Hi... This is my first post here, though I've been fortunate enough in the past to correspond with techAdmin on another forum.

I'm in the process of moving from a Windows 98 system to a new computer with Windows XP. I've just learned a bit about XP's default pathname for the "My Documents" folder that raises some questions about how XP will behave with a third party file manager. (In particular, I use Total Commander (formerly Windows Commander), and can't imagine using Windows Explorer and a mouse to manage files).

I know that a little knowledge is a dangerous thing, and I haven't used XP... but here is what I've been told and read. First, please advise what's correct and what's not correct...

a) By default, on XP, the Target or actual location of the My Documents folder is C:\Documents and Settings\user name\My Documents

b) I've also been told that in Windows Explorer, what you see is C:\My Documents, and that if you Drag and Drop your files to the apparent My Documents, they get copied to the right place.

c) I understand further that you can change the Target location by right-clicking My Documents, clicking the target tab, and specifying a new folder.

If the above is not correct, some of my questions might not apply....

How does a third party file manager see this directory?
Assuming (b) above is correct, will the file manager see the "real" pathname, C:\Documents and Settings\user name\My Documents, rather than the "aliased" C:\My Documents? I don't know whether you can say "DOS level file manager" with XP, but I'm talking about real pathnames.

Dragging and dropping in Windows Explorer vrs copying with a file manager?
Also, does XP treat files dragged and dropped into this "My Documents" directory, via Explorer, differently from files copied into it via third-party software?

I'm thinking here about the way "C:\windows\fonts" behaves on 9x. It's kind of a virtual directory situation... you get a very different file view of the Fonts directory in Windows Explorer than in a dir view... and (and this is very important), if you copy a font into C:\windows\fonts using TC, it does not install, whereas if you drag and drop it using Explorer, it does.

Pathnames and migration of data...
I'm about to migrate to the new machine and had planned to have the system builder to copy over all of my old data to, say, a D partition. Then, using Total Commander, I was going to copy over the My Documents folder on D to C:\My Documents. I'd hope to preserve my current data file structure (as well as all my shortcuts, including those configured on my Directory Hotlist, on my toolbar, etc) on C, and also keep a backup on D until I had everything working.

Preserving the data file structure is particularly important, as otherwise I'll be spending a month just typing in new pathnames on shortcuts, in miscellaneous program settings, etc.

I'm also concerned with the length of pathnames. When I've gone several directories deep in 9x, I'd needed to shorten some filenames for Total Commander to work. I don't know whether this was a Windows or a TC limitation... and I don't know whether this might give me problems... perhaps make some current data innaccessible, when I move it to "C:\Documents and Settings\user name\My Documents" in XP.

Reluctant to mess with defaults...
I should add that I'm totally new to XP and shy enough about the Registry in general that I don't want to try any directory hacks, etc. Also, my guess is that if I used right click to change the Target location for My Documents to be actually "C:\My Documents", there'd be a conflict of sorts. Thoughts on this?

Would appreciate comments as soon as possible, as I'd been scheduled to migrate data today and this has come as a big surprise...
Back to top

Keep this in mind, windows NT is a multi user system, translating pathname to dos compatible formats isn't a good idea, it's a big pain. Windows 98 was closer to DOS, some say very close, but NT is not. All DOS runs only in emulation mode, and it doesn't run very well to put it nicely.

I checked out that app, they claim to support XP/NT so there should be little problems I think, but don't quote me on that, the only way to find out is to try, give it a whirl, if you like that navigation system see if it works, no other way really to know. There's nothing fundamentally different about how Windows 98 and NT do file naming to be honest, except NT systems tend to have very long file names.

I always set up my NT installs the same way.

Partition 1, usually c:\ is for windows and programs. All data lives in other partitions, I use a lot, but let's keep it to 1. All data goes to your second partition, d:\

create a folder on the d drive called documents, or my_docs, whatever. Inside that folder create a user folder, say 'rc'. Name is irrelevant, keep it short and you might be able to use your dos tool.

