Re: reboot cycle, no log messages
In case someone searching for something finds this thread (or I find myself forgetting things I've done again) and the following information helps: I ended up filing a bug, and in the process thought of checking a few more combinations. Checking those led me to discover I had /dev/sda (n) and /dev/sdb(n) reversed on the partitions/file systems that were actually causing complaints. Specifically, it was not the LVM partitions giving me the problems, that was just where the messages were making it to the screen. It was me trying to mount /dev/sdb4 instead of /dev/sda4 (or was it the other way around?) in /etc/ fstab . (who'da thunk? especiallly with the zero-length partition thing 8-p) (I left the bug there because the error messages could be improved, but my problems are solved.) On Jan 1, 2010, at 5:00 PM, Joel Rees wrote: Just before it reboots, it gives me messages about old ext3 partitions that have problems mounting. [...] This is what I captured with a digital camera when it cycled: /dev/mapper/fc7-7[various]: clean [long list of partitions from the old system] /dev/mapper/fc7-7varwww: clean, 644/516896 files, 27268/516896 blocks [failed] *** An error occured during the file system check, *** Dropping you to a shell; the system will reboot *** when you leave the shell. *** Warning -- SELinux is active *** Disabling security enforcement for system recovery. *** Run 'setenforce 1' to reenable. sulogin: error while loading shared libraries: libfreebl3.so: cannot open shared object file: No such file or directory Unmounting file system Automatic reboot in progress. --- [...] Setting the passno entry in /etc/fstab to 0 or commenting out the entry just pushes the error message back. When a large partition at /dev/sdb4, which I use for backup, is the final partition checked in /etc/fstab, it gives me -- fsck.ext3: Attempt to read block from filesystem resulted in short read while trying to open /dev/sdb4 Could this be a zero-length partition? -- and then the messages about libfreebl3.so and the automatic reboot. yum provides libfreebl3.so tells me that nss-softokn-freebl is installed. Trying an erase to re-install it tells me that most of the OS and most of the apps seem to be dependent on it, one way or another. Hmm. This looks like I need to file a bug or two. Anyone know what's happening here? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Best way to get minimal system
On Jan 5, 2010, at 7:22 AM, Chris Smart wrote: [...] Ubuntu provides an "Install a command-line system" mode on the alternate installation media. Does Fedora have something similar? I didn't see any Anaconda options to do it. [...] I was wondering that, myself, but I think when I installed F12 last month I noticed a similar option for Fedora. It was hidden rather well, however. (And I think I had seen it once before, back in, maybe, FC7 days.) Hopefully, I'll remember to watch for it and report back when I upgrade (re-install to current) my rescue system partition later on. One note, for command-line only stuff, I sometimes prefer openBSD. Depends on what and why. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: reboot cycle, no log messages
Okay, I had a little more time this morning, so, ... On Jan 5, 2010, at 6:11 AM, Joel Rees wrote: On Jan 3, 2010, at 12:51 AM, Roberto Ragusa wrote: Joel Rees wrote: yum provides libfreebl3.so tells me that nss-softokn-freebl is installed. Trying an erase to re-install it tells me that most of the OS and most of the apps seem to be dependent on it, one way or another. Hmm. Bypass yum. rpm -V nss-softokn-freebl or/and rpm -Uhv --replacepkgs --oldpackage nss-softokn-freebl.rpm Well, yeah, I'm expecting to use that as a workaround. (Thanks for saving me the trouble of trying to remember the --replacepkgs option. :-) Ended up using yum reinstall. Didn't change anything. Are you on 32 or 64 bit? If 64, maybe the libs are incomplete. 32 bit. Seems odd this hasn't bitten others yet. I'm hoping it doesn't mean the hard disk itself is dying, or that LVM has issues with my setup and is corrupting the file system. So I checked my command history, and apparently I still thought the missing file was libfreeb13.so when I searched for it before, rather than libfreebl3.so. (erk) So it was probably never missing, which is comforting to know. Sort of. Anyway, I ended up commenting all the old file system partitions out of /etc/fstab and that allows it to boot. That's okay, because I can mount them by hand once I'm logged in, and I probably should copy all the files I need in from backup instead of just making them available where they were. But I'm going to file a bug on this, I think. Whatever the reason it won't mount the old stuff during boot, it shouldn't try to drop me to a shell and then force a reboot because it can't load the file necessary to allow me to drop to a shell. I had to boot from a rescue install at one point because the reboot cycle would never end. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: I can't startup my Fedora 12
On Jan 5, 2010, at 2:18 AM, 严晶涛 wrote: I have only install Fedora 12. My partition is /boot 200M LVM 290G LVM: fedora 20G other 10G opt 60G home 200G swap 3G fedora is Fedora 12's / opt is /opt home is /home All the filesystem is ext4 Do you mean, including /boot ? My memory is that ext4 /boot is still not officially supported. Did you reboot once, just to check? Then I want to install CentOS in other,I backup /boot and format / boot to ext3(because CentOS 5.4 doesn't support ext4) and copy backup to /boot Install CentOS to other,and format other to ext3. Where is Cent's /boot? What happens to F12's /boot that you attempted to restore? When I have installed CentOS,I can login CentOS,but when I choose Fedora 12,my system will stop in fsck,and I can't startup it... Has Cent OS's /boot by any chance overwritten F12's /boot? That's kind of what I'm expecting happened. I think what you wanted to do was set the host (MSDOS) disk label as /boot 200M: ext3 (I would name it "Fboot" or something in diskdruid, to make it easier to remember.) other 10G: ext3 (I like 40G or so, to play in, and I would probably name it "cent" or "otherOS") LVM 280: LVM In other words, make the entire Cent OS file system a single MSDOS level partition. Hmm. I just checked the current partitioning instructions. I think I'd better go raise a fuss or two. Anyway, the above should work, and allow you to do something like what you are trying to do. Recognize that, in the above, all of cent's file system, including "/" and " /boot", would reside in the "other" partition. (I'm not sure about the LVM interactions, or I'd suggest something like separate "Fboot" and "Cboot" MSDOS-level partitions for Fedora's /boot, and Cent OS's /boot, respectively.) Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: reboot cycle, no log messages
On Jan 3, 2010, at 12:51 AM, Roberto Ragusa wrote: Joel Rees wrote: yum provides libfreebl3.so tells me that nss-softokn-freebl is installed. Trying an erase to re-install it tells me that most of the OS and most of the apps seem to be dependent on it, one way or another. Hmm. Bypass yum. rpm -V nss-softokn-freebl or/and rpm -Uhv --replacepkgs --oldpackage nss-softokn-freebl.rpm Well, yeah, I'm expecting to use that as a workaround. (Thanks for saving me the trouble of trying to remember the --replacepkgs option. :-) Are you on 32 or 64 bit? If 64, maybe the libs are incomplete. 32 bit. Seems odd this hasn't bitten others yet. I'm hoping it doesn't mean the hard disk itself is dying, or that LVM has issues with my setup and is corrupting the file system. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Fedora Basic End User Rollout Support Operation
On Jan 2, 2010, at 4:33 AM, Robert E. Martin, VCM Network wrote: [...] Fedora has many advantages, but in its raw state it is not very user friendly. Actually, in its current state, it is very user friendly to those not steeped in the One True Microsoft Way. Except for system upgrades, and the high churn rate. I am looking to talk to someone who knows has to create a user friendly approach to Linux that can be effectively pitched to the consumer and SMB market. Would it be out of place for me to suggest that the "user friendly approach" may be more of an exercise in providing a Rosetta Stone, of sorts, to help people familiar with the Microsoft (cough) Way to figure out what they really wanted to do and what the tools to use are called? If you're looking for a project that throws an MSWindows veneer over a Linux distribution, such do exist, at some times and in some places. I forget what any of them have been called, though. Maybe someone else here remembers? My business model will simply be to set-up, configure systems, assist with package application selections and train this market for competitive price. Can someone point me to the right person or people who are doing it or potentially have the know-how? centos.org has been mentioned. Have you talked to the people at Red Hat? (You know, the ones who sponsor the Fedora Project.) BTW, I agree with several who have suggested that Fedora is not appropriate. It's kind of a playground for geeks and hobbyists, as well as a testing ground for some of Red Hat's enterprise products. It can be induced to work for small businesses, but I think the word you want to hear about that is "customized". By which I mean you have people building true custom systems. (And I don't mean packages customized to a market niche. I do not mean packages, at all.) And, of course, there are Ubuntu and other non-Red Hat distributions that have slightly different focus. Joel Rees, free-associating a little -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Installation plays hardball
Yes, and I did, mostly with custom, over and over again. Id est, I tried to increase the size of /boot any way I could, and never found any way to add a single byte. Are you talking about increasing the size of /boot before or after the install? (I can't tell.) Incidentally, my memory is that /boot will not be managed by LVM under any ordinary (much less default) setup. That means it will have to be either a "base" MSDOS partition or MSDOS "extended" partition. That puts it out of LVM's ken. Completely. LVM is entirely irrelevant to your issues, or it should be. Based on what you've said elsewhere, I suspect where you are getting stuck is in disk druid, the GUI tool that the install process uses to allow you to set up a partition map with both LVM and non-LVM partitions and not worry about what to do with which, meaning that it calls into the gparted and LVM tools for you. You aren't, by any chance, trying to change the size of /boot on a PPC Mac, are you? I'm not knowledgeable enough to do sophisticated partitioning -- that would be like a half-blind spastic (both of which I resemble at times) trying to shave with a straight razor. He might succeed, of course. Partitioning really isn't rocket science. I suppose that it may sound like I'm insulting you to say that, but I'm not. Trust me. The scariest part of partitioning is trying to guess how much you need where, and that was one of the original reasons for the existence of the LVM project. The other scariest part is that the old tools allowed you to declare the partitions to begin and end at certain places when setting up the labels, and then allowed you to tell the system they started and ended at other places, which is definitely, well, not rocket science, but a bit scary. If your eyes just glazed over, don't worry. It would be pretty hard to get any of the GUI tools to allow you to do that. You'd have to try really, really hard. So you really don't need to know about that. Alternatively, you may create LVM volumes and partitions inside them. It's all there in the GUI, and it's completely configurable. Nothing is forced down on you, AFAIK. I haven't the faintest conception what LVM is, much less what good it is to the Alpha Plus Technoids who understand it, but whose prowess I no more aspire to than they to expertise on the history of tongues. I did try, several times each, not only with Anaconda but by accepting the risk of using gparted and qtparted. All refused, every time, to let me add a single byte to /boot. Did you try deleting the partition after /boot first? That's the usual step. It may not be necessary when performing a fresh install, but once the partitions have been cut (labeled, really), you need special tools to move partitions, and if you resize one partition, the partitions after that one must be moved. (Unless you're using LVM.) That means that, if you have any important data (configuration files, etc.) in the partition after /boot, you need to back that up first. Okay, that's the other scariest part. If you have data in a partition that you can't afford to lose, you really, really should back it up first. Okay, okay, physically moving partitions is scary. Theoretically, it's just moving bits around, but calculating from where to where, well, yes, that can get close to rocket science in terms of being "hard". And, no, we do not have an equivalent to "Partition Magik" or whatever that was. (Been seriously bit by PM. I mean, to the point of copying an important project out of a botched MSWindows system on floppies. And even with access to Microsoft's documentation, Norton has a little trouble with that stuff.) There are gnu tools for moving the data when resizing partitions, but those are also easy to screw up with. The only thing I miss is the ability to use old-school fdisk instead of disk druid, but over time I learned to trust it to do its job as well as fdisk. :-) Fdisk is another of the things of which I know only how to spell them; life is too short The only thing you have to fear is fear itself. (Just kidding.) Hopefully, what I've written above will provide enough clues that you can figure out how to tell us what you're really trying to do, and then it will be easier to give you advice, suggest alternatives, etc. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: PythonCard?
