[vlc-devel] [OKFILTER] vlc-devel Digest, Vol 56, Issue 41

[John Freed]

A few observations:

1) I'm new to VLC but not to project management with volunteers. Shelley's
points are well taken.
2) To Karlo's point about "Bad/fake/broken promises are the cancer of
software industry" -- I totally agree when those promises are made as
marketing tools, that is, to users or potential users. A roadmap is no such
thing -- it is an *internal* guideline -- in essence, a promise (more of a
goal-setting) to *each other*.
3) We already have a roadmap. The very statement that a "feature freeze" is
proposed for May 31 means that, if agreed, the roadmap has a feature freeze
of May 31. That's rudimentary, but on the right track. A better roadmap
might be:
--Today through Feb. 15: brainstorm new feature ideas (like Shelley's on
bookmarks) for what would be 3.0 in my proposed numbering scheme. (Face to
face meetings are helpful for this, I find.)  No ideas are discarded during
this phase, even if wild, ridiculous or harmful to VLC.
--March 31: core team members finish prioritizing features they feel are
important and divvy up the ones they are willing to commit to working on.
Others are placed in the "projects for outside volunteers" bin.
--April 1 to May 31: additions to feature list considered and triaged as
they come up. None are written off unless considered harmful to VLC. Less
important ones, or ones the core members don't want to tackle, go into the
"projects for outside volunteers" bin. No further features are considered
for this release; they go into 4.0. (At this point, a 4.0 branch can be
opened.)
--Sept. 30: Target for first pre-release, version 3.0.x. Features that have
nobody working on them are reassigned to 4.0.
--Dec. 31: Target for public release. (3.1.0 in my proposed numbering
scheme.)

I'm fairly sure Shelley has something like this in mind. Milestones and
their dates can be debated, but I think a roadmap, or outline, is useful.
As I noted, JB and Remi are already debating one aspect of a roadmap. To
Karlo's point, nobody is penalized for missing the deadline; the feature
simply misses the 3.1 release and can always be added to 3.2 (development),
3.3 (release) or 4.0 when ready. Also to Karlo's point about fixing
release-blocker bugs: failure to fix them would delay the target dates, of
course, but that's why they're targets.

In any case, this is all theory based on what has worked for non-VLC
projects. I see that Christophe Mutricy has some real-world VLC project
management experience and wonder what his views are.


On Mon, Jan 9, 2012 at 6:52 AM, <vlc-devel-request at videolan.org> wrote:


> From: Shelley Horwitz <shelley at sjcomm.com>:
>
> I have not seen any comments about my 3 requests:
>
> 1. To have bookmarks saved in a file so I may reuse them on subsequent
> viewings of video and movies.
> 2. To be able to name the bookmarks (like bookmarks in my browser).
> 3. To have a bookmark button in the controls so I don't have to go to the
> menu.
>
> Any comments or opinions about the desirability of these features or
> priority for having them?
>
> Shelley
>
> From: Shelley Horwitz <shelley at sjcomm.com>:
>


> Kaarlo,
>
> I will try to remember to switch to plain text for all my messages. Thank
> you for the reminder.
>
> Now... to our discussion. I do not want to make this a personal debate....
> only to make VLC a better product in the world of players.
>
> Even in volunteer organizations, and I have worked in several (Boy Scouts,
> Girl Scouts, for example), most people react well to goals, targets, and
> schedules. Actually, I believe the majority of people working on projects
> in volunteer organizations appreciate and work better in a scheduled
> environment. The schedule is NOT a threat, but a guide that helps them
> shape their work and pleasure hours to get the most out of their
> contributing energies. In my experience, a lot of people do not focus their
> minds and activities when they have no clear idea of when something is due.
> Because there is no "time-limit," there is no focus on performing a task.
> Because there is no time limit, there is no need to work today... there is
> all the time in the world to do it tomorrow. Everybody procrastinates when
> there is no well-defined finish line in sight. People don't procrastinate
> when they have a goal and a timeline... either given to them or they set it
> themselves.
>
> [snip]
>


> Date: Sun, 8 Jan 2012 14:29:45 +0200
> From: Kaarlo R?ih? <kaarlo.raiha at gmail.com>
>


> [snip]
> And you are completely missing the point of volunteer vs. paid developers.
> Since most people related to VLC development do this as a hobby, it isn't
> feasible to make them work with any schedules or force them to do certain
> features. Even current releases are stressful events because only few
> people do release blocker bug fixing, and if someone would be forced to do
> that, I think they would quit pretty fast.
>
> IMHO roadmaps aren't useful if there isn't any guarantee for events in it
> to happen. Bad/fake/broken promises are the cancer of software industry.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/vlc-devel/attachments/20120109/8da706c5/attachment.html>