jdd wrote: > Shriramana Sharma wrote: > >> I was wondering why it is often said that NTFS support on Linux is >> incomplete when there exists this package: >> http://www.jankratochvil.net/project/captive/ > > > I could never make it work... > > jdd > > > Just tested, and it works :-)
In short: get and install the rpm from the above site. (captive-static-1.1.5-0.i386.rpm <http://www.jankratochvil.net/project/captive/dist/captive-static-1.1.5-0.i386.rpm> - not tested the tar.gz one) Then as root, run the command: captive-install-acquire which will try to find the *proprietary* drivers needed (ntfs.sys and ntoskrnl.exe being the 2 more important) They should be copied here: */var/lib/captive/* (http://www.jankratochvil.net/project/captive/man/captive.pod.html, "OPTIONS") As the search was taking some time, I did use directly the drivers in the Windows partition; better: no proprietary driver would be copied in Lin. ** So, using the following command, i was able to read &WRITE to my hda1 partition (NFTS partition on a dual boot box): captive-cmdline --disk --rw --filesystem=/windows/C/WINDOWS/ServicePackFiles/i386/ntfs.sys --load-module=/windows/C/WINDOWS/ServicePackFiles/i386/ntoskrnl.exe /dev/hda1 Where /windows/C is the way "C:" is defined (or i defined?.?) in fstab Where the --filesystem and --load-module paths are the location of ntfs.sys and ntoskrnl.sys (note that this could be in the Windows dir, but as Service Pack have them, it was better to use these alternate files just in case of... The windows defaults are WINDOWS/system32/ntoskrnl.exe and WINDOWS/system32/drivers/ntfs.sys) And where /dev/hda1 is the ntfs partition to be read...and WRITTEN!! quote: You do not want to use captive ntfs. It may work for some people, but trusting a driver written for a different operating system with different kernel stack size, compiled with a different compiler, no source available to debug, etc. is nothing I would do on any of my systems. Carl-Daniel Yes, that's a risk, and surelly should be avoided. Anyway, it can be on occasion a great ressource. So we can *finally* write a NTFS partition. But using what Carl-Daniel reminds us... Thus the real "hack" of the NTFS writing has not been done yet. Whatever happens, wish it won't last 20 years. But for now, i think it can be pretty handy for us susers. Cheers, Patrick M.
