On Fri, Apr 18, 2014 at 3:12 AM, Benoit Claise <bclaise@xxxxxxxxx> wrote:
Hi,
Hi Benoit,Thanks.
At 07:14 17-04-2014, Benoit Claise wrote:
We discussed during the milestones during one of the last plenaries.
Yes.
The discussion was around: the milestones are indications and not deadlines.
In Open Source project, these are deadlines.
In the IETF, I would love to find a middle ground between the two.
In the IETF, the only "deadlines" are the meetings, or to be more precise, the submission deadlines just before the meetings.
Lisa used to send monthly progress reports. It provided some visibility into what was happening within the area.
A milestone can be an estimate of when X can be completed. I might suggest one and the group decides whether everyone can target that date and plan accordingly. There are times when deadlines may be needed as an effort can end up in failure.
Exactly, and we should understand why!
I like this tool: <http://www.arkko.com/tools/lifecycle/>http://www.arkko.com/tools/lifecycle/<draft-name>-timing.html
For example: <http://www.arkko.com/tools/lifecycle/draft-ietf-netmod-interfaces-cfg-timing.html>http://www.arkko.com/tools/lifecycle/draft-ietf-netmod-interfaces-cfg-timing.html
Where is the bottleneck? Is this a process issue? The authors/shepherd/IESG/RFC-editors?
I don't want to finger point, but understand what we should improve.
Let's see. :-) The shepherd write-up does not provide me with much information. There's three years from the first -00 to the latest version of the draft. The author activity looks okay. There is less activity from the document shepherd (there is an explanation for that). There was a Last Call in April 2013. The AD asked for a revised I-D in May 2013. There was another Last Call in December 2013. The IESG didn't like something in the draft and it took seven months to address that.
There's still the two and a half years. There are seven months between the date in the charter and the first Last Call. It would take some effort to figure out what happened during the first two years. It looks like the working group was the bottleneck.
I'm more after analyzing trend than inspecting this particular document.
For example, https://datatracker.ietf.org/iesg/discusses/ doesn't look like a very long list to me.
I wonder if this NETMOD draft is typical of other WGs or not.
The idstats say it has been a WG draft for 388 days.
I think the original charter was about 160 days, but I can't tell since
all the relevant milestones were deleted from the WG charter page last week.
The problem is obvious. The WG spent lots of energy going over and over
the same issues as more new people read the draft for the first time. These people
often brought genuine concerns and new insights to the current solution, and the
WG was quite willing to ignore the milestones and accommodate the new feature
requests.
Regards, Benoit
Regards,
S. Moonesamy
.
Andy