projects / unix-history / blame
| Commit | Line | Data |
|---|---|---|
| 584fddaa JH |
1 | # $Id: TODO,v 1.3 1993/07/19 13:40:00 cgd Exp $ |
| 2 | ||
| 3 | 1) should understand "older", original High Sierra ("CDROM001") type | |
| 4 | ||
| 5 | Not yet. ( I don't have this technical information, yet. ) | |
| 6 | ||
| 7 | 2) should understand Rock Ridge | |
| 8 | ||
| 9 | Yes, we have follows function. | |
| 10 | ||
| 11 | o Symbolic Link | |
| 12 | o Real Name(long name) | |
| 13 | o File Attribute | |
| 14 | o Time stamp | |
| 15 | o uid, gid | |
| 16 | ||
| 17 | Except follows: | |
| 18 | ||
| 19 | o POSIX device modes | |
| 20 | ||
| 21 | I have no idea right now, we should check a REAL implementation | |
| 22 | for 386BSD..... | |
| 23 | ||
| 24 | o Limitation of 8 level directory( ISO9660 limitation ) | |
| 25 | ||
| 26 | Rock Ridge Extension are defined with the "CL/PL/RE" for getting | |
| 27 | rid of this limitation. But as far as I test the cdroms,I'v never | |
| 28 | seen this definition and we can access the over 8 level without | |
| 29 | it. (Another word, this limitation is NOT physical ISO9660's | |
| 30 | FORMAT limitation for unix stuffs.... I believe... ) | |
| 31 | ||
| 32 | 3) should be called cdfs, as there are other ISO file system soon possible | |
| 33 | ||
| 34 | Not yet. Probably we should make another file system when the ECMA draft | |
| 35 | is valid and do it. For doing Rock Ridge Support, I can use almost same | |
| 36 | code. So I just use the same file system interface... | |
| 37 | ||
| 38 | 4) should have file handles implemented for use with NFS, etc | |
| 39 | ||
| 40 | Yes. we have already this one, and I based it for this release. | |
| 41 | ||
| 42 | 5) should have name translation enabled by mount flag | |
| 43 | ||
| 44 | Yes. we can disable the Rock Ridge Extension by follows option; | |
| 45 | ||
| 46 | "mount -t isofs -o -norrip /dev/cd0d /cdrom" | |
| 47 | ||
| 48 | 6) should run as a user process, and not take up kernel space (cdroms | |
| 49 | are slow) | |
| 50 | ||
| 51 | Not yet. And addition, we should try to avoid a long seek by a absolute path | |
| 52 | with using the PATH TABLE or other method. | |
| 53 | ||
| 54 | 7) ECMA support. | |
| 55 | ||
| 56 | Not yet. we need not only a technical spec but also ECMA format | |
| 57 | cd-rom itself! | |
| 58 | ||
| 59 | 8) Character set change by SVD ( multi SVD support ) | |
| 60 | ||
| 61 | Not yet. We should also hack the other part of system as 8 bit | |
| 62 | clean. As far as I know, if you export the cdrom by NFS, the client | |
| 63 | can access the 8 bit clean (ie. Solaris Japanese with EUC code ) | |
| 64 | ||
| 65 | ---------- | |
| 66 | Last update July 19, '93 by Atsushi Murai. (amurai@spec.co.jp) |