Suggest a USB-installable Linux distro to try on a 256MB RAM PC

esuhl

Kernel_Sanders wrote: »

Thank you for that, I was trying to work out just why I couldn't understand it! He lost me when he said one difference with Linux is that other drives are mounted as a subdirectory in the root filesystem, as if the drives in Windows weren't subdirectories of the directory named 'My Computer' :undecided

"My Computer" isn't a directory and doesn't exist as part of the filesystem at all; it's just an abstract visual representation of a root directory. You can't change directory to "My Computer" and you don't refer to directories like this: \My Computer\C:\subdir.

In Windows, there is no root directory like there is in Linux, each drive has its own root: A:\, B:\, C:\, etc.

Kernel_Sanders

esuhl wrote: »

In Windows, there is no root directory like there is in Linux, each drive has its own root: A:\, B:\, C:\, etc.

Still trying to get my Microsoft-indoctrinated mind around this!
So what is the point of root in Linux? Is it anything more than just a / ?

-TangleFoot-

Kernel_Sanders wrote: »

So what is the point of root in Linux?

It's a bit like RAID, only... not. Essentially, all storage devices are treated as though they were a single unit. There's no drive letter because there's only one drive as far as the system is concerned.

/ = root
/dev/sda = first HDD
/dev/sda1 = first partition on first HDD
/dev/sdb = second HDD

esuhl

Kernel_Sanders wrote: »

Still trying to get my Microsoft-indoctrinated mind around this!
So what is the point of root in Linux? Is it anything more than just a / ?

The point of the root directory is to contain all other sub-directories for the WHOLE filesystem. Windows doesn't have a single root directory that contains all others; instead it assigns each drive partition a letter and mounts the whole partition using the drive.

In other words, Linux mounts every available partition somewhere under /. Windows mounts partitions separately using A:\, B:\, C:\, etc.

There's no real reason that one method is better than another -- the only reason that either models exist is so that you can reference files.

One obvious difference is that you can't tell just by looking at a Linux filesystem how the directories are shared across drives. You see the root (/) directory, and below it /usr, /home, /bin, etc. Those subdirectories might all be on the same disk or not... Using Windows you can only refer to a file by knowing which arbitrary letter has been assigned to it.

The Windows method seems (to me) to be less logical because it involves using arbitrary letters and forcing users to be aware of how their hardware is configured in separate partitions (which is unnecessary). Linux doesn't use arbitrary letters; instead filesystems are mounted to sensibly-named (one hopes!) directories.

Hope that helps! It's very simple... but surprisingly hard to explain clearly!

GunJack

esuhl wrote: »

The Windows method seems (to me) to be less logical because it involves using arbitrary letters and forcing users to be aware of how their hardware is configured in separate partitions (which is unnecessary). Linux doesn't use arbitrary letters; instead filesystems are mounted to sensibly-named (one hopes!) directories.

Interestingly, this to me is how windows is MORE logical...for instance, you have a file "splod", on F: drive, which happens to be a usb drive. You remove the usb drive, and you can't get the file "splod" that's stored on it. Because you're aware of the hardware config, you know you're wasting your time trying to access that file on the F: drive if it's not plugged in....IYSWIM

Nowt wrong with forcing peeps to be more aware of the system they're using

-TangleFoot-

GunJack wrote: »

Because you're aware of the hardware config, you know you're wasting your time trying to access that file on the F: drive if it's not plugged in....IYSWIM

The last time I plugged a USB stick into a system running a Linux distro the file manager displayed the device as a separate unit with its volume label (MY_LIL_PNY) in plain view. Isn't technology wonderful?

esuhl

GunJack wrote: »

Interestingly, this to me is how windows is MORE logical...for instance, you have a file "splod", on F: drive, which happens to be a usb drive. You remove the usb drive, and you can't get the file "splod" that's stored on it. Because you're aware of the hardware config, you know you're wasting your time trying to access that file on the F: drive if it's not plugged in....IYSWIM

Nowt wrong with forcing peeps to be more aware of the system they're using

Yeah... but it's not immediately obvious that F: relates to the USB stick as opposed to any other device. There is, of course the Linux device equivalent, e.g. /dev/sdf1 (i.e. the first partition on "drive f" -- the sixth hard drive).

It makes more sense to me to think about whether /mnt/USBstick contains mounted files than whether F: is mounted...

And imagine you had a system with 1000 remote users, and the filesystem was so big that you needed 50 hard drives to contain it. It would be really confusing to the users if they actually knew that they had 10 files on one drive, 3 on another, 6 on the next, etc.

To be honest, you can think of lots of scenarios where one method might be easier to get your head round than the other. It's all pretty arbitrary.

Personally, I find the Windows system easier to use (and easier to explain), but the Linux system is easier to think about, more flexible... and more elegant!

As my old dad used to say, "It's all swings and roundabouts".

esuhl

-TangleFoot- wrote: »

The last time I plugged a USB stick into a system running a Linux distro the file manager displayed the device as a separate unit with its volume label (MY_LIL_PNY) in plain view. Isn't technology wonderful?

That's the great thing about Linux. It's so flexible, and so easy to write scripts to automate everything. You want stuff mounted automatically in a subdirectory with the name of the volume label? No problem. You want the file manager to display partitions on removable devices as not being subdirectories of the root directory? That's fine too (an illusion, perhaps).

You just need the technical aptitude... which I don't really have. But I'm happy enough for others to do the hard work if they're going to make it available under an open source licence!

-TangleFoot-

esuhl wrote: »

Yeah... but it's not immediately obvious that F: relates to the USB stick as opposed to any other device.

Those everything-in-one card readers are particularly confusing. :laugh:

esuhl

-TangleFoot- wrote: »

Those everything-in-one card readers are particularly confusing. :laugh:

I know what you mean -- my friend's PC has half-a-dozen slots, creating half-a-dozen drive letters. There's two optical drives plonked randomly in the middle of the list (drives F: and K: ) and a virtual drive at H:. The first USB stick you plug in tends to get mounted at J:! It's nuts!