History Tracking #2
Conversation
|
Oh yes - its based on cocox's fork! |
|
I don't like the idea of fractional hours. It would be much better to store values as minutes instead of hours. My patches from 2009 at http://git.mantisforge.org/w/timecard/dhx.git?a=shortlog;h=refs/heads/modtimestamp already implement resolution at the minutes level, with an optional toggle to allow administrators to lower the resolution back to hours. This lets people choose whether they need fine grained resolution (there were complaints with an earlier version of my patches that some people didn't want to allow high resolution estimates). The issue on John's tracker with more information is http://leetcode.net/mantis/view.php?id=59 For this reason, I don't support this pull request. I do however agree with the key purpose of this pull request: to prevent estimates/times being leaked via issue history if a user does not have permission to view this information. It would be better to try and hide history at the time the history is being viewed -- although this is currently not possible with MantisBT's history system. I'd like to hear what John thinks about adding the toggle to disable recording of TimeCard history. |
|
Hi David, actually I agree about the fractional hours, but I was lazy and used that fork because it contains fixes for a couple of other issues as well. By the way, what is the purpose of the 'TimeCard' field as opposed to the 'Time Estimate' field? |
If access to timecard functions is restricted to say developers and above via the thresholds on the config page, anyone can still see the estimates and time spent in the Issue History. We are using this for internal estimating and don't want customers who submit issues to see that. Hence I've added a config option to enable/disable tracking changes to time estimates and time spent. I couldn't find a way to disable tracking on a plugin by plugin basis based on access level, so I've done it this way. Note tracking defaults to off.
Also fixed a problem in log_time.php: now uses auth_get_current_user_id() rather than auth_get_current_user_cookie().