On Wed, May 22, 2002 at 12:12:46AM +0200, petter wahlman wrote: > On Tue, 2002-05-21 at 22:19, Rodrigo Barbosa wrote: > > Attention: RANT LEVEL set to 7. > > > > On Tue, May 21, 2002 at 09:11:24PM +0200, petter wahlman wrote: > > > So, why are 'linux' filesystems case sensitive? > > > > I think the question is "Why is Windows case insensitive" ? > Or OS/2 and Novel witch are some of the other OS's I share code with. OS/2, Novel and Windows all shame the same historical codebase, that is, PC-DOS. No surprise they all share the same behaviour. > > Last time I checked with my language teacher (no matter what language), > > "A" and "a" are not the same. So, we can reframe your question as > > "Why do we have different cases for the same letter of the alphabet ?". > > In that case, I really can't answer. > Semantically speaking, I agree that 'A' != 'a'. The number of possible > filenames are also gratly increased with case sensitivity. > I would probably religously be against such a change if I did not > experience the previously mentioned problems. But this problems you experience is easily solved. I can even code a small script for you if you give me the details. > > As for Linux, that is the way Unix is since the beginning, in the 60's. > > And Linux is a Unix-like derivative. Other exemples include Solaris, > > AIX, Irix, MacOSX, FreeBSD, bsd386, NetBSD, SunOS etc. > I know, and can proably think of other examples, but that does not > change the fact that the lack of such a feature causes problems. > I do not think that changing the default behavior is the right think to > do, though. Good. So, what are our solutions ? If it is made a configurable behavious, all programs would have to deal with that. > > > Having two files or directories that only differ by case is IMO wrong, > > > but should at least be controlable with a generic mount option. > > Just becouse you don't like something does not make it wrong. I, for one, > > use different cases for different functionalities. Lets say I have a > > datafile (data.dat) I want to disable, but want to know what this file > > is. I simply rename it to data.DAT, or Data.dat. The application will not > > use it anymore, and I still know the correct name. That is one of several > > possible uses. > Wrong was probably a too strong word (i wanted to get some feedback on > my post :), and can absolutely see the value of having the possibility > you mention. Thanks. At least you read about the "rant level" of the mail, before reading the rest of it :-) > > You see, most people want to be able to control what a program (application, > > OS etc) do, and not just put some crap in and hope the program will work > > it's way. Old computer proverb: Garbage in, garbage out. > > If you need to conver filename, you can easily do it with a onliner. Same > > for file references inside a Makefile or anything else. > No, it is not that easy. I do not have the permission to change the case > on the include statements in all the source files I access, due to RO > and locked files. > It's not as easy as doing a 'find' ;) So, change the name of the files. Also, easy to do with a oneliner :-) And do understand your problem, but changing the VFS would only cause further problems (curing a headache by cutting the head off). This problems should be solved in userspace. []s -- Rodrigo Barbosa - rodrigob at tisbrasil.com.br TIS - Belo Horizonte, MG, Brazil "Quis custodiet ipsos custodes?" - http://www.tisbrasil.com.br/ Brainbench Certified -> Transcript ID #3332104 -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
