WinXP question: Is there any reason I should NOT upgrade from FAT32 to NTFS?

I got my new computer last week – P4/2.8ghz, 1gb RAM, 128mb ATI video card, Audigy 2 s/card, CD/CDR/DVD drive, DVD-RW drive, 250mb zip drive, 160gb h/d – and I’ve been abusing the hell out of the thing, with multi-tasking programs and LOTS of editing/copying/moving/burning/deleting files. So far it’s taken everything I’ve thrown at it :cool: but by this evening, the HD was screeeeeaming at me (not a “bad” scream, just a persistent “Hey! DEFRAG ME you S.O.B.” complaint.) I’m not sure how much of the drive is fragmented because my defrag program is a piece of Shi’ite that doesn’t give any kind of percentages (grr…) but it was estimating a 2+ hour defrag time before I aborted it.

The hard drive is a full 160gb partition formatted in FAT32 – I’m leery of upgrading to NTFS, because you “can’t go back” as they say, and I want to leave the option of possibly creating new partitions for FAT32 (and possibly even FAT) to try and run old Windows and even DOS configurations. (I already have a boot manager pre-loaded, though it’s not configured for anything but WinXP right now.)

Other than that…how much does NTFS rock? Does it take longer to defrag under heavy use? How well does it handle file management, etc.? Is there any reason I should NOT upgrade to NTFS this very second?

Overall my experience has been that NTFS does require somewhat less defragging, and has been reasonably robust so far under XP. The only possible negative I can think of is if your XP PC crashes and you need to pull data from the NTFS formatted harddrive hard drive. You will only be able to do so on another NTFS machine . FAT32 formatted machines will not read NTFS partitions.



This note is also helpful

that should read "FAT32 formatted machines with 98 and 95 OSes will not read NTFS partitions. "

Seems to me the FAT vs NTFS thing is pretty well over.

7 years ago, when NTFS machines were hard to find except in coprorate settings, it was important that a consumer be able to boot to DOS to recover files from a trashed Windows installation. Now there are decent tools recover tools for that. Odds are that your friend’s PC is NTFS-capable, if not NTFS formatted, so you can always pull the HD and take it over there to read files from it.

There are plenty of websites that weightily discuss the pros and cons, but many of those discussions haven’t been updated in 4 years and are talking about a different world than the one we currently live in.

FAT32 was, and still is, a kludge stopgap to let HDs break FAT16’s 2GB barrier. We passed that point a very long time ago. Most consumers will never know, much less understand, the significance of journalling, indexing, and a ton of other internals that mean that NTFS volumes are practically speaking,all but incorruptable.

Bottom line: Use NTFS, period. But if you are going to convert, first do general housecleaning to remove junk files (temps, IE history, etc), and uninstall any programs you were going to get rid of anyway. And 2) make sure you’ve got at least 20% free space before trying to convert. The books say 10%, but that’s one process you don’t want to malfunction midstream.

FYI, XP’s built in defrag tool does a damn good job on NTFS volumes. I’ve never used it on a FAT, so I can’t speak to that.

In our corporate setting we defrag servers every few weeks and desktops never. They go in the trash before their hard drives really need a defrag. If you’re doing a boatload of big install / uninstall, that’ll drive a greater need for defrag, but by and large the need for frequent defragging is a wive’s tale left over from DOS 4.11.

Not much to add except I build an almost identical computer about 6 months ago. For some unknown reason I was hesitant to switch to NTFS. All I can say is, switch, switch like the wind!

I use NTFS and FAT removable disks and get files called “EA DATA. SF” that can’t be erased on some of the disks. This is an “extended attribute system file” and I guess it carries information used by NTFS but not by FAT. While this hasn’t caused anything else to malfunction that I know of, it bothers me that there’s a file there I can’t control. For example, someone gives me a floppy and asks for a copy of something I’m working on, wanting just that one file, and I wind up writing two files - “What are you giving me this other one for?” “I don’t know and I don’t know how to write a floppy without also putting it there”. When this started happening I spent probably two hours studying it and trying to delete it and trying to read a file with two spaces in its name on DOS systems and checking for viruses and so forth.

I guess normal people don’t mind this, but somehow it still irks me. Computers already do too much I dislike and don’t know how to stop.

Okay, umm…what the heck is NTFS5? I’ve never heard of that before.

I’m well over 18gb in size so that’s a no-brainer – plus with the max file size of 4gb for FAT32 (hmm, never knew that before) it’s pretty much a done deal, since I’m gonna be toying around with files that size before long.

