On 31-08-2010 10:03, Robin H. Johnson wrote:
> On Mon, Aug 30, 2010 at 11:13:15PM +0200, Michael Weber wrote:
>> Hello fellow developers.
>>
>> On 08/30/2010 04:20 PM, Dirkjan Ochtman wrote:
>>> Sounds good to me, but I'd actually be more interested in having
>>> something the other way around; i.e. monitoring for activity in
>>> commits, bugzilla, IRC and maybe the -dev mailing list to see if
>>> people are still active and send them a message to encourage them to
>>> set devaway if they haven't been active in, say, 15 days.
>>
>> I think the intention was to force actually active developers to
>> remove their out-of-date .away message, which isn't very representative
>> for the project.
> .away age statistics, as of right now (2010/08/31, 07:27 UTC).
> - 53 developers with .away files.
> - Oldest: 2007/Mar/01 (1278.8 days old).
> - Mean: 153 days old.
> - Median: 55.5 days.
> - First, Third quartiles: 23.3, 136.5 days.
>
> What do the numbers mean? My opinion looking at them is that MOST
> developers are using the .away system correctly, however some developers
> just have forgotten to remove old .away files (they claimed they would
> be back by a date, and commits started up after that).
>
> I'll fully admit that I neglected to remove my last .away until I
> double-checked earlier today.
Yes, it was my case also. I thought I removed it though, but it was just
my memory playing tricks.
Thanks for raising the thread :)
>
> How about this as an idea:
> 1. Include a parsaable return date I suggest ("Returning:YYYY/MM/DD",
> "Returning:Unknown")
> 2. Automated emails when:
> 2.1. It's after the return date (weekly).
> 2.2. You start committing again.
>
Notifying about the away state when you commit, sounds great.
