|To:||Andi Kleen <ak@xxxxxxx>|
|From:||Michael Sinz <msinz@xxxxxxxxx>|
|Date:||Tue, 27 May 2003 09:24:16 -0400|
|References:||<1053694002.2887.1.camel@xxxxxxxxxxxxxxxxxxxxx> <1053697162.21472.51.camel@xxxxxxxxxxxxxxxxxxxx> <20030523134438.GC30288@xxxxxxxxxxxxx> <20030523150530.A31022@xxxxxxxxxxxxx> <20030524071709.GK27626@xxxxxxxxxxxxxxx> <20030524095245.A24074@xxxxxxxxxxxxx> <20030524091516.GM27626@xxxxxxxxxxxxxxx> <20030524093103.GA12181@xxxxxxxxxxxxx> <3ED344C0.1010700@xxxxxxxxx> <20030527120650.GA22306@xxxxxxxxxxxxx>|
|User-agent:||Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4b) Gecko/20030507|
Andi Kleen wrote: [...]
Yes there is: Unicode/UTF-8. That is where all the Linux distributions are going too. For legacy SMB support you will still need to support codepages,but that could be done by samba. For XFS I guess it would be enough to just support UTF-8. Supporting different code pages is probably not too useful anymore.
I would agree that a single standard is the best thing to do. Also, I fully understand why this has to be in the filesystem in order to produce a reasonable performance result. Too bad this also means that any hashing has to be redone. Also, too bad the volume needs to be mounted as one way or the other... I used to be a case-insensitive fan for any user-interface items. I still thing that the average user does not care about case in most cases other than for visual aesthetics. Preserving case is important. However, after having worked on the technical issues when you go international (and back in the day where 16meg of RAM was a HUGE amount and most had less than 1meg) BTW - I still remember when the first C++/C-front stuff was suggesting that the C++ files be named "file.C" vs "file.c" for regular C code. What a mess! -- Michael Sinz -- Director, Systems Engineering -- Worldgate Communications A master's secrets are only as good as the master's ability to explain them to others.
|<Prev in Thread]||Current Thread||[Next in Thread>|