On Jan 2, 2010, at 8:00 AM, Roger wrote: On 01/01/2010 07:17 PM, Joel Rees wrote: Anyone using it? Anyone know why attempting to install it would load the library parts of the package but fail to put the appropriate files in /usr/share? Do you mean PythonCad, if so yes but its not my app of choice. Roger No, I mean PythonCard, as in the Python GUI gadget inspired by HyperCard. I'm looking at it as a potentially amusing way to play with some GUI stuff in Python. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
PythonCard?
Anyone using it? Anyone know why attempting to install it would load the library parts of the package but fail to put the appropriate files in /usr/share? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
reboot cycle, no log messages
Just before it reboots, it gives me messages about old ext3 partitions that have problems mounting. History: this machine has seen upgrades and re-installs from FC2 or so. Most recently, I had succeeded in doing an upgrade from F7 to F9 using a netinstall CD and some patient cleaning up, then an uneventful upgrade to F10. Then I seriously botched a CD-less upgrade via the command-line upgrade tool because my /var/log partition only had 2G and I needed more than that for the packages. (Used lvm to extend /var/log to 5 G and then it did something really strange that I don't quite recall and left me with a system partially upgraded, trying to find th wrong kernel. And I didn't want to mess with that, so I cleared an old 40G partition I wasn't really using and wiped the botched boot partition, leaving the old stuff behind in the lvm-managed partition it had been in. Installed F12, and it booted fine. Then I added some old partitions I'd been using for backup to /etc/fstab, and added the old lvm partition, as well, for laughs. At that point, it would start boot cycling, repeating 2 to 7 or more times before something settled down and it would boot. This is what I captured with a digital camera when it cycled: /dev/mapper/fc7-7[various]: clean [long list of partitions from the old system] /dev/mapper/fc7-7varwww: clean, 644/516896 files, 27268/516896 blocks [failed] *** An error occured during the file system check, *** Dropping you to a shell; the system will reboot *** when you leave the shell. *** Warning -- SELinux is active *** Disabling security enforcement for system recovery. *** Run 'setenforce 1' to reenable. sulogin: error while loading shared libraries: libfreebl3.so: cannot open shared object file: No such file or directory Unmounting file system Automatic reboot in progress. --- After it successfully booted, I would be able to mount and access the old /var/www partition with no problem. I find this message in the forum: http://forums.fedoraforum.org/showthread.php?t=230419 but my /var/log/messages contains no EXT4-fs error messages. Setting the passno entry in /etc/fstab to 0 or commenting out the entry just pushes the error message back. When a large partition at / dev/sdb4, which I use for backup, is the final partition checked in / etc/fstab, it gives me -- fsck.ext3: Attempt to read block from filesystem resulted in short read while trying to open /dev/sdb4 Could this be a zero-length partition? -- and then the messages about libfreebl3.so and the automatic reboot. yum provides libfreebl3.so tells me that nss-softokn-freebl is installed. Trying an erase to re-install it tells me that most of the OS and most of the apps seem to be dependent on it, one way or another. Hmm. This looks like I need to file a bug or two. Anyone know what's happening here? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: How do you clear a botched kernel on a PPC system?
On Sep 27, 2009, at 3:20 AM, a helpful person wrote me off-list: (Leaving out the name in case the off-list was intentional. Or maybe it was because it was a reply to my mis-post to the -test-list.) On Sat, Sep 26, 2009 at 07:19:45PM +0900, Joel Rees wrote: it tells me that I should try passing in the init= parameter. I tried several permutations of what I thought was the probable syntax, boot: hd:3,/vmlinuz-2.6.29.5-191.fc11.ppc init=/ initrd-2.6.29.5-191.fc11.ppc.img yh Err..., init is a program you are running first. If everything else failed try 'init=/bin/bash' Okay, [...] init=/bin/bash ... VFS: mounted root [and something I missed as it rebooted.] Freeing unused kernel memory: 332k init Warning: unable to open an initial console. Failed to execute /bin/bash. Attempting defaults. . . Kernel panic. - not syncing: No init found. Try passing init= option to kernel. Rebooting in 180 seconds. Don't you have some earlier kernel which still boots? If that fails too then maybe indeed your /sbin/init is messed up but that would have nothing to do with kernels. The previous kernel did work, as I mentioned in another post, but only when I went in, renamed it (and the three other files) and so forth, as I described in the other post. I've yum remove-d kernel-2.6.30.5-43.fc11.ppc, and so it boots the old kernel and seems to run okay. So if the init= parameter is not supposed to be the initrd file, the bug in yaboot (under this ibook's openfirmware) might be that it's trying to pass an init= parameter when none is specified on the boot line. Hmm. I got a message back from the developers, from a comment I left where someone else reported this bug. Getting a fixed kernel is going to take a little time, they say. In the meantime, I'm avoiding updating, to keep the kernel at the previous level, and hoping the kernel/userland level mismatch will not bomb me out of anything important. Thanks. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: How do you clear a botched kernel on a PPC system?
On 27 Sep 2009 10:05:04 +1000, David Timms wrote, On 09/26/2009 08:27 PM, Joel Rees wrote: I find myself in another nice catch 22. ... So I'm having a hard time with things today. Can someone put me out of my misery? Yeah, mate, take that .22 above and... Thanks for the sympathy. ;-/ No seriously: Without having a ppc machine, I imagine that you can hit escape or any key during boot, so that you get the boot loader's option menu, and move the cursor to the previous kernel entry, hit enter... Yeah, you'd think. Open Firmware doesn't seem to be as tame as BIOS, however. It should allow typing the kernel name in at the boot prompt, but I couldn't get that to go, either. It would take the name of the down-level kernel, but it couldn't find the init. I tried specifying that with "init=", but it wouldn't take the path any way I could think of, without or without the drive spec, etc. As long as the older working kernel is still installed, it should be enough to yum remove kernel.specific.bad.version, and the rpm should take care of setting the previous versions item as the one to boot. Let me tell you what I eventually did: Inserted the netinstall CD and hit the power switch, held down the C key as one does on Macs. At the Fedora installer's boot prompt, boot: linux rescue After setting the language and keyboard and letting it start the network, it asks whether to try to mount the file system read only under /mnt/sysimage, and I took that option. Once it gave me a shell, I used mount with no parameters to get a look at what was where, then started umount-ing the volumes in order and then mounting them, so that they were mounted read-write. But I wasn't sure what to do with some of the more esoteric volumes. But it still wouldn't let me run yum under chroot /mnt/sysimage . Couldn't find the name server for some reason, and /etc was under /, which was still ro, and I really didn't want to umount everything again so I could mount / rw and edit resolv.conf to see if that would point the box back to the name server. (Got to put /etc in it's own partition next time, I guess. Or, if I had had my crystal ball out, I'd have set the network by hand instead of letting DHCP handle it.) So I gave that up and just mv-ed the offending kernel out of the way and ln-ed the previous kernel, config, initrd, and System.map to the old kernel's name. It boots up, and yum runs, but I'm worried about the naming games. Unfortunately, when I yum info the kernel, it looks like they haven't fixed the kernel yet after all, so I guess I'll yum remove the bad kernel. After filing a bug on it. Then I'd better practice using yaboot and, what was it? ybin? (And bash, too. Double checking those long version strings in the middle of file names is not something you want to do under pressure or late at night.) Thanks for the comments. I'm wondering if I'm the only PPC Fedora user left on the planet. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
How do you clear a botched kernel on a PPC system?
