On 01/27/2014 09:17 AM, Petr Spacek wrote:
On 27.1.2014 08:07, Martin Kosek wrote:
On 01/24/2014 05:23 PM, Simo Sorce wrote:
On Fri, 2014-01-24 at 17:17 +0100, Petr Viktorin wrote:
On 01/24/2014 04:57 PM, Simo Sorce wrote:
On Fri, 2014-01-24 at 16:48 +0100, Petr Viktorin wrote:
...
Technically we could alias the name so the attribute can be called
either way, but that is not necessarily a good option either.
If breaking master is unacceptable, we can use the old name instead.
ipaPermIncludedAttr is more consistent but ipaPermAllowedAttr isn't
downright wrong.
Ok, let's hear other opinions, I see a lot f value in consistent naming,
and not breaking a developer build is not that strong of a reason to
have substandard naming I guess. What do others think ?
Simo.
Hmm, I obviously see things differently here. I would rather "break"
the master
and let developers running on the git version to reinstall the servers
(including myself) than to have to live with suboptimal attribute name
for ever
or by adding unnecessary cruft to the code...
I think you got lost in the "not"s. You and Simo agree.
(Speaking as lab admin:) Please, break it! It will force people to
finally reinstall years old VMs! :-)
--
Petr³
_______________________________________________
Freeipa-devel mailing list
Freeipa-devel@redhat.com
https://www.redhat.com/mailman/listinfo/freeipa-devel
