I'm not sure. Yes, you can change it through disk management in the mmc, but I don't know how it will affect your games, registry entries, etc

Chrispy said:
Ugh that's so stupid.


Okay, so, can I change the drive letter with no problems?

you should be able to if you have no floppy. I haven't had this problem recently because I always make sure I'm installing to C: even though i don't have a floppy.

If any programs you use rely on a drive letter being in the path you may have to simply look for the EXE file in the directory where the program currently reside when you launch whatever program.

Worst case scenario is a full reinstall which no one is thrilled about doing.

here's a guide on changing drive letters in Windows 7

http://...about.com/.../change-drive-letters-windows-7.htm

It worked! I am currently installing Diablo 3 into my secondary drive, labeled P:\. Thanks!

Yeah. That's a legacy problem because even though floppies are a dead technology, the industry hasn't decided to let them die yet. You will probably be able to install to a HDD labled A:\ in 5-10 years. Think of if like VHS. VHS died years before they stopped marketing combo players/recorders because enough people still had/used them.

Chrispy said:
It worked! I am currently installing Diablo 3 into my secondary drive, labeled P:\. Thanks!

No Problem! That was literally the most wild guess ever, and I wasn't even sure it would have made a difference!

Sergio probably would have said:
huuhhuhuuhuhuhuh the P drive.

NCC-1701 said:
You will probably be able to install to a HDD labled A:\ in 5-10 years.

It would be quite sad if we're referring to drives by letters 5-10 years from now, wouldn't it?

The problem isn't that "A: is reserved for a floppy" it's that we're using twenty-six arbitrary Latin characters to uniquely identify our drives.

John Booty said:

It would be quite sad if we're referring to drives by letters 5-10 years from now, wouldn't it?

The problem isn't that "A: is reserved for a floppy" it's that we're using twenty-six arbitrary Latin characters to uniquely identify our drives....

Unless you just use drive names, but other than unique names, what else would you name them? I don't see the problem with using letters to identify drive letters to identify them on a system. Unless you mean like "Is that file on my C-Drive", cause yeah, that is kinda stupid, as different people designate their drives differently.

Straight Baby Thighs said:
Unless you just use drive names, but other than unique names, what else would you name them? I don't see the problem with using letters to identify drive letters to identify them on a system. Unless you mean like "Is that file on my C-Drive", ...

Well, you need a unique way of identifying each drive (that the OS uses, like 21EC2020-3AEA-1069-A2DD-08002B30309D or something) and a human-readable way.

It's just that drive letters are a really stupid human-readable way. And they're doubly stupid since they're redundant, since each volume already has a friendly name. If you've already named C as "MikesAwesomeShit" why should you also have to call it "C?"

I mean, I know why - it's just a backwards-compatibility thing because friggin' Windows 8 is backwards compatible to Windows 7 which was backwards compatible to Vista which was [ etc etc etc etc ] which was backwards compatible to 3.1 which was backwards compatible to DOS 5.0 and the whole fuckin' thing was (literally) a cheap CP/M clone in the first place.

Every OS I know of besides Windows has a saner way of doing the friendly name - skip the arbitrary drive letter and just use the partition name. On anything Unix-derived (OSX, Linux, Android, etc) you just go by the volume name, eg cd /Volumes/MikesAwesomeShit/SomeFolder on OSX instead of c:\ ; cd MikesAwesomeShit on Windows.

It's not a problem for OSX if you have two drives both named MikesAwesomeShit because that's just a friendly label that's there for the user. (I actually don't know what that looks like from the Bash shell though)

Different OSs and different filesystems have various schemes for the unique identifier. Usually it's a long unique number that is a combination of the drive's serial number and the drive's partition layout. For example, http://...microsoft.com/...ware/ff552563%28v=vs.85%29.aspx

Meh in a couple generations, "My Computer" will just be the root of your filesystem and all devices will just be mounted folders. Just like in my fantasy where Linux finally makes it big on the desktop and people stop pirating Windows.

Sweetums said:
all devices will just be mounted folders

The trend is probably the other way around - representing many physical storage devices as a single logical drive. Like RAID, but even more generalized/abstracted.

Generally these are called storage pools although Microsoft seems to have invented "storage space" in an attempt to be different and/or avoid trademark issues.

http://en.wikipedia.org/wiki/ZFS#Storage_pools

Note that previous versions of Windows Home Server had a feature like this (fucking awesome, assuming it works) storage spaces thing in Windows 8:
http://...anandtech.com/...dows-8s-storage-spaces-detailed

Ah, Thanks for clarifying. Wasn't too sure what you meant, but yeah, that'd be rad.

