' from you guys! We test 'cooker' you come up with solutions to our problems for the upcoming release of Mandrake-Linux 8.2 and as always we want it to be the best.\\!! After a few hours as the subjects says, KDE collapses and re-boots back to the 'login' screen by itself, is it caused by the screensaver: Jan 3 08:38:44 sedecomp kernel: Out of Memory: Killed process 4001 (krozat.kss). or maybe a memory leak in another application? just tell me what log file to send, and I will, as I consider this a very serious case. This behavior begun after installing the 20011224-cooker-snapshot, and later upgrades. Am I the only one experiencing this? that would help to find out what is wrong with my box. Athlon 800 512M MOBO ECS K7S5A Voodoo 5 5500 Maxtor 40 G IDE Acer 12x8x32 CDRW uptime 4 days 23:57 KDE running XCDroast running Mozilla Running Not my main machine-- free shows 386645k used as cache I was using 16M buffers for xcdroast Offhand, I would say look at the video drivers. Civileme Pentium-IV 128Mb RAM MoBo: Intel D-850 Nvidia drivers TNT2-64 w/32mb During the time using 8.1 on the same box, had no problems... during this time I have found a kdeinit process using as much as 98% of the CPU resources, have not been able to identify what process is this that it is consumming all system resources. Looks like a memory leak somewhere. Jan 3 08:10:00 sedecomp CROND[5151]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:15:00 sedecomp CROND[5159]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:20:00 sedecomp CROND[5167]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:25:00 sedecomp CROND[5175]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:30:00 sedecomp CROND[5183]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:35:00 sedecomp CROND[5191]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:38:40 sedecomp kernel: Out of Memory: Killed process 2102 (httpd). Jan 3 08:38:40 sedecomp kernel: Out of Memory: Killed process 2103 (httpd). Jan 3 08:38:41 sedecomp kernel: Out of Memory: Killed process 2104 (httpd). Jan 3 08:38:41 sedecomp kernel: Out of Memory: Killed process 2104 (httpd). Jan 3 08:38:43 sedecomp kernel: Out of Memory: Killed process 2105 (httpd). Jan 3 08:38:43 sedecomp kernel: Out of Memory: Killed process 2105 (httpd). Jan 3 08:38:44 sedecomp kernel: Out of Memory: Killed process 4001 (krozat.kss). Jan 3 08:38:44 sedecomp kernel: Out of Memory: Killed process 4001 (krozat.kss). Jan 3 08:40:00 sedecomp CROND[5206]: (root) CMD ( /usr/sbin/monitoring.pl) Jan 3 08:41:39 sedecomp kernel: Out of Memory: Killed process 5198 (httpd). Jan 3 08:41:41 sedecomp kernel: Out of Memory: Killed process 5199 (httpd). Jan 3 08:41:43 sedecomp kernel: Out of Memory: Killed process 5200 (httpd). Jan 3 08:41:45 sedecomp kernel: Out of Memory: Killed process 5201 (httpd). Jan 3 08:41:45 sedecomp kernel: Out of Memory: Killed process 5202 (httpd). Jan 3 08:41:47 sedecomp kernel: Out of Memory: Killed process 5203 (httpd). Jan 3 08:41:49 sedecomp kernel: Out of Memory: Killed process 5204 (httpd). Jan 3 08:41:51 sedecomp kernel: Out of Memory: Killed process 5213 (httpd). Jan 3 08:41:51 sedecomp kernel: Out of Memory: Killed process 3881 (kdeinit). Jan 3 08:44:39 sedecomp kernel: Out of Memory: Killed process 5214 (httpd). Jan 3 08:44:41 sedecomp kernel: Out of Memory: Killed process 5215 (httpd). Jan 3 08:44:43 sedecomp kernel: Out of Memory: Killed process 5216 (httpd). Jan 3 08:44:45 sedecomp kernel: Out of Memory: Killed process 5217 (httpd). Jan 3 08:44:47 sedecomp kernel: Out of Memory: Killed process 5218 (httpd). Jan 3 08:44:47 sedecomp kernel: Out of Memory: Killed process 5219 (httpd). Jan 3 08:44:49 sedecomp kernel: Out of Memory: Killed process 5220 (httpd). Jan 3 08:44:50 sedecomp kernel: Out of Memory: Killed process 5221 (httpd). Jan 3 08:44:52 sedecomp kernel: Out of Memory: Killed process 3870 (kdeinit). Jan 3 08:45:01 sedecomp CROND[5228]: (root) CMD ( /usr/sbin/monitoring.pl) SK
Re: [Cooker] [RPM] fluxbox-0.1.5
Han wrote: Han ([EMAIL PROTECTED]) wrote: Don't try to understand'em, just throw and rope and brand'em Ok folks here is the much more official Mandrake RPM. * Sat Dec 29 2001 Han Boetes [EMAIL PROTECTED] 0.1.5-2mdk - applied a fix for bbtools; will be included in 0.1.6 - added menu system. (ripped it from bb) - I didn't add the update-alternatives system yet. Gotta get positive feedback first. So far I vote for the update-alternatives from Alexander, but since blackbox not my package I would like to hear you all about it. I ripped the menusystem from blackbox. It _should_ work. But since I don't use the menusystem I have no way to realy test it. If anybody could test it for me I'd be very happy. BTW rpmlint goes berserk on this rpm (just like the bb-rpm) Something for the next release ;) I just gave it the /etc/X11/wmsession.d/99fluxbox If someone has a better idea about that number please post it. Oh, as allways, you can find my rpms here: http://www.xs4all.nl/~hanb/software/ Cya, Han. try 16 for the number--it won't tread on any others and Xtart will still run as well as chksession. Civileme
Re: [Cooker] Re: System Freezes
Dave Seff wrote: Ok, Not to reopen an old wound, But I think I've narrowed down this Freeze thing. Now my machine just froze a minute ago. I could, however, ssh from one of my other machines. and running top indicated that X was a runaway process. If I kill it (-9 is the only sig I could use to kill it), The systems goes into deep freze. Machine lockup, mouse froze, ssh session froze (Kernel panic?) another time this happened I ssh'ed in and ran strace against X. It was looping SIGALARM over and over, and over. I have a K7T266 Pro2 with an Nvidia GeForce2. (Nvidia's latest driver) So what are my options at this point? Is this X's fault, or NVidia's? -Dave Heeheehee Someone on the expert list mentioned thisd within the past week--his instability problems disappeared when he reloaded the older (1541?) driver. Civileme