http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4788
[EMAIL PROTECTED] changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |[EMAIL PROTECTED]
------- Additional Comments From [EMAIL PROTECTED] 2006-02-11 12:04 -------
My thought is this...
sa-update may exist in 3.1.0, but it isn't operational. 3.1.1 is to have the
first working sa-update meant for production usage. I believe we should either
decide to do it 100% now in 3.1.1, or defer it entirely until 3.2 if this is too
much risk for 3.1.1. A half-way job would be undesirable for several stated
reasons.
It comes down to a question of priorities. If this important enough to us for
our users to have rapid rule/score update capability long before the release of
3.2.0, then we should do it 100%. If this isn't that important, then we should
defer it.
Of course upstream is free to make any other decision. My participation here is
only to point out that updating files in /etc is a bad design and we in
particular wont be able to use the software if it relies on that location. It
makes a lot more sense to update versioned directories in a location recommended
by FHS. For Fedora and RHEL, this is an extra complication because it actually
matters in runtime due to restrictions within SELinux policies.
If 3.1.1 is shipped with production sa-update but without the location,
Fedora/RHEL may ship 3.1.1 at some point, however I will be forced to make a
decision like 1) Disable sa-update or 2) Patch in the 3.2 behavior in our
shipped package. Both of these options are very undesirable.
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