I find myself in another nice catch 22. Some time last week, I updated with yum (after several weeks of not updating) and, for some reason, I was able to reboot that day, but the next day attempting to boot Fedora resulted in a kernel panic. Looking things up, I see I'm not the only one with a panicking kernel. So, I got to work to clear the crudded kernel out of the way. Took me a while of looking around the web to rediscover the incantation to get to rescue mode with an install CD: "linux rescue" at the second stage boot prompt. Then I booted to rescue mode. I was thinking along the lines of, even though I don't get a grub menu like on x86, I should be able to change the configuration file to default to the previous kernel. Is the usual approach to that to use ybin? (And it looks like, with all the strange parameters, and the success rate I'm having today with arcane command parameters, it would likely take several tries to get it right.) Isn't there an easier way to do this? I tried passing the kernel name of the previous kernel in at the second stage prompt, but, even though I figured out (again) the arcane open firmware syntax to point to the kernel: hd:,/vmlinuz-.ppc (Since the boot volume is separate, the path to the kernel is empty.) it tells me that I should try passing in the init= parameter. I tried several permutations of what I thought was the probable syntax, boot: hd:3,/vmlinuz-2.6.29.5-191.fc11.ppc init=/ initrd-2.6.29.5-191.fc11.ppc.img and such, but all of them failed to execute, suggested specifying init= and left me at 180 seconds to reboot. So I'm having a hard time with things today. Can someone put me out of my misery? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: name server via dhcp, but don't want dhcp assigned addresses
On Mon, 21 Sep 2009 17:37:12, Sam Varshavchik replied, Joel Rees writes: The WAN side of the router runs dhcp to my ISP, and gets the dns server addresses by dhcp, as well. Check your router's documentation. The way that 99% of these routers are set up, is that they run a caching nameserver internally, and on the local LAN they give their own IP address as the DNS server's address, via DHCP. Well, yeah, it does that. That is, I think the one page of docs said that it did, and I think I remember testing it when I first got it six or seven years ago. (Sure didn't expect to be using it this long.) Small cache, but shouldn't be so small that I would notice delays or anything, even on a big YUM update. It's a black box, if it's using open source, and if NEC has published the source, they sure haven't made it easy to find it. Probably closed source. I seem to be able to telnet in, but it doesn't recognize any command I give it except "quit". (or was it "goodbye?") I don't really trust it, if I could afford the money and time to replace it with something I could load openBSD on, I would. (Come to think of it, it's rental, I should be able to justify the cost of replacement by how much it has cost to rent it all this time.) I guess, if I trust it to route, and if I can't shut the DNS function off, I might as well trust the DNS function as well. If somebody gets far enough into it to do a MIM on the DNS function, they can probably MIM the routes as easily. In the past, the ISP had told us to set the primary and secondary dns server addresses statically, so I had the router set to serve dhcp with those address. But I have also set the dns primary and secondary server addresses for all the boxes by hand to the dns servers Chances are that this is unnecessary. You should've just set your servers to use your router as the DNS server. It was the ISP's original recommendation. So, my problem is that I need to tell each Fedora box to accept the DNS server addresses provided by the DHCP server (the router, actually, which worries me), but not ask for a host IP address for itself, but the GUI dialogs in current Fedora don't provide that as an option. Why don't you test setting your server as full blown DHCP client, and see what DNS address your router gives you for your DNS server. Chances are that it's your router's IP address. In which case you just need to configure your servers to use a static DNS server on your router's IP address. The ISP recommends leaving the DNS addresses to be set via DHCP, rather than setting the router as the DNS server. Not that recommendations for the average customer are the only way to do things, of course. Well, since I seem to be able to set the Macs on the network to keep a static host IP address and use the DNS server addresses passed along by the DHCP server, I was hoping I could do that with the graphical UI stuff on Fedora. Or even with /etc/dhclient.conf. (Not really seeing how yet from the man pages, so now I'm wondering if that's actually part of the standard.) OK. Thanks for pushing me to think a little further about the implications of trusting the router. (And about whether I should consider investing in a router I can control, as an investment against the cost of more rent.) Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
name server via dhcp, but don't want dhcp assigned addresses
A quick search of the archives didn't turn up anything I could figure out easily. Background: I have ADSL, with a "modem"/router that does filtering, dhcp, etc. Since I want to refer to the boxes on the internal LAN (natted to local-address subnet) by name, I have the router set to only automatically allocate a piece of the subnet. Usually, all of the internal machines have their IP addresses set statically, thus, to addresses not included in the automatic portion of the subnet. The WAN side of the router runs dhcp to my ISP, and gets the dns server addresses by dhcp, as well. In the past, the ISP had told us to set the primary and secondary dns server addresses statically, so I had the router set to serve dhcp with those address. But I have also set the dns primary and secondary server addresses for all the boxes by hand to the dns servers provided by my ISP, so that I was ignoring the dns servers provided by the router. Recently, my ISP (a middle-rank provider in Japan) sent around notices that they were changing their setup, such that the dns servers would be advertised by dhcp and subject to change. That is, they said, more or less, "Clear the addresses you've set as your DNS servers in the DCHP client dialogs" (with example screen shots of Macs and MSWindows DHCP client dialogs) "so that you get the DNS addresses provided by the DHCP servers." On the Macs, there is an option to set the box as a DHCP client with static IP address, so I can still keep that much of my lan intact. But the Fedora boxes do not provide a static IP address client as a setting in the standard dialogs, as far as I can tell. So, my problem is that I need to tell each Fedora box to accept the DNS server addresses provided by the DHCP server (the router, actually, which worries me), but not ask for a host IP address for itself, but the GUI dialogs in current Fedora don't provide that as an option. Am I just missing something there, or will I need to disable the GUI and dig into text configuration files in /etc/dhclient.d? (nis.sh and ntp.sh don't seem to be the files I need.) Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
F11: Fedora's Netbeans, Eclipse, and Tomcat.
(This is going to break threading again, I apologize.) on Fri, 18 Sep 2009 18:13:55 -0700, Daniel B. Thurman" com> wrote From what I can tell, there is no possible way to run Fedora's implementation of Netbeans and Eclipse against Fedora's implementation of Tomcat 5 & 6. > [...] Against? Why against? I usually simply let each one load its own class tree where it wants to. Lots of duplication, but it saves a lot of time. Not pretty, but it generally gets things working. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: amaya in repos?
Okay, my report on trying to compile Amaya for ppc. On Jul 19, 2009, at 7:35 PM, Joel Rees wrote, Ed Greshko responded, Joel Rees wrote: > Amaya shows to be orphaned, last entry is Fedora 9: > > https://admin.fedoraproject.org/pkgdb/packages/name/amaya? _csrf_token=1e50eca0476c3f20d764e66baf5cfd9654b11dc6 > > > Neither yum info nor yum search seem to find it. At least, not on a > ppc machine. > > Have orphaned projects been removed? > > Would the best approach at this point be to download the source from > w3.org and build? Anybody use it? Care to comment? Especially, > concerning the ability to enter Japanese? > > Have you tried their rpm? http://www.w3.org/Amaya/User/BinDist.html Thanks, Ed. I have a habit of building from source things that aren't in the distro packages. And, now that I think of it, they don't have a binary for ppc. One of the source packages they provide comes with all the dependencies, and I don't recognize a couple of them. freetype and w3c-libwww were installed, redland was not, but was in the regular repos. Parts of Mesa seem to be installed, other parts seem to be n the repositories. I suppose the quickest way to find out if it's enough is to try building it. yum search wxWidgets gives me a list of stuff like bacula-console- wxwidgets, compat-wxGTK26, and, hmm. Maybe that would be wxBase and wxGTK and wxGTK-gl and wxGTK-media, which are already installed. I needed the -devel versions of gtk2 and Mesa-libGl (I think they were) in addition to what I already had, then the configure script ran to completion. Just to be helpful, I installed bison and the gnu fortran compiler. And I think there was one other -devel package that I had to intuit from the configure script's messages (using yum search). Anyway, it looks confusing, and I guess I was hoping there would be someone on the list here who has used it recently and could tell me the lay of the land before I wade in. Well, I guess I'll try building it after the family is all in bed. Or maybe set the build going now, before I start washing the dishes. Or maybe get the RPM, since it looks possible that I have all the dependencies, now that I've looked again. Took me a little longer to get at it than I thought, but I did try compiling, and ended up with a lot of compiler error messages about things not declared correctly and such. Then the X11 session ran wild and I had to use the virtual terminal to kill the entire X11 login (KDE) to get it back. So I didn't get a chance to grab the error messages this time. Okay, this project goes on a back burner for a while. Maybe I'll use a gnome session, since KDE on PPC feels a little fragile sometimes. Or, maybe I'll try building it on my AMD box first. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Linux "NULL pointer dereferece" in the News...
Woops. Didn't intend to mess threading up that much. (Apologies for messing up the threading yet one more time, and for cluttering up the fedora thread with more stuff that is only semi- relevant here.) -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
(no subject)
(Apologies for messing up the threading yet one more time, and for cluttering up the fedora thread with more stuff that is only semi- relevant here.) Michael Eager wrote, ... Compilers determine what modifications they can make to the code using the inferences they make based on the data flow through a program. They don't say "well, this is undefined, we can muck it up however we like". In the case of this *kernel* bug, the compiler determined that the pointer must be valid, because it had been previously dereferenced. Would it be entirely inappropriate to have a compiler switch that would cause the compiler to issue warnings when pointers are used without testing after being returned from functions calls? This allowed the compiler to eliminate a test which would always be false, *as long as the behavior of the previous code was defined*. Which is a problem, because the compiler authors are fully aware that dereferencing a pointer without testing is a common bug. This code is correct and well defined as long as the pointer is valid. Only if you can assume that the dereference after the function call and before any comparison must indicate that the programer knows the function will always return a valid pointer. Otherwise, the code, as it was, was not well defined. The behavior only becomes undefined when the pointer is null. The only way that the compiler could determine that the initial dereference of the pointer was undefined would be to insert a test for null before the dereference. That would be non-standard behavior, and potentially incorrect. And, if it could solve the problem, it would indicate that the dereference before testing could, and should have been flagged with a warning. It would have to do this for every piece of code which dereferenced a pointer which was passed into a function, dramatically impacting performance. Or perhaps it could issue a warning message saying that it couldn't determine that the pointer was valid, resulting in a warning that would occur thousands of times when compiling the kernel. If this kind of code really occurs that often in the kernel, we have serious problems. OK, what would reasonable, sane people do in that case? That's right, they'd fall back on the behavior of just doing what the program source code says, but no, gcc is too smart for that, gcc's undefined behavior shows how smart it is and therefore makes much more sense than doing the obvious :-). You can get exactly that behavior by not optimizing your code. Your code will run much slower, but that's OK, isn't it? Ah, you want optimizations? But you want them to magically decide that one is an error in the program and shouldn't be done, while in another place the same optimization should be done because it generates better code. OK, write a description of how to determine one case from the other and I'm sure every compiler developer will rush to implement it. Actually, it's fairly simple. The default warning option for discarding code should be to issue a warning about discarded code. Discarded code generally indicates that the programmer did not say what he meant, and that is usually a no-no in kernels. And, IMNSHO, in any other application, the programer should be required to insist that discarded code can safely be ignored, via pragma or switch. Yeah, this sort of debate would be more appropriate in the compiler developers' mailing list. I feel like letting a little steam off, and I don't feel like signing up for yet another mailing list. So sue me. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Linux "NULL pointer dereferece" in the News...
