Messages in this thread | | | Date | Mon, 21 Jul 1997 19:53:33 -0400 | From | James Mastros <> | Subject | Re: FAT/VFAT/FAT32 reimplimintation... |
| |
At 08:54 AM 7/28/97 +0200, you wrote: >> A) For normal files, the default is the same as the mountpoint, but a-x. > >So you won't be able to execute shell scripts and true Linux binaries >off a DOS partition. You have a point; but having everything executible is nearly as bad...
I would say that it should check for executibles by file magic (look for a ELF header/a.out header/script header), execpt that (1) I can't go through binfmt_* to check (I don't think there is a simple function to check if a file can be executed without acatualy executing it, I'm not in Linux now, check later...) (2) many scripts don't have a #! magic, but rely on the shell to exec them. (3) It feels like it would slow stuff down to much.
I could use some of the reserved bits to store full unix permisions/owner, but (1) I don't know if there are enough. (2) Microsoft might use them for somthing eventuly.
> >> 3) I'm planing on, eventually, making .lnk files work as symlinks... > >Do you know the format of these files? As far as I can tell, this >is OLE structure storage of monikers... > That much I knew, but that's all. I was planning on trying to guess... I took a look a while ago, and it looked like the filename was at a fixed offset, and the rest I can just throw out.
>> F) If you use the .lnk->symlink mount option, files with the .lnk >> extention will have it striped. If you don't, they won't > >So you are risking files with duplicate names in the same directory. Ouch. I hadn't thought of that.
> >> 5) I am in desperate need of a good reference to the fatish (sounds kinda >> kinky!) filesystem layout... (the directory entries, esph...) (Free is much >> preferable, I'm a starving highschool student... (OK, I'm starving because >> I can't concrantrate on eating... > >Have you checked Ralf Brown's interrupt list? I know what your speaking of, and though I have seen it before, that was 3+ years ago, and I can't seem to find it (but I havn't looked much yet).
> >Good luck, >Martin >
Thanks for the feedback, James Mastros
| |