So let's say you keep it all very short:
d:\docs\rc
will be your target folder for 'my documents'
Right click on 'my documents', change the target path to 'd:\docs\rc
Now all your data for 'my documents' will go to that folder. My Documents is just an alias. It doesn't exist, whatever path you've created is where the files go.

For Windows Explorer, which is an excellent and for some reason very underrated file navigator, I have to give MS their due on that one, it was good enough for Mac to copy after all, you can create various opening options through command line switches.

Right click on 'Windows Explorer', I always make a shortcut in the taskbar to that, in XP you have to first unlock the taskbar to make shortcuts there, then this would make it open always in d:\
%SystemRoot%\explorer.exe /e, /n, d:\
Use exactly that syntax.

The reason you see differences when you drag and drop fonts is that the fonts need to get installed, dragging a font into that folder triggers the install process, there may or may not be a command line way to do that, but I've never done it, no reason to.

I can't say how the file manager would deal with these issues, my guess is not very well.

No reason to do anything with the registry unless you really feel like hosing your system and reinstalling the OS at worst, creating hard to fix problems at best.

There is no 'my documents' folder, it's just an alias, doing that would not achieve anything unless you had created an actual physical folder called c:\my documents , which would be silly.

By the way, the default xp skin is horrible, the first thing I do to any XP box is turn all that garbage off, and return it to as close to classic windows style as I can, takes a while, but it can be done, except for a few extremely annoying things like the stupid 'search' feature, which as far as I know can't be returned to normal. Did I mention how much I hate Windows XP? Windows 2000 was very good, but in a way I'm grateful to MS for creating the XP monstrosity, it motivated me to start moving to Linux, since I know I'll never use XP, or any other windows os that will be released in the future except for testing purposes.
Back to top

jeffd - Thanks for your thoughtful and clear reply. It leads to some more questions, of course. If this were DOS, I think it would all be settled.

OK... the XP "My Documents" is an alias... and NT/XP are multi-user systems. I think I understand.

My clone-builder was pushing me to keep my data on C, whereas it made much more sense to me to keep it on D as you suggest. (He's a good tech, but used to dealing with companies, so there's a common denominator factor he's learned to take into account... and his English isn't good enough to discuss the nuances).

So, I'll create the target folder on D...
d:\mydocs

I'll be the only user, so I don't need a user folder. I'm assuming it's possible to avoid the user specification when installing XP. Is it?

In Windows Explorer, I assume this folder will show up as "My Documents." I assume it will not also show up as d:\mydocs.

If I understand you, I assume that, in the file manager, Total Commander, it will show up as d:\mydocs.

What about pathname specs in programs, ini files, etc?
But, what happens with pathnames that programs request, either upon install, or in their configuration menus, or in their ini files?

I have a bunch of ini files that I've been told to pre-edit to create settings so the fresh XP installs are up and running (eg, programs like PocoMail, NoteTab Pro, WS_FTP, etc)? Do I use the "real" pathname or the alias?... ie, should I edit the ASCII ini files so the pathname to the data is...
d:\mydocs\app_data

...or keep them as they are now...
c:\My Documents\app_data

(or would it be on d: ?)

What about more automated consumer level programs like, say, the new Quicken? I haven't seen the new version I'm going to install, but the last one had fields for typing in backup paths, etc. Do these want My Documents or the "real" pathname, or is it assumed my data will be in that main folder and thus that spec is superfluous?

If the latter, what happens if I put data elsewhere, outside of "My Documents"... say in d:\otherdata?

Also, what happens to the browse function in various programs I have an in download dialogue boxes? Is there any consistency?



I understand, and mentioned it only because the friend who first told me about the virtual My Documents warned me that I'd better stick to drag and drop for copying files into it... and I was worried that there might be some sort of weird install-like process with data files so the system could find them. At that time, I hadn't a clue what was going on, and he really couldn't explain in detail.

Re Windows Explorer, I use it occasionally, but frankly I hate drag and drop; and it's incredibly limited in so many ways it would take a month to describe. Take a look at Total Commander and you'll see what I'm talking about. It's really a joy to use. Interface derived from the old Norton Commander for DOS and refined from there, and I've been using versions of it for something like the past 20 years.



