More DVDs than Debian? - Forums Linux |
- More DVDs than Debian?
- CUPS admin
- Mutiple filename changes
- Puzzled about KDE or GNOME or what?
- Email server setup
- Fedora Core 5
- FC11, libusb, HP scanner (G4050)
Posted: 22 Sep 2009 12:58 PM PDT Manuel Rodriguez wrote: Difficult task. You can certainly put a LOT onto a 1TB drive. But certainly not all since the beginning. There have been too many distros and too many versions of distros since 1991. I do believe Debian is the largest. Just collecting all of the data could take months. |
Posted: 22 Sep 2009 04:41 AM PDT On Sep 22, 2:36pm, Unruh <ubc.ca> wrote: Now, now: the sudo convention is very handy for keeping users away from unnecessary use of the 'root' user, especially for shell access. rose is probably not listed in /etc/sudoers with the appropriate permissions, while chrisc is. If you need root shells, and are a member listed in /etc/sudoers, you can do 'sudo -s -H", which works pretty well. And that's excellent advice. Those 'system authorizaition' menus basically manage /etc/sudoers. But so far, Magnate didn't say what OS they're running, nor if their CUPS and in turn their related ghostscript and other drivers are. So Magnate? If you can give us some hints about which Linux flavor, and which version of CUPS and in turn the Epson setup, and look in the /var/log/ directories for information about what is failing, it might be more helpful overall than simply getting help with this particular workaround. |
Posted: 21 Sep 2009 08:29 AM PDT Aragorn schrieb: There exist chances to implement such additional information, similar to the implementation of the long filenames hack. See above. The *physical* layout of a FAT drive can be FAT12, FAT16 or FAT32, with the according file *allocation* table format (FAT entries consist of 12, 16 or 32 bits). The *logical* file *descriptor* format in the directory entries is the same for all FAT sizes (32 bytes per entry). VFAT added long filenames to the FAT descriptors, without changing the specification of the underlying FAT12/16/32 system. The difference between file descriptors and long filename entries is encoded in the file attributes bits, so that every VFAT filesystem can be damaged by the use of older (MS-DOS) tools, which are not aware of the long filenames hack. Since every file has a unique starting cluster in the FAT, this number could be used as the unique inode number of every file. A POSIX system could build the missing inode descriptors (in detail the exact file size) from a scan of the filesystem, executed whenever the filesystem is mounted. Hard links can be implemented by additional file descriptors, refering to the same start cluster - they would be treated by Windows systems as corrupt crosslinks, so that a problem, similar to the long filenames problem, would exist with such hard links, when Microsoft tools check the filesystem. Perhaps hard links can be hidden from such unwaware tools by special settings of the file attribute bits, similar to the implementation of the long filename entries. Please note: I do not encourage the implementation of hard links this way, I only want to point out chances for such an implementation. Microsoft has doented the old (8.3) DOS filename as "alternate", i.e. the primary VFAT filename is the long (UTF-16) version. A POSIX VFAT implementation should apply the *same* ranking on the filenames, and IMO this handling already *is* the Linux standard. A POSIX system can (should?) treat the two names as links to the same file, and act upon these links according to POSIX rules. I'm not sure about the best POSIX equivalent, but IMO the long filename is nothing but a symlink to the (immediately following) file descriptor. As already mentioned, a POSIX filesystem should ignore the alternate name and only deal with the long name. As long as no long name exists, the 8.3 name can be used instead, but on a rename a long name with the new name and spelling should be created. With regards to the case-insensitive handling of filenames, it only matters when Windows compatible access is really required. IMO this can happen only when Windows applications, scripts or doents come into the play, which want to refer to files in an case-insensitive way. Normal operation of a POSIX system can respect the case of the filenames, without breaking anything. The alternate (your primary) filename is stored when a file is created. There exists no requirement to change that name at any later time. Alternate filenames can be used as a second chance, when no long filename matches a search pattern (see my symlink suggestion above). Windows may update the alternate filename as well, when a file is renamed *and* the new name is compatible with the 8.3 naming rules. In this case an eventually existing long filename *may* be removed, just like a POSIX system would drop one of two identical links to the same file, in the same directory. But this only is a *behaviour* of the OS, not a requirement of the file system. Applying different rules, in a POSIX implementation, will not violate any FAT specs. The only real requirement for filenames on FAT volumes is, that no filename entries can exist in a directory at the same time, which *only* differ in case. POSIX imposes the essentially same restriction, different only in the handling of case sensitivity (no directory can contain multiple entries of exactly the same name). You seem to favor an exact emulation of the Windows behaviour on Linux. I see no need for such a slavish emulation. Its perfectly sufficient to conform with VFAT specs WRT the *stored* information only. I also see a chance for mounting options, indicating how Linux should treat an specific VFAT drive. Then it's up to the user to enable case sensitive interpretation of filenames, and the implementation of links and POSIX permissions, as far as possible with the stored VFAT data structures. DoDi |
Puzzled about KDE or GNOME or what? Posted: 21 Sep 2009 01:53 AM PDT On Wed, 23 Sep 2009 03:47:50 +0000, Chris F.A. Johnson wrote: I really thought criticism would smoke Sid out. I'm very disappointed. C'mon Sid! -- d.b. cooper |
Posted: 20 Sep 2009 10:01 AM PDT Aragorn wrote: Yes, just got it fixed. It was a laptop configuration problem. I agree. But it never really gave an error message, it just hung after login. But in any case I just got it fixed. All seems to be working. Way cool. Roger |
Posted: 18 Sep 2009 11:26 PM PDT On Tuesday 22 September 2009 15:53, someone identifying as *Sydney* wrote in /comp.os.linux.setup:/ What Moe was asking about is whether your wireless device might have a /native/ driver in Fedora 11. However, /ndiswrapper/ should already be in FC 5 as well. On this old Mandrake system here, with kernel 2.6.3, it is also already present. <snip> You need to use the /ndiswrapper/ that came with the kernel you are using, i.e. this one: /lib/modules/`uname -r`/kernel/3rdparty/ndiswrapper *Note:* Those "quotes" in the above path are not apostrophes but rather what you would in French call an "accent grave". The shell will substitute the command between these two accent graves by its output. Hope this helps. ;-) -- *Aragorn* (registered GNU/Linux user #223157) |
FC11, libusb, HP scanner (G4050) Posted: 17 Sep 2009 07:10 PM PDT On Thu, 17 Sep 2009 22:10:14 -0400, K.A. <net> wrote: With my canon mp150 usb scanner/printer, scanimage -l does not show it either, yet xsane does detect it, and allow it to be selected, along with my usb webcam. Try xsane, to see if it is detected. Regards, Dave Hodgins -- Change nomail.afraid.org to ody.ca to reply by email. (nomail.afraid.org has been set up specifically for use in usenet. Feel free to use it yourself.) |
You are subscribed to email updates from TextNData Forums - Linux To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
Google Inc., 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States |