(Sorry about hashing the threading.) Tom Horsley commented, On Sun, 19 Jul 2009 11:49:42 -0700 Michael Eager wrote, > OK, write a description of how to determine one case from the > other and I'm sure every compiler developer will rush to implement it. The fact that they already have a -fno-delete-null-pointer-checks option in the compiler seems to demonstrate that they know they are doing this specifically because it is undefined and they can do whatever they want (including checking a compiler option to choose a different thing to do). I guess thinking that the best choice is always to do what the source code says makes me a member of a looney minority though :-). Well, the -fno-delete-null-pointer-checks option was intended for a related, but different class of problems, specifically certain application areas where it is defined to access NULL pointers. (The standard does not define the behavior, but it isn't really against the standard for a specific application field to define the behavior. The code becomes non-standard C, however.) (Specifically, embedded applications, although I tend to wonder if the hardware design could not usually get around the reasons for putting valid stuff in the bottom of memory. Actually, I'm beginning to wonder, what with modern logical address spaces usually significantly larger than applications need, whether it doesn't make sense to devote address bits to the problems of classifying pointers.) The C standard does not specify that every function returning a pointer return invalid pointers. Moreover, it does not specify how to determine a valid pointer, even though we customarily assume the NULL pointer idiom. I don't disagree that it feels awkward that they know the pointer has been referenced and therefore would seem to be valid, thus, non-NULL, but aren't able to back up the parse stack and see that the access looks like a test-after-use bug. Also, we humans walk back up parse stacks with relative ease, but it slows compilers down to be walking back up the parse stack with assumed valid code. I'm talking, worst case could be slow enough to be completely unusable. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: Linux "NULL pointer dereferece" in the News...
Rahul Sundaram wrote, On 07/19/2009 11:31 PM, Tom Horsley wrote, > On Sun, 19 Jul 2009 23:11:52 +0530 > Rahul Sundaram wrote, > >> It is not so simple. This is not a compiler bug. I suggest you read >> through http://lwn.net/Articles/341773/rss to understand why. > > I did. It is a compiler bug no matter what a bunch of language lawyer > holier than thou compiler developers say :-). The kernel developers claim it is a kernel bug and the compiler developers claim it is not their bug and this everybody agrees with and yet you don't agree with all of them? Who is being holier than thou here? Kernel first dereferences a pointer, and after that checks whether it's NULL. It is quite common as a compiler optimization to compile out code like this. Rahul I vaguely remember some of the reasons (mainly the difficulty of untangling the semantics of pointer tests), but, I must admit, after using Java long enough, it seems odd to me that the C compiler would be smart enough to catch the theoretical "don't care" and not smart enough to distinguish between the reasons for the don't care. I think I remember, way back when, using C compilers that would warn about code that isn't reached and is discarded, and even warn about code that dereferences a pointer before testing it after the pointer is returned by a function call. So it definitely feels like both a coding error and a (design class?) bug in the way the compiler options interact. It would be nice if certain combinations of options would at least issue "potential evil compiler switch combination" warnings before a compile started. Maybe it does already, and the warnings are being swallowed in the make process? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: amaya in repos?
Ed Greshko responded, Joel Rees wrote: > Amaya shows to be orphaned, last entry is Fedora 9: > > https://admin.fedoraproject.org/pkgdb/packages/name/amaya? _csrf_token=1e50eca0476c3f20d764e66baf5cfd9654b11dc6 > > > Neither yum info nor yum search seem to find it. At least, not on a > ppc machine. > > Have orphaned projects been removed? > > Would the best approach at this point be to download the source from > w3.org and build? Anybody use it? Care to comment? Especially, > concerning the ability to enter Japanese? > > Have you tried their rpm? http://www.w3.org/Amaya/User/BinDist.html Thanks, Ed. I have a habit of building from source things that aren't in the distro packages. One of the source packages they provide comes with all the dependencies, and I don't recognize a couple of them. freetype and w3c-libwww were installed, redland was not, but was in the regular repos. Parts of Mesa seem to be installed, other parts seem to be n the repositories. I suppose the quickest way to find out if it's enough is to try building it. yum search wxWidgets gives me a list of stuff like bacula-console- wxwidgets, compat-wxGTK26, and, hmm. Maybe that would be wxBase and wxGTK and wxGTK-gl and wxGTK-media, which are already installed. Anyway, it looks confusing, and I guess I was hoping there would be someone on the list here who has used it recently and could tell me the lay of the land before I wade in. Well, I guess I'll try building it after the family is all in bed. Or maybe set the build going now, before I start washing the dishes. Or maybe get the RPM, since it looks possible that I have all the dependencies, now that I've looked again. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
amaya in repos?
Amaya shows to be orphaned, last entry is Fedora 9: https://admin.fedoraproject.org/pkgdb/packages/name/amaya? _csrf_token=1e50eca0476c3f20d764e66baf5cfd9654b11dc6 Neither yum info nor yum search seem to find it. At least, not on a ppc machine. Have orphaned projects been removed? Would the best approach at this point be to download the source from w3.org and build? Anybody use it? Care to comment? Especially, concerning the ability to enter Japanese? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: ssh clarification needed
On Jan 5, 2009, at 12:01 AM, Todd Zullinger wrote: Kevin Kofler wrote: * authentication keys - those are what you use to log in instead of a password. They're one per user and machine unless you explicitly copy the private key to a different machine or user account (something you normally shouldn't do I presume you mean only the latter part (copying the private key to another user account) is something that you shouldn't do? I share the same ssh private key between my desktop server and my laptop (both as the same user). I don't see much reason to have two separate keys for that. Personally, I might use the same passphrase, but I think I would re- generate the keys on each device. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: ssh clarification needed
empirical. Well, ... On Jan 5, 2009, at 1:44 AM, Robert L Cochran wrote: Mail Lists wrote: On 01/04/2009 11:24 AM, Mike Cloaked wrote: By the way if your root partition was not encrypted then someone with physical access to your machine could boot into single user mode and get root access - hence encrypting the root partition is probably the only way to avoid that - unless someone knows a different way in? Just boot a CD, DVD or USB key I own the whole laptop - as a bad guy i would prefer to boot my own OS anyway not yours. I disagree with this. I know from experimenting that if I boot another Linux OS (regardless of media used) and then try to access the data on separate a LUKS encrypted device, I can't see that data without providing the passphrase. Be careful to distinguish between "can't see the unencrypted data" and "can't see the encrypted data". As a matter of fact you are prompted to supply the passphrase. Remember, that's with the standard tools, of course. If you boot a Microsoft Windows OS you can't see the data anyhow...Microsoft Windows doesn't recognize non-Microsoft filesystems such as ext3. Again, that's with the standard tools. How much you have to worry about things like that depends a lot on what you have on your storage device, of course. Well, I should say, it depends on what your opponent thinks you have on the device. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: ps2 mouse pointer wanders away
Tying up some loose threads as I have time -- On Aug 16, 2008, at 8:15 AM, Joel Rees wrote: Ed Greshko wrote: Joel Rees wrote: I find that I often lose control of the mouse pointer. It seems to happen more when I'm seeding torrents, or running a diff or a message digest check on the download, or doing other things that load the system. And it just got worse, both in Fedora 7 and Fedora 9. Sempron 2600, single processor, VIA KM400something+8237 chipset, 760M RAM. Generic PS-2 mouse and keyboard. I can plug in a USB mouse and use that, and sometimes plugging the USB mouse in brings the PS2 mouse back, as well. In your /etc/X11/xorg.conf file what is mouse protocol set to? No mouse definition in that file whatsoever. [...] I've seen what I think you are describing when set to infrequently seen/used setting. FWIW, the one I use most often is... Option "Protocol" "IMPS/2" Thanks for reminding me where that file is. I found an example of the mouse definition section on the web, but the problem now is getting the box back on the web first. Oh, yeah, checking the box and letting the gui network manager or whatever it was manage the network was something I mentioned finding in another thread, and probably my fat fingers while I was dozing at the keyboard were the culprit. The out-of-control mouse was just a bad mouse. A little high-tech wire-wiggling with a multimeter, then 5 cm. clipped off the end of the wire, re-soldered, and it works fine. Flaking out with Fedora 9 appears to have been just a coincidence, or maybe some timing thing in nine made it more sensitive there first. More noise, my apologies. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
update notification dialog conflicts with authenticate dialog
What I was doing: I was checking to see if a couple of volumes had mounted properly. Got the "Enter the root password." authenticate dialog. (Why isn't that a sudo dialog like the Mac OS presents, so you don't have to give out the root password to everybody?) Then the updater dialog pops up. Freeze. The updater dialog is in front, covering all buttons on the authenticate dialog. I'm not sure it would matter that the buttons are covered, because the close box does not respond. And no button on the updater dialog responds to mouse clicks. In fact, nothing on the screen responds. I switched to a character console, logged in as root, and killed a notification-something process, and returned to the X console to find it responsive again, with the updater notification dialog gone, of course. Anybody seen this? is it already in Bugzilla? And can anyone tell me how to get the updater to refrain from running when ordinary users log in? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: FireFox 3 EULA
On Sep 16, 2008, at 11:55 PM, Steve Hill wrote: On Tue, 16 Sep 2008, Joel Rees wrote: If you do not agree to the GPL, you have no license at all, and the only thing that allows you to use the software in any way is fair use. Fair use does not necessarily cover running the software. Certainly not the case the world over. ISTR that UK copyright law has an exemption that basically says it won't prevent you doing whatever you need to do to use the product (e.g. installing the software on a hard drive, copying a database into RAM, etc). That would be, I'm assuming, that you have bought the product and have some sort of license thereby to use it? Remember, in the case of GPL-ed software, we are not trafficking in software in the usual sense. joelrees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: FireFox 3 EULA
