Hi Andrei, On 2012-04-02 05:28, Andrei POPESCU wrote:
On Du, 01 apr 12, 10:51:57, David Prévot wrote:Le 31/03/2012 15:36, David Prévot a écrit :Le 31/03/2012 04:36, Andrei POPESCU a écrit :[...]In order to be a little more constructive, let me share some thoughts we recently had while discussing the publicity workflow (those are mainly if not exclusively Francesca's ideas, who is taking care of mostly if not all the DPN work these days, so she may correct if needed). In order to write smoothly the DPN, we could really better share the work: it's OK to write and organize the last missing bits the Friday before the issue, but it's less effective if we write it all at the last minute form scratch. We add some TODOs when we see DPN material during the two weeks, but we are not really good to draft stuff before the last Friday, even if some people actually replied to the recent call to help the publicity team. Maybe should we have a regular (short) IRC meeting once every two weeks (the weekend we don't actually send the DPN), in order to share the TODO items of the DPN and of other publicity fronts. Any thoughts about this proposal?While I know IRC is very valuable, IMVHO you may be excluding potential contributors[1] if you don't somehow include those following only the mailing list(s). [1] me excluded, not been much of a contributor anyway :p How about posting the IRC log?
I strongly support efforts to make Debian [more] transparent, but this particular issue doesn't worry me. I care to see discussions that lead to decisions affecting what we ship, but if the meeting is merely for sharing the workload, I don't care that it's private to its victims^Wparticipants, as long as we don't start giving ownership of news items to their writers. Potential contributors should simply participate to the meeting or write news items before the meeting so they don't miss any discussion anyway. If they can't do that, the meeting's assignments wouldn't be binding, other contributors can still beat whoever had an item assigned to writing the item. Or, if the person assigned beats a potential contributor and the potential contributor is not satisfied with the result, he should just get the news item changed.
Unless you mean that potential contributors would be hesitant to write an item thinking that someone else may have it assigned and write the item at the same time. In that case, I imagine that assignment of TODO items could be commented in the repository after the meeting.
However, logging #debian-publicity would surely help make the team's activities in general more transparent. This was already discussed (see https://lists.debian.org/debian-publicity/2011/08/msg00035.html ).
That being said, I'm skeptical about the realism / helpfulness of such meetings. Is it frequent that a writer gets to commit an item and realizes someone else just wrote it? I wasn't aware that the DPN was written at the last minute until David noted it. What is causing this? Are we short on manpower so everyone waits hoping that someone else will write before? Or are writers just waiting as much as possible so they don't need to update an entry about an ongoing event? If it's just more optimal to wait as much as possible before writing an item, meetings might not help much. On the other hand, if the problem is manpower / procrastination, I imagine it will be difficult to hold meetings. If we have a manpower issue, we may look at crediting contributors [more] or at easing contributions.