It's basically just me complaining about all the backwards-compatibility stuff clogging Windows up.

I'm soooo happy this PC hardware thread is BLOWING up lately.

Has anyone else had trouble with SSD firmware? May as well mention this as a preemptive warning, I guess...

I endured months of random BSODs on my SSD RAID before finally trying to upgrade the firmware. Like, as a first time SSD owner I had no idea it played such a big role in the drive/motherboard interface. So, potential SSD upgraders beware: YMMV, but, It's A Thing.

Not really a big deal to fix, though. Just keep in mind you can't use a drive and upgrade it's firmware at the same time.

I had some fairly early SSD drives (OCZ Vertexes with the first-gen Indilinx controllers) that seemed to have semi-endless firmware updates. The issues weren't on the level of BSODs.

Having been "into" SSDs since 2008, the good news is that things seem to be settling down a bit. Manufacturers are figuring this stuff out and in a few years I think SSD firmware issues will be as rare as HDD firmware issues.

That said, I've gone for peaceful existence over raw performance with my last couple SSD purchases. Last four SSDs I've bought are Intel G2 and Intel 320-series (essentially the G3) based drives.

They don't have the highest raw transfer rates, but they're quite competitive for small random read/write performance, and honestly I'm just glad to have a really trouble-free experience.

Trouble-free is nice, especially since OB's server runs on two of the aforementioned drives.

quick question. I've been told never to raid on a SSD. Why is that?

Rexall said:
quick question. I've been told never to raid on a SSD. Why is that?

I know why you were probably told that, but it's not necessarily true.

When you delete a file, newer OSs (Win7, etc) send a "TRIM" command to SSDs that helps them clean up their free space more efficiently. In RAID mode, those commands don't reach the drives. That can lead to your drives slowing down over time.

That's not necessarily a dealbreaker. Most newer drives do a pretty good job of housekeeping ("garbage collection") to clean up that free space even without TRIM, so they usually work fine in RAID mode.

I would slightly question the wisdom of RAID-1'ing your drives. Very long story short, while you will get double the maximum transfer speed, that's not going to have a lot of impact on the stuff you actually care about - how fast the OS boots, programs load, etc. And obviously (in a typical RAID-1 setup w/ two drives) then you're doubling the risk of drive failure too.

That said, I do have two SSDs in RAID-1 in my PC... haha! Fuck it, it's fun to make shit go fast!

So the actual answer is not "never raid on a SSD" but more like "research to make sure RAID works okay with your particular SSDs." My experience with RAID 0/1 is that the onboard Intel ICH-xxx controllers are actually really awesome at RAID 0/1. You don't need a dedicated controller unless you're doing RAID 5 or something.

John Booty said:
And obviously (in a typical RAID-1 setup w/ two drives) then you're doubling the risk of drive failure too.

That's more with RAID0...lose one drive, the whole array is toast. RAID1 will still do double-duty to a disk, but losing one won't trash everything.

Personally, RAID on an SSD doesn't seem to make a lot of sense. RAID was largely designed as a way to quickly increase drive capacity and provide a balance between disk failure/replacement and having to restore from backup, with mechanical disks. Since SSD's really only wear out because NAND flash degrades with each write, RAID is kind of unnecessary.

I am still waiting for MRAM (Magnetoresistive Random Access Memory) tech to become more prevalent and hopefully supplant NAND as the dominant flash memory product (as well as DRAM). But NAND is cheap while MRAM isn't (yet), so no one is interested in setting up fab plants to produce the stuff.

http://en.wikipedia.org/wiki/Mram

Wow I totally didn't know about the hazards if RAID-1ing SSDs! But now that I know I still don't give a shit, haha.

The system configuration is so volatile on my desktop that I may as well be running every component this thing's got right into the ground.

Kumba is 100% correct. I was saying RAID-1 when I meant RAID-0.


RAID-1 is mirroring (2x reliability, 1x performance, 1/2x capacity)

RAID-0 is striping (1/2x reliability, 2x performance, 1x capacity)


When "enthusiasts" talk about RAID they mean RAID-0 most of the time. Very few "enthusiasts" would be mirroring drives for redundancy. :-)