https://bugs.documentfoundation.org/show_bug.cgi?id=95416
Eike Rathke <[email protected]> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |[email protected]
--- Comment #35 from Eike Rathke <[email protected]> ---
(In reply to Tor Lillqvist from comment #33)
> But isn't it *using* stuff from a wrong include file that indicates
> something is wrong, not where the include file is located?
Yes it is, though maybe inevitable in some filter code cases, but moving the
file to a higher hierarchy level and adapting the include effectively hides
that and invites for further abuse all over the place.
> Should we instead then have a difficultyInteresting easyHack (a
> contradiction in terms...) to find these includes and fix the code to not
> need stuff from such a "wrong" include file at all?
We can try.. but having taken a look at the mentioned ScDocShell case it's
definitely not just a five minute task [1] ... and no quick idea how it
actually could be solved. difficultyInteresting certainly is applicable ;-)
However, there *are* the easy cases where moving a header is appropriate, but
it has to be decided on a case by case basis keeping header scope in mind, and
reviewers should be aware of the underlying problem.
[1] http://hea-www.harvard.edu/~fine/opinions/fiveminute.html
--
You are receiving this mail because:
You are the assignee for the bug._______________________________________________
Libreoffice-bugs mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs
