Slashdot Mirror
Windows Home Server Corrupts Files
crustymonkey points out a ComputerWorld article which says that "Microsoft Corp. has warned Windows Home Server users not to edit files stored on their backup systems with several of its programs, including Vista Photo Gallery and Office's OneNote and Outlook, as well as files generated by popular finance software such as Quicken and QuickBooks."
Crustymonkey asks Don't back up your files to Windows Home Server, as recommended by Microsoft themselves? I'm not exactly sure what the point is in having a home server if you can't back up files on it."
.... If there's a user friendly alternative to Windows server for Joe Enduser? I run a Debian box with Samba on a computer that does hardware RAID 1 for my file sharing needs (I also have an SCP turned on so I can shove files onto it from outside my network too). But that's not something that I can suggest to my friends and family. So what can I suggest to them that is as "user friendly" as Windows Server?
This is my opinion. To make sure you don't steal it, it's covered by the DMCA.
There's a difference between "backing up" and editting files on the server. So when one opines "I'm not exactly what the point is in having a home server if you can't back up files on it."... it shows they don't understand that difference.
Now an actual valid complaint would be what good is having a file server if you can't STORE files on it... but that's a world of difference between simply backing those files up.
I have a feeling this problem will be fixed in less time than it takes Apple to shut down a blogger.
Oh please. ``a, because a PST being written and read over a network is slower, and if the connection goes down, the file may be corrupted... just like working with any other file over a network.'' That might be the gospel according to Redmond, but for those of us outside the horrific networking decisions Microsoft have made, terabyte-class Oracle databases work just fine over NFS. Remote access via GigE to dedicated filers is faster than local spindles unless those spindles are in exotic raid arrays, and why would a `network' be any more likely to induce corruption than, oh, a fibre channel network?
From Microsoft's site:
When you use certain programs to edit files on a home computer that uses Windows Home Server, the files may become corrupted when you save them to the home server. Programs affected: Windows Vista Photo Gallery, Windows Live Photo Gallery, Microsoft Office OneNote 2007, Microsoft Office OneNote 2003, Microsoft Office Outlook 2007, Microsoft Money 2007, and SyncToy 2.0 Beta. Additionally, there have been customer reports of issues with Torrent applications, with Intuit Quicken, and with QuickBooks program files. Our support team is currently trying to reproduce these issues in our labs.Finally, they say:
This issue may occur because of a recently discovered problem with Windows Home Server shared folders and with certain programs.Fraggin' scary.
Don't tell me to get a life. I'm a gamer; I have LOTS of lives!
Quote from Microsoft's support article: "When you use certain programs to edit files on a home computer that uses Windows Home Server, the files may become corrupted when you save them to the home server."
A large amount of Microsoft's profit, in my opinion, comes from selling unfinished software, and then getting money for "upgrades". Microsoft won't get money for the fix to this problem, but I think you will agree that Microsoft is the largest supplier of unfinished software, and making the whole world a beta tester is cheaper than selling a finished product.
Therefore, MOD PARENT UP.
I notice that people are inventing nonsense about this; the problem appears not to have anything to do with editing backups.
UNC paths (\\computer\share\filepath) have worked for as long as LAN Manager networking has been built into Windows, and before that, as long as network redirectors were able to be loaded into DOS via TSR (Novell, Artisoft, etc.)
Apart from file dialogs, most of your argument is semantic. "Special" is your invented word for "not working the same as it does in UNIX". "Redirector" vs. "director", please. Look, it's doing the same thing as mount points and path resolution do in UNIX land.
I could argue that, if anything, it's confusing in UNIX land to see all devices (local and remote) rooted in the same hierarchy, because it breaks the physical local folders metaphor. At least in Windows (excluding DFS trickery), with a UNC path you immediately know the machine at which a path resolves and don't have to consult mount.
And where in any X11-based window manager is there a file dialog that lets me mount remote filesystems from within it (equivalent to Windows' "Map Network Drive"). There's not. Why? Because in UNIX, network files are "special" too.
There are no karma whores, only moderation johns
That statistic is not the least bit true.
