Knoppix Variant Offers Full NTFS Write Support
mache writes "Full NTFS write support for Knoppix is under discussion on Knoppix Ideas forum and it looks that Knopper will include Captive into Knoppix 3.4. The best part of Live CD with full NTFS write support is that it actually exists in LinuxDefender, a remastered Knoppix distribution made by Bitdefender, presented at LinuxConf 2003, the annual Romanian Linux Users Group (RLUG) conference."
..is how it can offer better hardware detection and often better features than other, "commercial" Linux distros?
Anyone has internal information on how Knoppix is developed and maintained?
Doesn't the kernel now support this natively?
If I'm skimming the Captive homepage to quickly, but it seems to me like Captive is using Microsoft DLL's to read/write NTFS filesystems.
Seems to me like that would or will violate the Microsoft EULA and leave Knoppix users open to problems if MS changes parts of these DLLs in subsequent service packs or releases.
Conformity is the jailer of freedom and enemy of growth. -JFK
As always, NTFS read-only works fine, writing is very limited unless you want to corrupt your filesystem. Knoppix would be using an approach that uses Microsoft's ntfs.sys to handle writing. Of course, you need Windows installed (or at least a copy of ntfs.sys) for this to work.
LOAD "SIG",8,1
I mean, realistically, wouldn't the best part of full read/write NTFS support be the fact that it exists?
obviously no. If someone has a valid XP licence they should be allowed to use it in any way they wish to and this includes the NTFS driver.
Due to DMCA and other wonderful stuff, you'll probably never see a native OSS driver for NTFS...it's an encrypted filesystem. AS for WHY...almost all PCs sold nowdays come with NTFS...the OEMs sell them that way. Short of asking a user to reformat their drive, you can't get fat32 on newer boxes...and Knoppix is aimed at users who don't want to/can't modify their windows install.
> Allthough this will be extremely usefull for the people having to cope
> with ntfs, i'd rather wait until the kernel supports it fully (there's
> allready a "partial" driver in the kernel 2.6). But personally, i just
> stick to FAT32.
The usefulness of this is primarily geared toward situations where NTFS is
already extant (e.g., OEM installs of WinXP). In these scenerios, if you
want to multiboot and share data between the two OSes, use Knoppix as a
rescue system, or anything along those lines, you *need* read/write NTFS
support. You don't need this if you have your choice of filesystems,
because you can just use another filesystem, but if you are in a situation
where you need this, nothing else will do. So it's important. It's
especially important for Knoppix, which is often used as a rescue system;
now it can be used as a rescue system for NT/2K/XP, as well as for 9x/Me.
No, you wouldn't choose to use this on a new install when you have your choice
of filesystems. For that you'd pick Reiser or ext2/3 probably, or FAT for a
data partition in a multiboot scenerio (since that gives the best compatibility
and works with every major OS and most minor ones as well). But that's not the
intention of captive-ntfs. It's for working with existing filesystems.
Cut that out, or I will ship you to Norilsk in a box.
> It says they use ntfs.sys and even ntoskrnl.exe from your XP partition.
> Wondering if there are legal problems with this.
One supposes that if you have an NTFS partition with these files on it, the
files are licensed for you to use and therefore legal. (If not, you have a
problem that goes beyond captive-ntfs.) Unless there is some specific
verbiage in the EULA that expressly prohibits use of the drivers when the
NT kernel isn't running, or some such restriction, I'm not sure what the
legal problem would be. I find it difficult to imagine that the MS legal
team would have dreamed up that kind of restriction, since it's not the sort
of thing they would expect people to do. What happens when the EULA is
revised is another matter, but this would have to get on the MS radar for
that to happen, which will take at least a year, then another six months or
so until the next update/revision cycle, and hopefully by then the native
read/write NTFS support in the 2.6 kernel will be of such quality as to make
the whole point moot.
Cut that out, or I will ship you to Norilsk in a box.
NTFS is not an encrypted file system. Windows' layered file system architecture supports encryption, yes, but that's not quite the same thing.
I think it's important to note that the issue with the current Linux kernel's NTFS support is its capabilities, not its quality. It can only write to a file without increasing its size - and no creating or deleting files or directories. It's fine if you want to mount a loopback file on an NTFS partition (say if you want to install Linux onto a Windows machine without repartitioning - a few distributions support running Linux out of a file on the Windows partition in this way).
It's also fine if you just want to read files from an NTFS partition - I use it for this all the time.
I hope accomplishing this didn't involve modifying undocumented internal structures.
Raymond Chen will be pissed.
Monstromart: Where shopping is a baffling ordeal
Could a kind soul who has already gotten it please make a torrent out of it? Thanks!
I feel fantastic, and I'm still alive.
Will this driver speed up the creation of a native one?
Since now NTFS.SYS is working under Linux although through emulation it should be quite easy to spy to what it's doing, and try to improve the native driver based on that.
- Boot from CD to try it out.
- Convert to dual boot. There would be a utility to re-partition, install and configure for dual-boot. Let the user keep it dual-boot while they find substitutes for any Windows-only programs that Wine can't handle.
- Convert to Linux only
You could give these out like AOL disks and slowly convert the installed base. There could be a utility to detect existing win32 programs and check their status in the Wine application list.This would be the logical extension to Bruce Perens' UserLinux idea.
It is legal. The technique used by captive has been used by other products, such as the Systems Internals NTFSdos product. MS works with and even promotes this company so they can't now complain when others do the same thing.
Not to mention the fact that Windows has some stupid approach to NTFS encryption. All you have to do to crack a NTFS encrypted file is just to access it with the Linux-NTFS driver.
M$ really needs to get their act up.
Now that there is reliable NTFS write support, maybe we could get a kernel modification to use the pagefile.sys as a swap partition.
mkswap
swapon
What about using the windows temp directory for storage of highly used apps and libs?
-metric
> I think it's important to note that the issue with the current Linux
> kernel's NTFS support is its capabilities, not its quality.
Capabilities aren't part of quality? Do you want a car that can't make right
turns? What about a car that can't make turns at all -- it would still be
fine for driving straight ahead, and you could even put it in reverse...
> It can only write to a file without increasing its size - and no creating
> or deleting files or directories.
This limits its usefulness in important ways. Not that it is if no value,
but I would really like to think that in a few months someone will figure
out how to (safely) lift these restrictions.
Cut that out, or I will ship you to Norilsk in a box.
They are. But NTFS is a huge project. It takes a lot of time. Today it's developed as a hobby, couple of hours a week or not even that much.
> I would really like to think that in a few months
> someone will figure out how to (safely) lift these restrictions
The developers said in the forums: somebody having lots of time must start active coding (no need for reverse engineering any more) or hire coders and finance the development.
ntfs also supports bigger file sizes and better performance for bigger partitions along with fact that its a journaling file system like ext3 and reiser.
Remember Microsoft is a Corperation of the USA...linux hackers are just Citizens [or maybe even Foriegn citizens...the horror]
- But do you think they'd tollerate full R/W support from Linux without trying to break it?
Breaking NTFS is part of the technology evolution. ext3 "broke" ext2, reiser4 "broke" reiserfs, etc. So why you can't mess your filesystem? Because everytime these changes happen the on-disk filesystem version, signatures are also incremented. Not doing so Microsoft would damage itself, for example by W2K, XP destroying Longhorn NTFS. If this version change is detected then no data harm is possible. The old, unmaintained NTFS driver didn't check the NTFS version so when W2K came out the NT4 driver destroyed the W2K NTFS. Kernel developers fixed this only much later on thus the damage was already done.About WinFS: it's not a filesystem. It's a storage layer on top of NTFS.
Such a restriction would also block current, commerical rescue-offerings, at least, UNLESS they licensed the driver themselves, somehow. And you would be committing a felony unknowingly just by using such software, until those software companies declared their (up to now anyways) secret licensing of this driver. It just isn't worth the political trouble for Microsoft. As long as captive-ntfs requires ntfs.sys from microsoft, Microsoft would be harming its friends more than its enemies by fighting it.
:)
I bet they regret not patenting ntfs though
Did they file patent on winfs??
Or M$ just uses the next service pack/patch to revise the EULA.
I don't mean this a troll, but when has M$ ever cared about stabbing a "friend" in the back? They do it every time it becomes convenient. I remember a cellphone company story a while back who partnered with M$ and a last I heard, we suing for stealing technology as just one example.
obviously no. If someone has a valid XP licence they should be allowed to use it in any way they wish to and this includes the NTFS driver.