Can you clarify this just a bit... I can't quite follow.

I'd love to see something replace Windows. Am absolutely dreading this move, but for the time being there's a lot of Windows software I need to be able to run.

One more question... re the c:\ partition 1, for Windows and programs... How large would you make this partition?

I'll be using a 120-gig disk. Right now, on Windows 98, the OS, programs, and all of my data occupy only 6-gigs. Obviously, my apps and data will change with a newer, faster system, but any sort of rough guideline will be helpful.

Thank you.
Back to top

Disk Partitioning
Lazy, stupid, what can you say? When your os fails, if it fails, often even if your harddisk c partition is wiped out, say by the latest windows virus, there is your data safe and sound on its own partition, every one who knows what they are doing does that, which means anyone familiar with *nix type OS's, but even good windows admins do that.

No excuse to do otherwise, only Windows users use 120 gigabytes of space in one partition, os, software, email data, all other data, that's just plain dumb, nobody in the unix / linux world would ever do that. But sounds like you got your d: going.

Size: 10 gigabytes for c:, 12 Gigabytes if you want to be completely sure you never need to resize again, after about 3-4 years I'm using 6 gigabytes out of 11 available in my os/programs partition, and that's a lot of software. Remember also, you don't need to install programs in the c: if you don't want to, you can install them whereever you want, some network admins install just the OS in c:, programs in d:, and data in e:, for example, but I've never found a need to do that, though there is an argument that can be made for that.

Programs like Fireworks and Photoshop actually expect a dedicated swap partition, I give them the first 2 gigabytes of my second harddrive for that, it's better to put the swap partition close to the beginning of the harddrive, it reads and writes faster, that's also by the way why it's better to put your os in the beginning of a large drive, the data last, the data gets accessed much less than programs and swap files and os stuff.

If you put all your data, including emails, into d:, that's more than enough. I used to recommend 6 gigabytes, but after having to repartition my current harddrive a few times to expand my primary os partition I now recommend 10, that's very safe as long as no data is stored on it.

It's also much easier for an OS to deal with itself on a smaller partition, easier to defrag etc. Plus windows likes having its swap file on the same drive as the os, linux of course always has it on a dedicated partition, for what should be obvious reasons to anyone who understands how this stuff really works.

Multi User Systems
A multiuser system always has 1 user, administrator. If you always work in admin mode, that's 1 user. If you work in power user mode, which is slightly safer, but not much, and a pain to do on windows, that's a second user. There is also a built in 'guest' account, that has very few privileges. That's at least 2, possibly 3 accounts. So the answer is, create: d:\docs\rc and assume one day you'll be adding more users.

Also, it's a good idea to use a guest account with restricted permissions if anyone else is ever going to use the system, people can do amazing damage with the most innocent and ignorant intensions.

Paths in Windows File system

You assume wrong, it will show up as both, my documents on top, and the actual folder in the d: directory.

Yes, it's a real folder, so it will show up in your filecommander as such.

Use the real path, it's easier, less typing, and guaranteed to work every time.

Use the real path, it makes no difference at all, less typing, again, and will always work, since you're sending it to a physical path.

I never use the 'my documents' path for anything, all paths I assign I assign explicitly, it's easier that way. Plus spaces are a pain to make work in many applications, especially for web stuff, so I don't use them in file or folder names, that means they always work everywhere for all OS's.

Nothing happens, it goes where you put it, it doesn't matter, it's a directory on your hard disk, the data is where you put it, it doesn't matter, unless you want to access all your data through 'my documents'. My data is all explicitly assigned, to different partitions for different purposes, a web one, a business/personal one, a audio/video one, etc. I don't think I've even used my documents on my main box for a few years. Although I do access the stuff in there.

Yes, it's completely consistent from what I've seen, you get the explorer navigation in the file open and save dialogue boxes, those usually have my documents + of course all your real folders, makes no difference which you use.

