Re: user/group permissions in linux

[deepak <deepak2000@xxxxxxxxxxxx>]

hmmm....... it can't be a bug , since it is well defined in many good books and they never say that is a 
bug. i think it provides better control over the file access to the owner. 

infact by ur logic we may not even require permissions for the owner & always grant him all access 
since the owner knows what should/shouldn't be done to files created by him !!

to give some idea about better control here is an eg

since most of the file created by me would be in my folders through which i pass on often, i may 
accidentally give some command say 
$echo "deepak" > deepak
without realising that a file named "deepak" already existed there with some important data.
since other ppl in the group have to specifically write the full path to write something there so(hopefully) 
they would be knowing what they r doing . by turning off the owner write permissions i could prevent my 
self from overwrititng its content accidently.    

hope it is a good eg , may be someone comes up with better example. another (similar) could be with 
execute permisions , though an eg in support of  "read " access would be tough.
_______
deepak


>consider the below case:
> 
> file : temp
> perm: ----rwx---
> owner: myself
> group: myself
>
> i.e a file temp is owned by me and i'm the only member in the group
>"myself". if i try to open the file it gives an error that permission
>denied for opening the  file.
>though the file cannot be opened based on the user permissions, it has to
>be opened due to the rwx permissions for the group.
>
>this is because linux kernel makes accessibility test based only on the
>user permissions if the user is the owner of the file and never considers
>the group permissions in that case.
>
>is this not a buggy feature......or is there any reason behind this
>design.
>please comment.
>
>sriram,
>rec,surat.
>

________________
  reply soon
  bye
  __d ee p a k__
________________