[advocacy-discuss] "What is OpenSolaris" page

[Jason King]

On Tue, May 6, 2008 at 5:40 PM, Ceri Davies <ceri at submonkey.net> wrote:
> On Tue, May 06, 2008 at 07:51:16PM +0100, John Levon wrote:
>  > On Tue, May 06, 2008 at 11:25:21AM -0700, Rich Teer wrote:
>  >
>  > > > I'm genuinely confused: what parts of the OpenSolaris don't make Solaris
>  > > > a better Solaris? Which features are retrograde steps?
>  > >
>  > > /usr/gnu being at the front of the default PATH for one.
>  >
>  > Please give a detailed technical reason for why that's an issue. AFAIK
>  > pretty much every GNU tool implementation is both more functional and
>  > significantly faster than the ancient system V ones.
>
>  For the OpenSolaris/Indiana release branch, it isn't an issue.
>
>  For the "Solaris will look like this now", it certainly is.  Changing
>  my .profile isn't the issue, it's the users and ISVs I'm concerned about
>  there.  For an example of extreme difference in simple output:
>
>   ceri at grudge:~$ ls -vd /
>   /
>   ceri at grudge:~$ /bin/ls -vd /
>   drwxr-xr-x  27 root     root          33 May  5 20:55 /
>      0:owner@::deny
>      1:owner@:list_directory/read_data/add_file/write_data/add_subdirectory
>          /append_data/write_xattr/execute/write_attributes/write_acl
>          /write_owner:allow
>      2:group@:add_file/write_data/add_subdirectory/append_data:deny
>      3:group@:list_directory/read_data/execute:allow
>      4:everyone@:add_file/write_data/add_subdirectory/append_data/write_xattr
>          /write_attributes/write_acl/write_owner:deny
>      5:everyone@:list_directory/read_data/read_xattr/execute/read_attributes
>          /read_acl/synchronize:allow
>   ceri at grudge:~$
>
>  So a) the GNU tool is missing functionality and b) this might have
>  broken anything and I (and much less Sun) have no way of knowing what.

It's slightly worse than that.  It would appear that in a default
setup, none of the common tools will support ZFS acls on Indiana.
This to me is a big loss.  IMO, one problem *solaris has is that a lot
of the very useful (and cool) features aren't well known.  Making
things even less accessible is a step backwards.  I don't think anyone
is against adding new features, but not at the cost if losing (or
burying) existing useful functionality.

I would rather see a merging of features from the GNU and Solaris
utilities (as much as there's no mutually incompatible overlap with
options).  Which code base gets the modifications is of lesser
importance.