I really don't have time for this tonight. On Sep 15, 2008, at 9:01 PM, Steve Hill wrote: On Mon, 15 Sep 2008, Joel Rees wrote: [...] Some apps show you a EULA for the GPL when they install. And if you can7t agree to the GPL, you can click disagree and refrain from installing. The GPL is not an EULA - the end user is not required to agree to it in order to use the software. In fact, requiring the user to agree to any licence (GPL or otherwise) before they can use a piece GPLed software is itself a breach of the GPL. If you do not agree to the GPL, you have no license at all, and the only thing that allows you to use the software in any way is fair use. Fair use does not necessarily cover running the software. There is a reason the README is included in every distribution, and there is a reason the GPL license notice is shown wherever copyright notices are shown. I thought I had more to say, but I'm going to work on the less-is- more assumption tonight and go to bed. [...] Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: lose network after re-boot
On Sep 14, 2008, at 6:20 PM, Timothy Murphy wrote: Joel Rees wrote: I just tried the Save item in the File menu of the Network Settings tool. Rebooted and the interface is still down. But this time, when I try to bring up the Network Settings tool, I get a message dialog, about trying to bring the interface up, which goes away after a second or two, to be replaced with an error dialog about being unable to bring the interface up. As a matter of interest, is the Network Configuration=>Network Settings tool intended for use with Network Manager? Funny you should ask. There's a little checkbox in there that asks whether the nic should be managed by the network manager. For some reason, after install, it was unchecked. So, I just checked it and rebooted and the settings took. Thanks. Now I can see if I can navigate the procedure for updating the signatures. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
lose network after re-boot
I lose my network after every re-boot. I have the same thing happening on my AMD desktop box and my iBook G4. I can get it back up with the nice GUI network tool, just by hitting the enable button. Then it works. Then I reboot and it doesn't work, until I click that button in the GUI network tool again. I just tried the Save item in the File menu of the Network Settings tool. Rebooted and the interface is still down. But this time, when I try to bring up the Network Settings tool, I get a message dialog, about trying to bring the interface up, which goes away after a second or two, to be replaced with an error dialog about being unable to bring the interface up. (I went and logged in in Japanese again, so I'm translating.) So, I select eth0, hit the green button, and ping a box in the local network, and mail ships me my dmesg like I asked. Shouldn't be hardware issues -- I'm dual-booting Mac OS X with no network problems, and I netinstalled the box via the built-in nic. I'm assuming that the results of using the GUI interfaces to the network is supposed to be sticky. Is my assumption there wrong? I seem to be too clueless to know where to look. Anyone got a clue for me? Here's my /etc/sysconfig/network-scripts/ifcfg-eth0 (after bringing up the interface, with internal details obfuscated/effaced): --- # Apple Computer Inc. UniNorth 2 GMAC (Sun GEM) DEVICE=eth0 BOOTPROTO=none BROADCAST=172.19.255.255 HWADDR=**:**:**:**:**:** IPADDR=172.19.36.26 NETMASK=255.255.0.0 NETWORK=172.19.0.0 ONBOOT=yes DNS1=**.**.**.** DNS2=**.**.**.** SEARCH="reiisi.homedns.org" NM_CONTROLLED=no GATEWAY=172.19.138.95 TYPE=Ethernet USERCTL=no PEERDNS=yes IPV6INIT=no my /etc/hosts: # Do not remove the following line, or various programs # that require network functionality will fail. 127.0.0.1 moth.reiisi.homedns.org moth ::1 localhost6.localdomain6 localhost6 172.19.36.24reiisi.homedns.org reiisi 172.19.36.25federales.reiisi.homedns.org federales and my dmesg: Using PowerMac machine description Total memory = 768MB; using 2048kB for hash table (at cfe0) Initializing cgroup subsys cpu Linux version 2.6.25-14.fc9.ppc (mockbuild@) (gcc version 4.3.0 20080428 (Red Hat 4.3.0-8) (GCC) ) #1 Thu May 1 05:51:33 EDT 2008 Found initrd at 0xc1a0:0xc1da7800 Found UniNorth memory controller & host bridge @ 0xf800 revision: 0xd2 Mapped at 0xfdfc Found a Intrepid mac-io controller, rev: 0, mapped at 0xfdf4 Processor NAP mode on idle enabled. PowerMac motherboard: iBook G4 via-pmu: Server Mode is disabled PMU driver v2 initialized for Core99, firmware: 0c console [udbg0] enabled Entering add_active_range(0, 0, 196608) 0 entries of 256 used Found UniNorth PCI host bridge at 0xf000. Firmware bus number: 0->0 PCI host bridge /[EMAIL PROTECTED] ranges: MEM 0xf100..0xf1ff -> 0xf100 IO 0xf000..0xf07f -> 0x MEM 0x9000..0x9fff -> 0x9000 Found UniNorth PCI host bridge at 0xf200. Firmware bus number: 0->0 PCI host bridge /[EMAIL PROTECTED] (primary) ranges: MEM 0xf300..0xf3ff -> 0xf300 IO 0xf200..0xf27f -> 0x MEM 0x8000..0x8fff -> 0x8000 Found UniNorth PCI host bridge at 0xf400. Firmware bus number: 0->0 PCI host bridge /[EMAIL PROTECTED] ranges: MEM 0xf500..0xf5ff -> 0xf500 IO 0xf400..0xf47f -> 0x nvram: Checking bank 0... nvram: gen0=292, gen1=291 nvram: Active bank is: 0 nvram: OF partition at 0x410 nvram: XP partition at 0x1020 nvram: NR partition at 0x1120 Top of RAM: 0x3000, Total RAM: 0x3000 Memory hole size: 0MB Zone PFN ranges: DMA 0 -> 196608 Normal 196608 -> 196608 HighMem196608 -> 196608 Movable zone start PFN for each node early_node_map[1] active PFN ranges 0:0 -> 196608 On node 0 totalpages: 196608 DMA zone: 1536 pages used for memmap DMA zone: 0 pages reserved DMA zone: 195072 pages, LIFO batch:31 Normal zone: 0 pages used for memmap HighMem zone: 0 pages used for memmap Movable zone: 0 pages used for memmap Built 1 zonelists in Zone order, mobility grouping on. Total pages: 195072 Kernel command line: ro rhgb quiet root=UUID= mpic: Setting up MPIC " MPIC 1 " version 1.2 at 8004, max 4 CPUs mpic: ISU size: 64, shift: 6, mask: 3f mpic: Initializing for 64 sources PID hash table entries: 4096 (order: 12, 16384 bytes) GMT Delta read from XPRAM: 0 minutes, DST: off time_init: decrementer frequency = 18.432000 MHz time_init: processor frequency = 1199.97 MHz
Re: Secrecy and user trust
I scanned this thread and was a bit disappointed. Lots of apparently competent people arguing from false assumptions. None of you know me from Adam. If we met at a key-signing party, you would have no reason to trust this guy with an odd scare and a week's growth of whiskers on his face. Face to face is only useful if the face matches a familiar pattern. Shoot, I know most of you better from the patterns in your posts than I would know you if we met face to face. You could show me your passport and other ID (including the electronic ones), and I would still know more about you from the patterns of your posts on the list. In the real world, if Jack Smack comes up to me and tells me my cousin is introducing him, I check with my cousin, and I still don't really trust him until I get to know him. Dorothy Developer who is an acquaintance of a friend of a developer known by a couple of guys I met at the last Linux conference I attended has an advantage in numbers, but I still don't trust her programming until I have run her stuff a couple of times. And checked my logs Chain of trust is an illusion. So is a web of trust. What I do have is a running OS and log files in my firewalls. And memories of domain names I have surfed by. That the keys posted on the mirrors and the main site match is good evidence of a number of people cooperating. That is why I trust the keys enough to try the OS, and I could trust them no more were they to be signed by someone chained back to verisign or whoever. Speaking of verisign, trusting a single entity as a "root of trust" is not far removed from trusting some charismatic religious or philosophical type to be your moral root. The CAs can serve a purpose, but they tend to be used against that purpose more than for. The best use of the keys is as a checksum. Isn't that clear? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: [FWW] method for minimum install
On Sep 13, 2008, at 5:59 PM, Frank Murphy wrote: On Sat, 2008-09-13 at 17:30 +0900, Joel Rees wrote: I suppose, when I calm down, I'll now have to go start playing with rawhide and join the developers list and learn how to write code to fix this junk, but, haven't they had a Carly Simon moment on this -- a "What on earth was I think of?" moment? Joel Rees Not being a programmer, the reason I think is, the inter-dependence of various apps\libraries. Until such time I suppose until the "Linux Standard Base" fully takes off this will be the case. Frank When I loaded F9 on my AMD box (from the live CD), I deselected mono. It ended up with Tomboy, which I had also deselected, (but no f- spot). Checking dependencies, the only thing that removing mono caused to be removed was Tomboy. Thus, I do not believe such an argument is valid, and, considering the provenance, I am concerned that the argument could gain some default respectability. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
[FWW] method for minimum install
Not sure if I'm asking a question or trying to start a flame war here, but, about four hours ago, I was getting a net install (PPC) of Fedora 9 going and needed to go to the store for the wife, so I thought I'd be sensible for a change and not install the kitchen sink, and wait until the base system was running before I started messing with the package manager. So I de-selected desktop, developer's, and server general sets from the quick selection screen in the install, left it to customize after the install, hit the start button, and went to the store. When I got back, it was installing 889 packages. Is 889 packages a minimal install? Is de-selecting all the general sets not a minimal install? Why? Where is the minimal install button? Why? Can I file a bug on this as a vote for a real minimal install option without having to go into the screens for selecting and de-selecting individual package? And, really, seriously, why on earth is that IP-trap tomboy in any of the default installs? I was shocked that it was in the i386 live CD, but in PPC, too? In what ought to be a minimal install? At least I don't see any evidence of f-spot or whatever that silly photo "manager" was. (Not sure why rhythmbox and the video player have to be in there, either, but at least they don't have mono.) No, really. Fedora is not a part of the Novell microcosm. Fedora is trying to be clean with licenses, right? Where is the logic in allowing a single pseudo-text editor of questionable general utility to drag in the taint of mono? (I suppose it would hurt less if I could remove it without having to have access to the repositories.) No, I really don't want to start a flame war. But I would like to know the logic of this. Why are Tomboy and mono considered more "default" or more important or whatever than, say, parted and the LVM volume manager? And why isn't there a true minimal install without going through and hand-de-selecting stuff? I suppose, when I calm down, I'll now have to go start playing with rawhide and join the developers list and learn how to write code to fix this junk, but, haven't they had a Carly Simon moment on this -- a "What on earth was I think of?" moment? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: multibooting ppc with Mac OS 9 & X on an ancient iBook
On Aug 26, 2008, at 12:12 PM, Craig White wrote: On Mon, 2008-08-25 at 15:24 +0900, Joel Rees wrote: (Giving a little more detail:) On Aug 23, 2008, at 12:07 PM, Craig White wrote: Actually, I pre-partitioned the yaboot partition and the Linux big partition with the Mac OS X utility, as well. And the Mac OS 9 utility. I had hoped that someone with experience with multi-boot on Macintosh would have piped in because I lack this experience myself but I got the impression that the Fedora installer created the yaboot partition itself but I may have just let it. When you let it do the automatic setup it does. But Anaconda still isn't able to use it on my iBook. looking at a Mac with Fedora 8 installed, the anaconda-ks.cfg reports... #clearpart --all --initlabel --drives=hda #part appleboot --fstype "Apple Bootstrap" --size=1 --ondisk=hda #part /boot --fstype ext3 --size=100 --ondisk=hda #part / --fstype ext3 --size=1 --grow --ondisk=hda #part swap --size=256 --grow --maxsize=512 --ondisk=hda So evidently, I must have allowed the druid to create the partitions but note that anaconda on a ppc permits an fstype of 'Apple Bootstrap' so yeah, if you are gonna install multi-boot, you gotta have 1 mb partition designated for it. this may be useful (apparently you can't use fdisk on an Apple system) # parted -l Model: Maxtor 6Y080L0 (ide) Disk /dev/hda: 82.0GB Sector size (logical/physical): 512B/512B Partition Table: mac Number Start End SizeFile system Name Flags 1 512B32.8kB 32.3kB Apple 2 32.8kB 1081kB 1049kB hfs untitled boot 3 1081kB 106MB 105MB ext3 untitled 4 106MB 641MB 535MB linux-swap swap swap 5 641MB 82.0GB 81.3GB ext3 untitled obviously the yaboot partition is #2 That's a 100G drive? I haven't done multi-boot Mac's but I have done a bunch of different Mac setups. I'm not really a fan of multi-boot on any hardware choice. Different requirements. I need to take time out to learn how to program for ODF and PDF, so I can build programs that will do some things I presently use AppleWorks documents to accomplish. Until then, I need to be able to do a lot of work in the Mac OS. Appleworks is a black hole...it's virtually impossible to get documents out of Appleworks unless you are willing to open each one individually, save it as the Microsoft equivalent which is a time hog. Well, actually, Apple's iWok^Hrk suite, now that I look things up, is supposed to import a lot of the old AppleWorks stuff okay now. Just not the vector stuff. As for being a black hole, yeah, the data format is archaic. OOo has a terrific set of tools for document creation...I think it puts Appleworks to shame. Different strokes for different folks. OOo feels clumsy to me. Feels too much like MSOffice. That may be in part because I've been burned by MSOffice so many times that even the slightest whiff of MS turns my stomach. But what I'm using that is indispensable to me right now is a little feature where you can bury a spreadsheet in a "draw" document, show small windows on the printable page, and use the sort function with random keys to randomize the words in the printed list. Kind of a slick way to make word bingo cards, among other things. And I'm doing a lot of less program-like stuff with the drawing software, too. That SVG editor (inkscape?) that was in the news recently may help a lot for when I need to do drawings. Thinking about this, I suppose I should be able to build something similar with hidden worksheets in the MSOffice-centric way of thinking. Maybe I'll try that with OOo sometime. (I do use it on other, more powerful hardware. And I have to admit, I should not be surprised if the AppleWorks version might turn out to be less intuitive to my fellow teachers who might want to use the teaching materials I'm creating than an MSExcel "workbook" or whatever that would be called. Also, I've thought, several times, that I might be able to get even better effects by programing a plugin for OOo, just haven't had the time and luck to figure out how to get a handle on that yet.) Anyway, if there is an openfirmware incantation that I can use instead of installing yaboot, I could try that the next time I have time to work on this. I don't think so (on ppc...this is all you get) - OFI is for the Intel based Macs I think you mean EFI, which is iNTEL's take on openfirmware? Hate to be noisy, but I sure appreciate any pointers I can get. you probably got the best I had to offer - I actually made a clean install of OS 9.1 that I then made into a dmg file with 'Disk Utility' so I could just dump a full OS 9 setup onto any computer runnin
Re: question on a kernal "uhhuh" message
Boot off the CD/DVD. At the "boot:" prompt, enter "memtest86" and off you go. Is there a "memtestppc", as well? Listing bootable images showed a "check". When I gave that a run, it went for a while, to something more than 70% and then wandered off into the ether -- black screen, as if it openfirmware had put it to sleep, but it wouldn't wake up. The Fedora 9 powerPC netinstall doesn't seem to have any alternative images at all. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
Re: non-disclosure of infrastructure problem a management issue?
I think most of us were more peeved about not getting a *clear* warning, promptly, and wanting to know whether it really was a safety issue (do not download) or just broken servers (downloads may fail). They didn't say hardware, they didn't say source code control or other distribution software, they didn't say specific packages or distros, they didn't run around screaming, "Chicken Little was right! The sky is falling. RUN FOR THE HILLS" So we should have assumed that there was some ambiguous state typical of a breach discovered in the early stages. From the information so far, that's what it was, and the post-mortem in such cases does take time. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: multibooting ppc with Mac OS 9 & X on an ancient iBook
(Giving a little more detail:) On Aug 23, 2008, at 12:07 PM, Craig White wrote: On Sat, 2008-08-23 at 09:58 +0900, Joel Rees wrote: I recently tried installing Fedora 9 on an iBook that I need to boot both Mac OS 9 and Mac OS X on. I also have some other partitioning constraints, so I ended up with 6 partitions before I started the Fedora 9 install. Specifically, a Mac OS 9 partition (I know this is not necessary, but it does make the whole lludge more robust.), two booting Mac OS X partititions (Not case sensitive or journaling formatted, 10.2 is too early for that.), a UFS partition to serve to the web (Required by the configuration of one of the Mac OS X systems.), a UFS swap partition (Not strictly required, but, again, it makes the system more robust in my case, if not in the general case.), a partition intended for sharing with Linux Actually, I pre-partitioned the yaboot partition and the Linux big partition with the Mac OS X utility, as well. And the Mac OS 9 utility. Well, my memory is a bit confused, since I tried this several times, several different ways. I don't remember for sure, but I don't think the 1M partions I created with either the Mac OS 9 utility or the Mac OS X utility were accepted by the installer as boot partitions for yaboot. Don't remember all the ways I tried to get the installer to use them, but I couldn't. Fedora 9 happily created lots more partitions, ... when I told it to. but then Mac OS 9 (and X, IIRC) refused to boot. Couldn't recognize the format of the disk. (Might have helped if I had got a 120 G HD instead of a 160 G HD?) I think this was due to having more partitions than Mac OS 9 could deal with. I think, in one case, Mac OS X did boot, in Also, on a separate iteration, when I tried to force the yaboot partition to 1MB, that also apparently made the partition map unacceptable to Mac OS 9. I had to wipe the disks with the Mac OS 9 formatter and start again. (Lost a day or so of my time, but no data.) Since I thought I had time, I tried an install of just Fedora, but the current partitioning software wouldn't create the partition for yaboot any smaller than 16MB, IIRC, and then it wouldn't install yaboot in anything bigger than 1MB. I am currently successfully multi-booting Mac OS 9 & X (Jaguar) and openBSD, but openBSD is not using yaboot. I give the four-finger salute on startup and type "boot hd:,ofwboot /bsd" at the open firmware prompt. That doesn't really bother me, even though I have to remember that the keyboard map is US and doesn't match the Japanese keyboard. :-/ openBSD seems to take a little more nursing than I currently have time for, and I am primarily interested in getting the Gimp and openoffice.org running. Well, probably some of the edutainment stuff, as well. In about two weeks, one of the partitions will be freed, so I should have three partitions to give Fedora 9, and I am thinking of trying again. But I won't have the day or so necessary to re-build the Mac OS 9/X sides of things if the Fedora partitioning software makes the map unreadable to Mac OS 9 again. So, I am wondering a couple of things: One, Is anyone is currently successfully multi-booting Mac OS 9, Mac OS X, and Fedora 9 on any system, especially one with a boot HD larger than 120G? Two, would it be possible to boot with openBSD's approach, invoking an openfirmware script on the Mac OS 9 boot disk? (I haven't been able, yet, to untangle the web of what happened when yaboot became usable.) I haven't done multi-boot Mac's but I have done a bunch of different Mac setups. I'm not really a fan of multi-boot on any hardware choice. Different requirements. I need to take time out to learn how to program for ODF and PDF, so I can build programs that will do some things I presently use AppleWorks documents to accomplish. Until then, I need to be able to do a lot of work in the Mac OS. Anyway...I'm not sure why you would want or if you can have separate boot partitions for Mac OS 9 and OS-X and wonder why you would want to do that because if you create an HFS partition for both OS-9 (Classic) and OS-X you would normally keep them on the same setup and use the control panel 'Startup Disk' to choose which would boot. It's slightly more robust to put Mac OS 9 on a separate partition. My kids like to play a game called Bugdom that only exists in a Mac OS 9 version, and I've seen that game do funny things to Mac OS 9, so I'm a little superstitious about it. I may not absolutely need to, but cutting the partitions again will require me to install from scratch again, and I won't have time for that for a while. Use the Disk Utilities option on Mac OS-X install to create the partition for OS-X/Classic and leave appropriate 'unpartitioned' space for BSD and/or Linux. Yeah, bei
security blankets (was Re: non-disclosure of infrastructure problem a management issue?)
then I have to assume you don't trust the Fedora Project. I did trust the Fedora project. Now I'm not so sure anymore. Then who are you going to trust? Uhm, no, I guess that's not the right question, it only reminds us that we want to stay with F/OSS. Let me suggest to anyone who is still hot under the collar about the current situation, two things: One, if you want to understand the appropriate level of paranoia, go spend a day working backwards through the openbsd archives. Try http://marc.info/?l=openbsd-misc That will be plenty interesting. Two, if you've been paying attention to the news from more than a month ago, you should at least know there are active DNS exploits in the wild. ACTIVE DNS EXPLOITS IN THE WILD They haven't been shouting because it shouldn't be necessary. Under the circumstances, we should be significantly more paranoid and more cautious than we usually should be. The original announcement should have been enough, even if it wasn't perfect. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: H.D. install problem -
On Aug 25, 2008, at 5:31 AM, Bob Goodwin wrote: Russell Miller wrote: You now have your partition list, which you can mount. --Russell Ok, I can't mount it on / because that's already used but I can mount it on /mnt and copy files at least. Thats a start, more than I've been able to do so far! mount /dev/mapper/VolGroup00-LogVol00 /mnt I'll need to figure how to get grub to boot it ... If you succeeded in mounting it, look in the old /boot (in other words, in /mnt/boot) If you don't recognize what you should be looking for after checking the manual entries, etc., ask again. I've done this several times, including several times when I was fighting the old problem where you didn't want LVM handling volumes on more than one drive in a system. I don't know if that old limitation still exists, you'll want to check. (I probably want to check, too, but not today.) The short answer to booting the old system is to remove the new drive. The long answer includes adding to the new grub configuration, new entries to either chain or directly boot the old system, but modifying the drive numbering appropriately. If you chain, I think you'll end up editing the old grub's configuration as well, because of the drive numbering problem. Wish I had time to write a tutorial on it, but such do exist on the web. Some will be old and send you in the wrong direction, so look for something recent. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: non-disclosure of infrastructure problem a management issue?
I don't mean to be rude, but, ... [...] One thing this incident has taught us is to take regular backups of that mirror so that we can roll back to a non-suspect version of the Fedora updates. Didn't have that before, really missed it the last couple of weeks. Consider that a lesson well learned. And, while it may not have been the most convenient time to learn it, things could have been much worse. It's one of the costs (and, actually, one of the benefits) of working with open source. With "Proprietary" you have "guarantees". When they fall down on the job, or when other bad stuff happens, you can theoretically get some sort of compensation. But when you look at the record, the compensation you get isn't worth it. With opensource, you have both the responsibility and the privilege to run your own install servers and backups. And you don't have the guarantees that seem to fool the bean counters. Are you using site specific kickstart config files that install local yum config files, ssh keys, sendmail setup and sudo config files so your admins can access the hosts without typing pass words? Yes, to all. Unfortunately that regime isn't 100% adhered to, which is something we work on. Equally unfortunately, we have had to give the footwork guys sudo access to a limited set off commands. Sudo with or without passwords have different security implications, we've landed on "with". "With" is not a bad alternative. Balancing resources is always a problem. No matter how you choose, sometimes bad stuff happens. Again, if accounting or management is coming after you, point to the actual results (not the promises and fudged guarantees) that could be obtained from the proprietary alternatives. F/OSS, while better than the alternatives, is not some magic utopia. Now, I think they're handling this pretty well so far. I'm considering things from the overall perspective. A certain "Proprietary" vendor has put the entire world's infrastructure at risk, and they've managed to delay things with weird legal and political games for more than ten years, putting society at further risk. What we hear in public is not the worst that could happen (or is happening, really), and anyone whose infrastructure is dependent on that "Proprietary" vendor, is basically living on borrowed time and illusions. It's definitely time to run a tight ship now. [...] Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
japanese keyboards and xorg (and macs?)
When running the installer, my Japanese iBook takes the US keyboard map, which is not totally evil, just a little bit confusing at times. In openBSD, I have been able to get most of the keyboard mapped to the Japanese keyboard, merely by copying the jp keyboard mapping file from /etc/X11/xkb/symbols into /etc/X11/xkb/symbols/macintosh and setting the keyboard to jp in xorg.conf. But two keys remain unaccessible, and one key does not remap, leaving me unable to type underscore, right bracket, and right brace in the Japanese map. Again, on the openBSD system, I tried the xev utility, and determined that the two keys that are unaccessible are just not generating events. If I get Fedora 9 running on this iBook I expect I'll see the same thing. Has anyone here has seen this kind of problem? or could anyone give me some suggestions? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
multibooting ppc with Mac OS 9 & X on an ancient iBook
I recently tried installing Fedora 9 on an iBook that I need to boot both Mac OS 9 and Mac OS X on. I also have some other partitioning constraints, so I ended up with 6 partitions before I started the Fedora 9 install. Fedora 9 happily created lots more partitions, but then Mac OS 9 (and X, IIRC) refused to boot. Couldn't recognize the format of the disk. (Might have helped if I had got a 120 G HD instead of a 160 G HD?) Also, on a separate iteration, when I tried to force the yaboot partition to 1MB, that also apparently made the partition map unacceptable to Mac OS 9. I had to wipe the disks with the Mac OS 9 formatter and start again. (Lost a day or so of my time, but no data.) Since I thought I had time, I tried an install of just Fedora, but the current partitioning software wouldn't create the partition for yaboot any smaller than 16MB, IIRC, and then it wouldn't install yaboot in anything bigger than 1MB. I am currently successfully multi-booting Mac OS 9 & X (Jaguar) and openBSD, but openBSD is not using yaboot. I give the four-finger salute on startup and type "boot hd:,ofwboot /bsd" at the open firmware prompt. That doesn't really bother me, even though I have to remember that the keyboard map is US and doesn't match the Japanese keyboard. :-/ openBSD seems to take a little more nursing than I currently have time for, and I am primarily interested in getting the Gimp and openoffice.org running. Well, probably some of the edutainment stuff, as well. In about two weeks, one of the partitions will be freed, so I should have three partitions to give Fedora 9, and I am thinking of trying again. But I won't have the day or so necessary to re-build the Mac OS 9/X sides of things if the Fedora partitioning software makes the map unreadable to Mac OS 9 again. So, I am wondering a couple of things: One, Is anyone is currently successfully multi-booting Mac OS 9, Mac OS X, and Fedora 9 on any system, especially one with a boot HD larger than 120G? Two, would it be possible to boot with openBSD's approach, invoking an openfirmware script on the Mac OS 9 boot disk? (I haven't been able, yet, to untangle the web of what happened when yaboot became usable.) Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: Infrastructure report, 2008-08-22 UTC 1200
On Aug 22, 2008, at 9:00 PM, Paul W. Frields wrote: [Information on the intrusion, etc.] This time through has been a little bit rough. Nothing like a first time ... Not all the information I want, yet, but definitely much better than what Microsoft gives out. Thanks. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: Infrastructure status, 2008-08-19 UTC 0200
On Aug 22, 2008, at 8:45 AM, Jim Cornette wrote: Michael Schwendt wrote: On Wed, 20 Aug 2008 23:35:20 -0400, Jim Cornette wrote: I myself am not on the announce list Why not? Because I never thought the list was informative. This is a perception that needs correcting. Unfortunately, the people that foster that perception are anything but under the redhat umbrella. (Happen to be trying to tear it down, right now, if you know what I mean.) Just look at their record to see the difference between what they preach and what they do. In many parts of Japan there is a over-the-wire emergency broadcast system that pre-dated the internet. (And has since been at least partially re-purposed in many small towns to bring broadband in before any independent ISPs showed interest. I think, if you think about it, you can think of some implementation issues there.) The original purpose of the system was to broadcast earthquake and storm (typhoon, etc.) warnings and such. There are some lists in which no news is good news. That does not mean the list is useless. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: sudo GUI frontend
This touches on one of my gripes about Fedora (and RHE) -- the GUI authentication stuff is oriented towards su instead of sudo. On Aug 20, 2008, at 8:21 PM, Anne Wilson wrote: On Wednesday 20 August 2008 12:05:34 Gilboa Davara wrote: Hello all, Can anyone please point me to a good sudo GUI front-end? (Password dialog box). While I'm using KDE 4.1, I don't mind using a gtk box. I tried using kdesu with or w/o the kdesurc - but it doesn't seem to work. '/bin/su -' does work, though in some cases you will need to add $u (or is it %U?). These su dialogues requires the user to know the root password. If the user knows the root password, the user can log in as root. If the user picks up a keylogger and authenticates, whoever owned that keylogger now owns the machine. If you think carelessly, you might think that a running keylogger implies that the machine is already pwned, but you're ignoring the non-ideal behaviors of browser technology. Also, there's the less tangible benefit, in that the correct use of sudo raises the general awareness that we shouldn't surf the web naked. (So to speak.) -- Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: (slashdot)Package Managers As Achilles Heel
Just being alarmist, here, On Aug 18, 2008, at 5:42 AM, Mikkel L. Ellertson wrote: Björn Persson wrote: Mikkel L. Ellertson wrote: Marcelo M. Garcia wrote: http://it.slashdot.org/article.pl?sid=08/07/10/227220&from=rss Two things bother me about this. First of all, most users are not using the same mirror all the time, so there would only be a brief window that the system would be vulnerable. The second thing is that yum is not going to install an older package, and the package version is not dependent on the file name. It is part of the information in the RPM. So they could delay the installation of an update on some systems. By default, yum picks a mirror at random from the mirror list to help spread the load on the mirrors. I found this in their FAQ: | Q: I use a service that distributes my requests to different mirrors for my | distribution (like MirrorManager). That means I'm not vulnerable, right? | A: The good aspect of these systems is that it may spread your requests | across multiple mirrors in the normal case. However, when testing some of | these systems, we were able to target the clients that used our mirror and | exclude them from using other mirrors. This means that if an attacker wants | to target your organization, these services may help the attacker do so. It's not clear whether Yum is vulnerable to getting locked to the malicious mirror, or how they did it. Björn Persson By default, the mirrir list is fetched from http://mirrors.fedoraproject.org/mirrorlist?repo=fedora- $releasever&arch=$basearch and a mirror is picked at random from the list. You can override the mirror used with the fast-mirror plugin, or by editing the repo configuration file. So yum is probably not one of the clients they could do that to. Can yum install something that would overwrite its own configuration file? Now, if you used a DNS bug to hijack mirrors.fedoraproject.org, then you could lock in the mirror used by suppling a list that only contained pointers to the malicious mirror. Mikkel -- Do not meddle in the affairs of dragons, for thou art crunchy and taste good with Ketchup! -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: Cannot boot after the last updates
On Aug 16, 2008, at 10:40 AM, Craig White wrote: On Sat, 2008-08-16 at 10:31 +0900, Joel Rees wrote: On Aug 16, 2008, at 9:52 AM, Paul Smith wrote: On Fri, Aug 15, 2008 at 8:12 PM, Craig White <[EMAIL PROTECTED]> wrote: After the last updates were installed, I cannot boot in my machine: it stops when the line GRUB appears on the screen. Any ideas how to repair the problem? I am running F9. probably too much to ask people to search the archives... https://www.redhat.com/archives/fedora-list/2008-August/ msg01404.html I do not have any rescue/installation disk. Can it be done with Knoppix? If so, how? I have tried # grub-install /dev/sda Probing devices to guess BIOS drives. This may take a long time. Could not find device for /boot: Not found or not a block device. # Well, let's hope that means grub left the drive alone. How did you install? And, if you've booted the box under knoppix, can you mount the partitions with knoppix? most importantly, he would have to chroot to get it to install/find /boot Heh. I'd be likely to boot knoppix to download and burn the rescue CD image. I always seem to take the long route around. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: Cannot boot after the last updates
On Aug 16, 2008, at 9:52 AM, Paul Smith wrote: On Fri, Aug 15, 2008 at 8:12 PM, Craig White <[EMAIL PROTECTED]> wrote: After the last updates were installed, I cannot boot in my machine: it stops when the line GRUB appears on the screen. Any ideas how to repair the problem? I am running F9. probably too much to ask people to search the archives... https://www.redhat.com/archives/fedora-list/2008-August/msg01404.html I do not have any rescue/installation disk. Can it be done with Knoppix? If so, how? I have tried # grub-install /dev/sda Probing devices to guess BIOS drives. This may take a long time. Could not find device for /boot: Not found or not a block device. # Well, let's hope that means grub left the drive alone. How did you install? And, if you've booted the box under knoppix, can you mount the partitions with knoppix? -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: Cannot boot after the last updates
