"Erich E. Hoover" erich.e.hoover@gmail.com writes:
On Fri, Dec 19, 2014 at 11:13 AM, Alexandre Julliard julliard@winehq.org wrote:
... I don't think you should be defining libport functions at this point.
I’m really hesitant to make such a change, the proper location for this code is somewhere that can be shared by the wineserver and ntdll. We’ve been field testing these patches for over a year and, to the best of our knowledge, any odd complexity in the code is to combat quirky behavior that we’ve seen as a result of user feedback.
This is actually the reason we implemented DOS Attribute support at all, we wanted to reduce the size of the patches for SD/ACL support to a more manageable size so that the patches would be easier to review.
TBH I'm not convinced that adding a custom xattr that no other application recognizes is the right way to do ACL support, as opposed to for instance mapping to Unix ACLs.
There may be good reasons for doing things the way you do, but if so you have to present convincing arguments. Testing patches in the staging tree is of course a good thing, but it's not a substitute for a proper design. Particularly for things that are stored on disk, getting it right from the beginning is very important, since you can't easily change it later.