Get better friends. Just kidding, that's silly, makes no difference except for a very few cases, such as the font one, where a programmed behavior is triggered by the drag and drop operation, that's one of the only cases I can think of where that would apply though. With one caveat, however, in windows, when you drag and drop a folder/file from one partition to another, it only copies it, leaves the original. I don't know if Windows has an equivalent to the *nix 'mv' type command, that physically moves the file/folder n matter where it is.

Windows Explorer etc

Can't argue with that, it shouldn't be that different, windows 98 used the same explorer logic as xp/2000 do, same type of path naming conventions, very little difference actually.

If you make that the target your windows explorer will automatically open on your d drive, there's no real reason for it to open on the c: drive, which is the default, or to my documents, I can't remember which it defaults to, but opening to your data drive makes sense, no?

Windows Replacement? No way

Hopefully nothing will replace windows, ideally windows will just go on being windows, and the rest of world will move on to a real os, like Linux, leaving windows for those who want it, choice is good, I don't want linux to be like windows, although it would be nice to see some of the rough spots in the windows managers like kde and gnome, and some of the bugs in the gui software apps, get worked out, but given the desktop marketshare linux currently has, it's totally amazing how good it really is already as a desktop, it's server performance of course needs no comment, it's taking over very quickly, about as fast as MS thought their junk would take over the server world, something about their exhorbitant licensing schemes just seems to turn network admins and IT decision makers off, that plus the insecurity...
Back to top

jeffd

Your answers are so helpful I'm going to ask some more questions. I've got it about pathnames. Would appreciate your thoughts first about partitioning and swap files, and then about multi-user setup. Here's the overview....

Disk/System Setup
I'm going to have two 120-gig hard drives, plus two optical drives, and one-gig of RAM... with a 3.0-ghz P4.

I've long thought the same thing that you just suggested, that defragging is easier on small partitions.

At the same time, I felt that too many partitions would get messy... that sans-serif i's, j's, and l's are hard to read (said only slightly tongue in cheek)... and that I think I'd run out of room on my two-panel file manager task bar (as currently set up) after about 8 or 9 drives on each. Also, most Windows users I spoke with discouraged smaller partitions.

My original plans were to have 3 partitions on each drive, with drive-2 for backup only (not as a concurrent redundant drive, but as a backup to protect saved data from a disk crash, virus attack, or whatever). I'd then periodically burn CD-Rs of data and DVDs of the C system/programs partition and keep them in a separate location.

One of the things I'd want to be able to do with drive-2, in case of failure of drive 2 during a project, is to swap it in instead of drive-1.

I should add that I'm not a gamer and not a power user... and not yet into big media files.

Another consideration is that, during the installation/migration process, I need to use one of the partitions, probably on the second drive, to hold a copy of my current C drive, not to run any programs, but just to have access to various setting files, legacy data, etc, that frankly are too much for me to think about all at once.

There's also a peculiar consideration... I have a bunch of old DOS data and software (that probably won't run under XP, but you never know) on my E partition of my current setup... and for reasons too numerous to mention, I'm stuck with some "E" defaults on some of the software. What that means is that the E partition is a Windows partition but I'll be holding my DOS data and software on it... may change this later. I have it now at 2-gigs, just so it doesn't get lost, though there's only roughly 200mb of data on the partition. So, whatever I do, I need a small E partition.

I don't want to think about separating the OS and the program files right now.

Partitioning disk 1
I could just go with C = 12-gigs; D = 106-gigs; and E = 2-gigs, or I could add another partition after E.

I should mention that I've gotten both Norton Ghost and PartitionMagic 2005 at good upgrade prices, but I've been advised that PartitionMagic can cause problems and that I shouldn't use it to resize partitions.

Swap files
If I understand what you're saying, the Windows swap file should be at the end of C... but that a dedicated Photoshop etc swap partition should be at the beginning of my second disk.

I'm right now using Photoshop 5.5. Someday, when I recover from the cost of this new machine and software, maybe I can upgrade to a newer version. I'm not sure if 5.5 is that demanding, but I should allow for the upgrade.

Partitioning disk 2
This suggests, if I read you right, that I include a partition at the head of disk-2 as a swap partition for Photoshop... a second partition to mirror the operating system and program files on C using Ghost... and a third for backing up data files on D (holding a backup of E in a separate folder on this partition).

