On Sun, 2008-11-16 at 16:30 +0100, Mads Kiilerich wrote: > Oh. Please provide constructive feedback and help getting the problem > solved. I think I've been doing that. I also was just venting after a really frustrating experience... > > Christopher A. Williams wrote, On 11/16/2008 03:25 PM: > > Installation was a nightmare (as witnessed by a friend who also > > frequents this list). The main thing that was wrong was video. Simply > > put, it just doesn't work > > Point taken. Looks like > https://bugzilla.redhat.com/show_bug.cgi?id=465634 (which also suggests > boot parameter "xdriver=vesa" as temporary workaround) > > Let's hope it gets fixed. Indeed. This looks like the one. > > > 2) Running the autologin loads a session, which immediately goes to a > > blanked out white screen after trying to load 2 "untitled" windows (not > > sure what these are), but just makes the desktop unusable. You must > > switch to runlevel 3, login as root, load X, and then install. > > > > So this is with vesa? It looks like a separate problem. But I don't > understand exactly how you get it. And what you call "must" is a > workaround that worked for you? I suggest you file this issue in bugzilla. Just to clarify - yes, this is a work around that worked for me. Also, yes this is with vesa loading as the video driver. My "fixed" USB drive will boot. Letting it automatically logon loads the desktop, but something else is also starting that eventually causes the screen to go white. I think if we fixed the first problem, the second might not be an issue. Whatever is loading apparently doesn't load when you login as root. > > > 3) Prior to installing, the hard disk MUST BE MANUALLY PARTITIONED with > > fdisk such that ALL partitions are removed. If partitions exist on the > > drive, the installer fails when trying to re-partition the drive. > > That is not what other testers see. How did you see the failure? Any > exact messages? How did you reach your conclusion? I guess I'm unique then. I'm sure I don't see things other testers do. That said, I can't remember the exact error message, but it happens when partitioning the drive when clicking the "Write Changes to Disk" button. The system then responds that there was an error with partitioning, that it "was serious", and that the install can not continue. If I delete all partitions, reboot, and then run the install the same way, clicking "Write Changes to Disk" works normally, and the install continues without further incident. > > > 4) Something in the installation process takes a REALLY LONG TIME to > > finish (I mean 20 minutes or more). I later found out that it has > > something to do with kernel install / updates. > > > ... > > First, run all of the rawhide patches. ANOTHER PROBLEM - the kernel > > update alone took over 20 minutes to complete. Why??? It only seemed to > > move forward after I put selinux into permissive mode - at which point > > several hal oriented messages came through. I think this is also why the > > last part of the install took 20 minutes (looking like it was dead at > > the time). > > > > That is not what other testers see. Again, I apparently don't see things they do, but this is clearly reproducible. > Could you try to run something like > mkinitrd -v -f /boot/initrd-2.6.27.5-109.fc10.i686.img > 2.6.27.5-109.fc10.i686 > and see if that takes 20 minutes - and which part of it that takes time? Good Call!! Indeed, this is what is apparently taking forever and a day to complete. Here's the key output: # mkinitrd -v -f /boot/initrd-2.6.27.5-109.fc10.x86_64.img 2.6.27.5-109.fc10.x86_64 Creating initramfs <---------THIS ONE TAKES A REALLY LONG TIME!!!!!!! get_netlink_msg returned No buffer space available <--IS THIS WHY? Found root device /sys/block/dm-0 for /dev/VolGroup00/LogVol00 Looking for driver for device /sys/block/dm-0 Looking for driver for device sda2 Looking for deps of module scsi:t-0x00 Looking for deps of module pci:v000010DEd000007F0sv000010DEsd0000CB73bc01sc01i85: pata_acpi ata_generic <----------------- THIS ONE TAKES A REALLY LONG TIME TOO!!! get_netlink_msg returned No buffer space available <--IT HAPPENED AGAIN! Found swsusp device /sys/block/dm-1 for /dev/VolGroup00/LogVol01 Looking for driver for device /sys/block/dm-1 Using modules: pata_acpi ata_generic ... (everything else runs as fast as you would expect...) /etc/ld.so.conf.d/mysql-x86_64.conf -> /tmp/initrd.yFK3b5/etc/ld.so.conf.d/mysql-x86_64.conf /etc/ld.so.conf.d/atlas-x86_64.conf -> /tmp/initrd.yFK3b5/etc/ld.so.conf.d/atlas-x86_64.conf Running ldconfig ...and it finishes! > > > * Flash works kind of - there's no sound. > > > > 64 bit machine or Flash 9? > http://docs.fedoraproject.org/release-notes/f10preview/en_US/What_is_the_Latest_on_the_Desktop.html#sn-Web_browsers 64-bit, F10, Flash 10. Installation per the famed release notes. No sound in Flash. Tried other suggestions from the branched thread. Still no sound in Flash. > > * Speaking of sound, the sound tests work, but I don't actually GET any > > sound to play until a few seconds after something tries to play. Thus, > > the login sound never gets heard, neither does the logoff sound, etc. > > > > File a bug. OK - I'll happily do that. -- ================================== By all means marry; If you get a good wife, you'll be happy. If you get a bad one, you'll become a philosopher. --Socrates -- fedora-test-list mailing list fedora-test-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
