[tasklist-idea] Re: GNU task list management

[Richard Stallman]

    * The traditional task list (prep/tasks)

    Now obsolete.

    The traditional task list at www.gnu.org was a texinfo file where to
    put GNU tasks. The entry point for the task list was
    address@hidden, pointing to a RT queue. 

I'm not sure what "entry point" means in a context like this.

    There is a link to prep/tasks on www.gnu.org/help 

What is in prep/tasks now?  Is the obsolete file still there?  Does it
still contain a list of tasks, or does it have a pointer to the list
on Savannah?  Should we delete this file to avoid confusion?  Should
we delete the link?

    * The "GNU Help wanted items" savannah project 

    Some months ago the traditional task list (at prep/) became
    obsolete. The replacement was the "GNU Help wanted items" savannah
    project, managed by Toby. The rationale for that migration was the
    great savannah capabilities to reach developers.

    Now we are using the job posting and news savannah facilities for
    publishing the requested tasks. 

Yes, this is the new method.  

Is there any reason we need to change it? 

    There is also the tasklist-idea mailing list.

What is the purpose of that mailing list?  Is there any text in which
we urge people to use that list for some purpose?  If so, could you
show it to me?

    Toby and i agreed some days ago in that we need to implement a more
    structured work protocol for the task list, in order to keep track of
    each task history, and therefore making easy any later modification of
    it (for example, to mark a task as filled).

If we had infinite human resources, this might be nice to do.  But
since we don't, the question is, should we allocate our scarce
resources to changing this rather than to something else?

What problem are you trying to solve with this proposal?

    I am thinking in an organization similar to the evaluation group,
    maintaining a unique task list on CVS. The task list would be composed
    by several files, each one containing an outline and the history of a
    task. Each task would be identified by a timestamp. Then, tasks
    related request would be sent to address@hidden, where the person
    monitoring it would fill an initial outline for the task, assigning it
    a timestamp. Then, the person monitoring address@hidden would send the
    task outline to address@hidden, and update the CVS. A script
    would extract a texinfo file with all tasks and publish it on
    prep/tasks, periodically. Also, people from tasklist-idea would
    maintain the jobs and news of the tasklist savannah project synched
    with the list master copy.

That sounds like a lot of work to carry out.  I think it is a mistake.
We never have an excess of human resources.  We need to keep this
simple.

I think that one or two people looking at suggestions and deciding
whether to add them to the list is an adequate method.  Let's avoid
doing anything more labor-intensive than that.  Okay? 

This won't take up much time, so that various volunteers could do
something else instead of this.  For instance, we desperately need
more webmasters.