Are you one of those nuts that thinks guns have a mind of their own, and "could go off at any time"?
Aww, no fair comparing windows networking to DECNet and VMS. The first system to get file access right and consistent across disks/tapes/clusters/networking/carrier pigeon. Well, other than the 9-character limit in the filename part itself. And being different than any of the other slash-based filesystems around there were a lot of cross-platform tools that didn't cross over to VMS well, but I digress. With the distributed lock manager built into Files-11, everything was pretty transparent.
NODE"user pass"::device:[dir.subdir]filename.type;ver
I remember the first time I was working in Houston on a VMS node on THEnet and specified a file on a server in Austin.
At least with Linux tools now we have vfs implementations that hide most of it. I'm not sure Windows-based systems have even gotten that far.
heh - old DECnet (note the spelling) memories.
ever play with 'hidden areas'? at DEC we had so many nodes in the EASYnet that we had to use hidden areas (EASYnet was the name for our corp. network inside DEC):
node1::node2::NODE"user pass"::device:[dir.subdir]filename.type;ver
a form of explicit routing. once you are 'at' a given node, it uses its local area.node (I can't believe I remember this stuff) lookup tables (or routing) to then figure which next IP^H^HDECnet node to send the packet to.
--
"It is now safe to switch off your computer."
Nothing for 6-digit uids?
You can backup files to the system PERFECTLY FINE. It's just that once they are backed up, there is some problem editing the files on the server. While that's obviously a problem, why are people needing to edit their backups? I thought the point of a backup was just to be a copy of whatever is on your system. If you edit the file directly on your server you are leaving your PC with an older version - that doesn't make any sense.
If I wake up at 3 AM because someone is rummaging around my house I really don't think it's out of line to assume they're planning to do something bad to my person or effects...
Playing twenty questions in a situation like that often results in them giving a 21 gun salute for you...
Browsing at +1 - no ACs, I ignore their posts. So refreshing!
I used incorrect wording - the rest of my post was in 3rd person, and this should have been as well. I do not advocate that activity; I was pointing out that the statistics in question don't take such situations into account. Were I personally to be in a situation with an intruder in my house, I would hope my reaction would be to assess the threat and act appropriately, which could include shooting to kill. I really wouldn't thingk I would actually hold anyone at gunpoint; that's just stupid.
Now, onto your comments:
"It is likely to both be the case that people predisposed to violence are more likely to own guns, and that owning a gun will make people more likely to be violent than if they didn't own a gun."
Where on earth did you get that statistic? I'm also confused where you say that the situation is more nuanced than "owning guns causes people to be violent" or "being violent causes people to own guns", but then simply restate the positions with more words. "Nuanced" doesn't mean "couched in pseudo-psychological terminolgy."
And this:
"And the statistics don't show all the incidents where gun owners just didn't get a chance to use their gun in any way, in which case, all that the gun did was make them more likely to get shot."
Again, how is it that mere possesion of a firearm makes it more likely to be shot? Not from a statistical point of view, but a behavioral view. Statistically, owners of cars are more likely to be in car accidents; but that doesn't mean I should sell my car if I am a good driver and take proper precautions.
And finally:
"Don't overestimate the usefulness of a gun in defending your home. Real-life burglars have many advantages over the gun-owner, which include (a) figuring out when you're not likely to be home anyway, (b) being able to pick when and where to strike, (c) surprise." combined with "Don't spend money on a fantasy of an intruder breaking into your home and you handling it by the book with a gun. Spend the money on better security measures than guns; e.g., good alarm systems, good locks, metal bars over the windows, etc."
You are trying to have it both ways - the sophisticated burglar you posit will easily defeat those passive protections you advocate. And you misspelled "homeowner" as "gun-owner" - everything you say applies equally to someone who isn't armed. You seem to be saying that it is better to offer no resistance once one's house has been invaded.
I do not advocate everyone owning or using guns - there are plenty of people in the world who shouldn't deal with fireplaces, much less firearms. But I'm not going to tell them they CAN'T have one, and I have a problem with people telling me I can't have one "for my own protection"
"As God is my witness, I thought turkeys could fly." A. Carlson