Next question: by “one way trip” does that mean I can never re-format any part of this drive to FAT/32? I still want to leave that option open as a separate partition for a dual-boot (or tri-boot) O/S.

Automatic NTFS5 Conversions: What You Need to Know

Mac user here, so sprinkle this opinion with grains of salt to suit your taste.

If I had a PC, I think I’d have two physical drives, the primary boot drive formatted as NTFS and the secondary drive, where I’d save most of my documents, FAT32. I’d install a lean 'n mean version of the OS on that one to have the option of booting from it in cases where the primary wasn’t feeling very bootable, to do repairs and whatnot, but otherwise programs and OS would go on the NTFS and document files on the FAT32.

Then if the whole system got badly hosed by viruses I could snag my documents by mounting the FAT32 volume on a non-Windows computer that would be immune to Windows-based viruses.

(Would also give me more options for recovery in case of directory-level corruption, more scavenging utilities available under more operating systems).

Actually this isn’t too different from what I’ve done.

I had a very expensive, albeit old soundcard I used for my studio computer, but it was no longer supported. I had just built a state of the art XP pro box, so what was I to do? I stuck my old 98 C: drive in, then installed XP as NTFS on a blank D: drive. The NTFS can read the 98 drive as well, so I can shuttle files one way.

No longer necessary, now that NTFS is fully readable by Linux (including MandrakeMove and other CD bootable versions)

NTFS-5 is simply the latest version.

[pointless history trivia mode]Windows NT introduced the original NTFS (hence the catchy name), and when Win2K came out they made some small internal changes. A PC running original WinNT can’t read a NTFS-5 volume, but a Win2K or later machine can read an original NTFS-4 volume. They added the ability to read NTFS-5 into WinNT with, IIRC, SP4. NT machines still running SP3 or earlier are very rare anymore, so there’s no difference practical between NTFS4 and -5[/pointless history trivia mode]

Your question doesn’t exactly make sense. You seem to be confusing “formatting” with “partitioning”. A single physical disjk drive can be partitoned into several different logical drives( ie. C: D: E: F: , etc). Or the physical hard disk can be left as one big partition. You can also leave unpartitioned space on the drive which is unused now but which can be converted into a partition later.

Once you have some partitions, then you can format it/them. Each individual partition can be formatted differently from the others, but each individual partiton can be formatted only one way: FAT, NTFS, Mac, Unix, whatever. When you boot an OS, it will be able to read and display those partitions it understands, and the others will be completely invisible.

In general, partitions cannot be resized without deleting 100% of the data on them. There are programs that claim to be able to do that, and they work most of the time. But when they fail, 100% of your files are gone and you get to reinstall the OS and all the programs. Such fun.

Also in general, once a partiton is formatted, that format type is fixed forever. To change it, you reformat the partiton in some other style, incidentally erasing 100% of the data on it. As an exception to that rule, Windows includes a feature to reformat a FAT32 partiton as NTFS in place without damaging the data. That’s cool, but it’s also a one-way trip. Having converted the partiton from FAT32 to NTFS, there is no way to convert it back to FAT32. Not that you’d want to.

Formatting a particular partiton to NTFS does NOT limit what you can do to any other partitons you may have. They can be formatted, re-formatted, or deleted entirely. With 100% of the data on them being lost of course.

So in answer to your specific question, BEFORE you install the first OS, you can decide how to divide up the hard drive. Your, say, 80GB hard rive could be set up with partition #1 of size 3GB for booting Windows, partition #2 of size 50GB for storing Windows files, partion #3 of size 15GB for installing Linux & its file storage, and leaving the remaining 12GB unpartitoned.

When you start to install the first OS, it’ll ask you whether and how to partition the drive. Assuming you’r doing Windows first, you’d tell it you want the 1st 3GB partitoned as C:, the 2nd 50GB partitioned as D:, and leave the rest unpartitioned. For each of C & D you’d then get to pick FAT32 or NTFS. Windows would then go through the whole install process and you’d find you’ve apparently got two hard drives, a 3GB C:, and a 50GB D:.

Now you can go back and install, say, Linux. That can be 5 minutes, or 5 months, later. It’s install program will detect the unused 27GB and ask you how to partition it. When you’re done, Linux will read its partiton and Wiindows will read its partitons and neither will know the other exists. (As others have pointed out, some flavors of unix can read NTFS and will see the Windows partitions if you want.)

Again, you can’t change the size of a partition with data in it, at least not without special utility programs and then not with enough safety for my taste.

Hope that’s clear and simple enough. (For the pros, yes, I know about NTFS dynamic volumes, but let’s keep this at the home user level)