Thoughts... comments?

I remember reading somewhere that I've got to set up .net somethings to be able to use Ghost, but I'm not sure. I'm trying to set up the system so I can get it running.

Multi User Systems
I'm running out of steam here, but I want to touch on the whole question of setting this up. I haven't a clue about networks and permissions, and had sort of hoped to defer it for a while, but it sounds like I've got to make some decisions to get started. I can't imagine I'll be ready for power user mode for a while... and, per note below, it would make using Total Commander a little more difficult. I can see the virtue of the guest account.

Sounds like 3 accounts might eventually might make sense, but I'd like to set up what I need to now just to get myself going and allow later flexibility.

Is there any difference in functionality between d:\docs\rc and d:\rcdocs, if you follow? Second account would then be either d:\docs\sa or d:\sadocs.

Incidentally, what's the install order to have a user-dedicated "My Documents" to right-click on to then specifiy d:\docs\rc? By this, I mean it sounds like I've got to define myself as an account and set up my default "My Documents" before I change its location. Is this correct?

Re my Total Commander questions, I just got this response...



I know zilch about networks and permission.. thus far have no local network... though I'm going to have to set up a LAN when I set up broadband and a router, which will be very soon. Can you point me to an introductory info source?

Enough for one night... Your further comments welcome. Thanks.
Back to top

1. Have enough partitions but not too many.
drive 1: c: d: e: as you suggest. You have no control over the drive letters assigned when installing windows, it's automatic, you can change all the ones except c: through the NT/XP administrative tools interface.

If you want to boot off the second drive as a ghosted image of drive one it has to be a ghosted bootable image of drive 1, you can't have it both ways, either it's a ghosted image or it's not a ghosted image.

There are ways around this but it's an expert level thing. Unless Ghost has gotten a lot easier to use, but if your primary os parititon is on the c drive, and you ghost it to the x: partition, which is not the first partition in the case you described, that first being the swap for photoshop, of drive of drive 2, you're going to have problems, the system won't boot if you try booting second drive sometime.

Many of the questions you're asking don't have fixed answers, you just have to play with the stuff, that's how we all learn. the stuff your friend said about partition magic is nonsense, and simply reflects some sloppy practice on their part. XP requires PM 7. Rule number one however is treat the partitioning process with respect, do not carry out too many operations at once. that means execute the partitioning commands one or two at a time, it takes longer, but it's much more stable. Also of course when PM asks if you want to make emergency recovery disks, make them.

External backup drives are the way to go, I use firewire. Internal backup drives are next to useless.

I used to try internal drives for backup until I realized how silly the idea is and was.

For maximum stability and system recovery the 2 120 gig harddrives should have been installed as Raid 1, mirrored drives. Requires completely identical drives, same manufactorer, model number, and ideally manufacturing date. Any good raid 1 system supports drive failure and automatic rebuilding of drives when the dead drive is replaced.

However I assume you'r not going there.

Easiest by far method? Create firewire drive, with same partitions more or less as drive 1. To backup, drag and drop all contents of each partition into backup drive partition.

To use as rebootable drive, simply install the mbr in the backup drive should you ever need it.

God, I hope not, doubt it though, the old Ghost runs in dos mode, that's how it works on live partitions.

No, it's just a logic thing, I want all my_docs folders to be in the my_docs folder so I know where to find them. Pathnames are like notes to yourself that should work long after you remember why you made it, so in this case I look at the root directory of the partition, see 'my_docs', and think, hey, that's the my docs folder. Then all my user folders are in there. No brainwork required.

It's irrelevant, the default my docs is created when you first loginto your account, when you change the path to the new one it asks if you want to move all the contents of the folder to the new path folder, say yes. Only requirement is that you actually made the new folder before telling mydocs to go there.

They are correct, only admins can see all the user stuff, users can only see some of their stuff, that's standard non-admin priviliged users.