On Aug 16, 2008, at 9:51 AM, Craig White wrote: On Fri, 2008-08-15 at 17:39 -0700, Dave Stevens wrote: On Friday 15 August 2008 05:12:54 pm Craig White wrote: On Fri, 2008-08-15 at 20:00 -0400, Paul Smith wrote: Dear All, After the last updates were installed, I cannot boot in my machine: it stops when the line GRUB appears on the screen. Any ideas how to repair the problem? I am running F9. probably too much to ask people to search the archives... https://www.redhat.com/archives/fedora-list/2008-August/ msg01404.html Craig and too much to expect upgrades to not screw up the boot process? no - and this is a second report but unless someone invests the time to figure out why it happened (like Peterboy reported), then there's little to be learned (and thus nothing to be fixed). Update or upgrade? If he's going from Fedora 8 to 9, yeah, he should be prepared to re- install grub. If he's just doing a yum update on a running Fedora 9, I would hope that nothing in an update would muck with the boot sectors. Craig -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: avasys drivers and PM-A850 scanning
(Apologies for the threading.) Joel Rees wrote: > Sorry, I was not very clear. > > I tried scanning from the gimp. IIRC it was xsane that complained > that it could not find the driver. ok. did not mention what you tried either. :0) It was a late night when I posted, I'm running^H^H^H^H^H^H^H trying to run four or five projects at once between semesters, I'm closing on 50-ish (8-p)), ..., lots of excuses, you see. |-( so i ran a google for you on 'epson pm-a850'. Thanks. this relates to 'epkowa 2.10.0' http://www.sfr-fresh.com/linux/misc/sane-backends-1.0.19.tar.gz:a/ sane-backends-1.0.19/doc/descriptions-external/epkowa.desc note lines 25 -> 45. I'll work through that and see what I can find. Epson's Japanese site had this list of common printer models on the download page for the scanner driver for Mac OS X (the family computer here): CC-500L / CC-550L / CC-570L / CC-600PX / CC-700 / PM-A850 / PM-A850V at http://www.epson.jp/dl_soft/readme/819.htm so that gives me something else to work from. if you have not seen this before, log; http://www.google.com/advanced_search? hl=en&output=linux&restrict=linux run search against lines 26 -> 34. also, line 43 [unless that is what you were using]. have you tried line 45? That looks familiar. I'll have to look in my Fedora 7 partition, to see if that's what I've got configured there. And I guess this thread should get moved to a CUPS list or forum from here. Thanks again. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: ps2 mouse pointer wanders away
Ed Greshko wrote: Joel Rees wrote: I find that I often lose control of the mouse pointer. It seems to happen more when I'm seeding torrents, or running a diff or a message digest check on the download, or doing other things that load the system. Sempron 2600, single processor, VIA KM400something+8237 chipset, 760M RAM. Generic PS-2 mouse and keyboard. I can plug in a USB mouse and use that, and sometimes plugging the USB mouse in brings the PS2 mouse back, as well. In your /etc/X11/xorg.conf file what is mouse protocol set to? No mouse definition in that file whatsoever. I thought I'd ssh the whole file over to this box, to prove it, and discovered in the process that, while the interface was set up properly by the installer, the hosts file was set up to tell the fedora 9 box that it was the external interface of my dsl router. That boggled my mind enough that I forgot to take a copy of what I found in /etc/hosts, so I'm going to have a hard time filing a decent bug report on it. And, in the process, I discovered that there seems to be an intermittent bug in the language setup, such that I was not able to switch from Japanese to English so I could make sure I understood the firewall settings widget. And, now, using one of the widgets instead of working directly on the configuration file, I find that the box has forgotten what the dns servers were, so the box is completely off line. And there's something wedged in the firewall, as well, apparently, can't even ping it. I was going to start a rant about how I didn't expect Fedora 9 to be rawhide, but I helped my wife with the laundry instead and have calmed down a bit. This is, after all, the only way to figure out how to deal with the new security model. I expect that we'll end up proving that the whole idea of access control lists is a specification level bug, but only time will tell, and I really can't think of any other way of proving it is, other than to subject a mixed technical user group like this to the spec. Something good should shake out of the process, but it won't look much like the theory they teach at school. I've seen what I think you are describing when set to infrequently seen/used setting. FWIW, the one I use most often is... Option "Protocol" "IMPS/2" Thanks for reminding me where that file is. I found an example of the mouse definition section on the web, but the problem now is getting the box back on the web first. I find myself wondering whether I should bother, though. If I've got the time to nurse the box this way, maybe I should just go straight to rawhide. Thanks. Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
ps2 mouse pointer wanders away
I've got Fedora 9 up and running multi-boot while I wait for time to back up the data from Fedora 7 and install it properly. I find that I often lose control of the mouse pointer. It seems to happen more when I'm seeding torrents, or running a diff or a message digest check on the download, or doing other things that load the system. Sempron 2600, single processor, VIA KM400something+8237 chipset, 760M RAM. Generic PS-2 mouse and keyboard. I can plug in a USB mouse and use that, and sometimes plugging the USB mouse in brings the PS2 mouse back, as well. Anyone else noticing this kind of stuff? (I'm not much on filing bug reports. Seems they're always changing the front door, and I never can find my way around until after I've solved the problem. So I'm hoping it's probably someplace I need to tweak in X11, and I guess I'm being too lazy to go digging into the X11 settings today.) Joel Rees (waiting for a 3+GHz ARM processor to come out, to test Steve's willingness to switch again.) -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: avasys drivers and PM-A850 scanning
g wrote: Joel Rees wrote: > Or, preferably, point me > to a properly open solution? have you considered xsane? Sorry, I was not very clear. I tried scanning from the gimp. IIRC it was xsane that complained that it could not find the driver. When I went looking, that's when I found out that the avasys drivers are an opaque blob with a license that only allows reverse engineering far enough to comply with the LGPL, no source available. I suppose I have been using the avasys drivers already in Fedora 7, but if there are properly open drivers, I would like to use those instead. (I am not happy with this printer. Stupid profit-on-the-ink marketing games. Still, I'd like to use the scanner since I have it.) Joel Rees On 平成 20/08/13, at 23:09, Joel Rees wrote: I found a thread about Kooka being taken out of the distro. I've been using Kooka with my Epson PM-A850 all-in-one scanner- expensive-printer, but I have not been able to scan at 1200 dpi. Now I want to scan at 1200 dpi. I was able to take two scans, then it wedged. I hit the full scan button and nothing happens. Move reasolution back to 300 dpi and no problem. Back again to 1200 dpi and no scan. That was on F7. So, I decided to boot up my experimental install of F9, and I discover, no Kooka. Tried the scanner tool, but it complains that it can't find the sane library for my printer, or that there is some conflict. I really haven't been messing with loading lots of stuff, so I figure it's not likely to be a conflict, but I'll check later, if I can break out some time. Then I went hunting and found out that Kooka has been dropped. I also found Avasys. It's a blob. LGPL-compatible licensed. Gag. Anyone care to comment on Avasys's blobs? Or, preferably, point me to a properly open solution? I'll keep looking tomorrow, it doesn't look like I have time to do anything further today. But if anyone cares to give me a few clues before I go hunting tomorrow, I'd appreciate it. (Sempron 2600, Pasokon Koubou's private Librage brand with a K4 mobo, IIRC.) Joel Rees Joel Rees (waiting for a 3+GHz ARM processor to come out, to test Steve's willingness to switch again.) -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Re: apple boot partition causes error in install
Status: On 平成 20/08/01, at 10:45, Joel Rees wrote: Been mucking around with my vintage clamshell ibook, trying to get fedora 9 to install in a multiboot configuration. Today I hit what appears to be a show-stopper for me. I try to cut a 1 MB apple_boot partition and the partitioning software gives me an error, says the size change can't be communicated to the kernel until the next reboot. If I cut an apple_boot partition larger than 1MB, the installer will not proceed, telling me the partition is too large. I don't (yet) have enough experience with yaboot to try working around this catch-22 in the installer. I figured I had enough time to install Fedora single-boot on the iBook for a couple of days and see what I could figure out. (Installing from the net-install CD.) But I hit another snag. This time, I decided to just let the installer do the default partitioning scheme. It build, IIRC, the apple boot partition and a big partition for LVM. No swap partition. Then it gave me an error again where the partitioning stage could not proceed. Sorry for not recording more information, but it looked like LVM itself is not part of the net-install CD. At that point, I figured that it was going to take more customization than I have time to learn how to do with Fedora, so I'm going to go with openBSD for this project. But, Anyone successfully running F9 on an iBook? I'd still like to hear from anyone successfully running F9 on an iBook, or even on any PPC Macs. Is multi-boot possible? Issues/ successes with the net-boot CD? or with one of the other install CD/ DVDs? Is there a Fedora list specifically for Macs? Joel Rees Joel Rees (waiting for a 3+GHz ARM processor to come out, to test Steve's willingness to switch again.) -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
avasys drivers and PM-A850 scanning
I found a thread about Kooka being taken out of the distro. I've been using Kooka with my Epson PM-A850 all-in-one scanner- expensive-printer, but I have not been able to scan at 1200 dpi. Now I want to scan at 1200 dpi. I was able to take two scans, then it wedged. I hit the full scan button and nothing happens. Move reasolution back to 300 dpi and no problem. Back again to 1200 dpi and no scan. That was on F7. So, I decided to boot up my experimental install of F9, and I discover, no Kooka. Tried the scanner tool, but it complains that it can't find the sane library for my printer, or that there is some conflict. I really haven't been messing with loading lots of stuff, so I figure it's not likely to be a conflict, but I'll check later, if I can break out some time. Then I went hunting and found out that Kooka has been dropped. I also found Avasys. It's a blob. LGPL-compatible licensed. Gag. Anyone care to comment on Avasys's blobs? Or, preferably, point me to a properly open solution? I'll keep looking tomorrow, it doesn't look like I have time to do anything further today. But if anyone cares to give me a few clues before I go hunting tomorrow, I'd appreciate it. (Sempron 2600, Pasokon Koubou's private Librage brand with a K4 mobo, IIRC.) Joel Rees Joel Rees (waiting for a 3+GHz ARM processor to come out, to test Steve's willingness to switch again.) -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
apple boot partition causes error in install
Been mucking around with my vintage clamshell ibook, trying to get fedora 9 to install in a multiboot configuration. Today I hit what appears to be a show-stopper for me. I try to cut a 1 MB apple_boot partition and the partitioning software gives me an error, says the size change can't be communicated to the kernel until the next reboot. If I cut an apple_boot partition larger than 1MB, the installer will not proceed, telling me the partition is too large. I don't (yet) have enough experience with yaboot to try working around this catch-22 in the installer. Anyone successfully running F9 on an iBook? Joel Rees -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list