The flaw in your argument: use of "obviously" and "should" instead of giving a reason why Microsoft would allow this. (Allow? Yes. Until their EULA is declared illegal or void, it's a legally binding agreement with MS software users - even when it's stupid.)
Not allowed to use it without an installed copy of NT (or another OS with that file) - probably legal. Not allowed to use it at all even if you own a copy of NT... probably not.
Then again, it could be a grey area... it's like the "am I allowed to mod my XBox, am I allowed to help other people mod theirs to use it for what the want" issue.
Here's an example when you might want to use Captive....
You do video editing, XP has some good tools, Linux has some good tools. Video editing is best done on a file system that can handle large files like EXT2/3, Reiser, or NTFS (FAT 32 has a 4GB maximum file size)
Since XP doesn't support any of the Linux filesystems in read/write mode (at full speed) even with opensource add-on drivers there are 2 options for a video editor who wants to use both paltforms.
#1 buy 2 computers and network them and pass the project back and forth over the network
#2 format a common drive in NTFS, put the project on that drive and dual boot between XP and Linux using Captive in Linux to access the Filesystem in read/write mode at full speed.
- On the menu go to Knoppix -> Root Terminal
- type qtparted and press Enter
If someone can use Partition Magic, they should be comfortable with QTParted, it even resizes NTFS.vi +
> Also NTFS is preferable over FAT because FAT has no concept of file ownership.
This would only matter for situations with multiple users who don't trust one
another, which basically either means servers (where if you're using Windows
I feel sorry for you) or some kind of unusual desktop situation. I've seen
many, many Windows desktops, but they were all either used by one person or
were used by multiple people who all shared the same user account. I've not
yet seen anyone actually using the Windows multiuser stuff in real life. I'm
sure it happens, but it is not the typical situation.
And quite obviously the multiuser stuff is unimportant in a multiboot scenerio,
since that's invariably a computer used by one person, or else different people
use different OSes (e.g., the spouse uses Windows). Also, the other OS can
mount the NTFS partition (albeit perhaps in read-only mode, at least if you
don't want to screw things up -- until now), so security is already gone.
Frankly, even in a single-boot scenerio, a bootable removable drive pretty
much removes all security these days, what with Knoppix and tomsrootboot
and all those sorts of things. ISTR there's a floppy disk image out there
that's made for changing the passwords in the WindowsXP registry. If you
need real security, you frankly have to keep the user physically away from
the computer, or use an encrypted filesystem that requires key entry on boot.
> Its almost like running windows 98.
I believe that was the primary #1 selling point of Windows XP. If it *weren't*
almost like using Win98, it would have received about the same user response
as Windows 2000, i.e., "make it go away". After Win2000 flopped in the market
place (not flopped compared to NT4, but flopped in terms of being received by
consumers as the next version after Win98), Microsoft backpedaled their
no-more-versions-of-Win9x position, did some market research to find out
what differences regular people saw between the two OSes, released WinMe
(to tide over the OEMs who were carping about their position trying to sell
either Win2000, which people wouldn't buy, or Win98SE, which was officially
obsolete), and poured the results of that market research into WinXP. So
yeah, it's almost like Windows 98. If you squint. At the time, security was
not a major concern for most users. (It has in the time since become a more
significant concern. Yes, end users can be slow catching on to things.)
Cut that out, or I will ship you to Norilsk in a box.
> Such a restriction would also block current, commerical rescue-offerings,
> at least, UNLESS they licensed the driver themselves, somehow.
They probably already do, but if not it would be trivial for Microsoft to
arrange agreements with each of them.
Cut that out, or I will ship you to Norilsk in a box.
> Or M$ just uses the next service pack/patch to revise the EULA.
That would still take 2-3 months. And that's *after* captive-ntfs gets on the
Microsoft radar, which is six months *after* it's widely deployed, which is
six months from today. So we're still talking fifteen months -- enough time,
in theory, for someone to code the desired capabilities into the native NTFS
driver. (One of the big distros should probably pay someone to do this, though,
as it hasn't happened up till now. Probably because most of the serious kernel
hackers haven't used Windows since roughly 3.11 for Workgroups, as they have
a tendency to live deep in server/developer space, where most of the "users"
are programmers of some kind and even the managers have heard of Unix.)
Cut that out, or I will ship you to Norilsk in a box.
Last I checked, ext3 only broke support when a program designed to manipulate ext2 wanted to write.