By the way, windows permissions are the main cause of viri/malware, almost all default windows install put the main user with full admin rights, which gives all malware full admin rights. But to run in power user mode means you have to log out, log back in as admin, do your things like av updates, windows update, etc, then log out, log back in as power user name, having lost all your work in the process each time you log out. This is retarded of course, Windows xp slightly helps this process by their user switching mode, but that itself creates it own problems, it's buggy, doesn't work very well.

Linux, and all unixes, which were built from the ground up with security in mind, of course don't have this limitation, you always work in protected user mode, then when your try to do something that requires admin priviliges an alert pops up asking for your admin password, then lets you do that particular process. Or you can do it through the su command line option, all while the standard desktop stays put underneath the processes you run. Windows will never be secure until windows is rewritten to be much more like unix, but that will never happen, Bill Gates would then have to admit that unix is s a superior system, especially for servers, which he will never do.
Back to top

jeffd,

My ongoing thanks.... I'll try to pick up with some additional questions in the next day or so, but I want to let you know before then how much I appreciate your input.

Software install put off until Monday.



This makes a huge amount of sense, and I coincidentally just saw today a graphic example of what can happen when malware has admin rights. Not pretty. What more can you tell me about the user switching mode? Power user mode doesn't sound practical.

Can the security setup be changed after a default install? I don't want to have too much of a learning curve right off, as I'll have updates of almost all my large software packages plus a new OS, and I need to be up and running right away.
Back to top

Power user is ok, except some applications don't let you do things in it, like Norton Antivirus auto update, or I think Quicken, a few others.

Windows XP has two modes of logon/off. One is the older, Windows 2000 based on, real log off and on. That deletes all your programs and files open. Second is an XP specific one, a fairly feeble attempt to get around the problems of switching users, called fast user switching. That retains the desktop state in memory while you are logged out. It has a lot of problems, tends to clog up the system more, some apps simply don't support it, varies app to app.

Of course once you enter into the wonderful world of service pack 2 you're on your own, see this forum for links to some of the problems you may encounter.

Windows default 'security' is no security at all. If your system was installed with fat32 instead of NTFS filesystem you have no security, forgot to mention that. Hopefully your friends didn't tell you not to use ntfs.

To make Windows even moderately 'secure' you need the following:
a real firewall, ideally router.
a real software firewall, not the XP junk, turn that off, install for example zonealarm.

All file permissions need to be reset from the defaults, too big a topic for this thread, but Windows native permissions have exactly the worst possible security choices made. That's to make windows 'user friendly'. Default windows installations are defaulted to fully insecure. XP service pack 2 gives the illusion of fixing that, but don't be deceived. Read the threads on xp sp2 'security' here.
This is however about as good as Windows can do with their design limitations, it's better than nothing.
Back to top

jeffd - Thanks for the SP2 comments. Once I've explored the recommendations of The Register article, I may start a thread on the whole subject of security settings. I'm almost tempted to start one now for my next questions, but let's stick with this one for the next round.



I don't know how real my $80 or whatever LinkSys router is, but that's what I'll be using. Can't afford one of the real real ones.

I've previously gotten a least one recommendation that I shouldn't use either the XP firewall (or the Norton Personal Firewall, which comes with the Norton package I have), but instead rely on the router in conjunction with Spybot, Adaware, and Spywareblaster.

Would you comment on ZoneAlarm in addition to, or instead of, these (including the Norton firewall).

In the past, I'd read about all sorts of conflict problems people were having with ZoneAlarm. I gather various firewalls, routers, and communications software can interact.
Back to top

Your router is a real firewall, a hardware firewall, but I tend to also use something like zonealarm so I really know what's going on in the local machine. For example, I don't like giving MS word access to the web, it's risky, and I can block that with a real firewall. XP does not have outbound packet blocking, which means it's not a real firewall, I don't know why they bothered releasing it. Well, I do know, they wanted to 'improve security' but they didn't want to actually make the system secure, because a secure system requires more care than joe user can give, and joe user is the Microsoft bread and butter, both home and corporate.

I'd second jeffd here, use a router, a software firewall, then all that other stuff, spybot, adaware etc, and it will be better than it would have been without all that.

I've used zonealarm for years, no problems.
Back to top