Chronological Thread 

< Chronological >      < Thread >

---

On Thu, Oct 9, 2008 at 8:55 AM, James Lawson <j.lawson at auckland.ac.nz>
wrote:
> It appears nobody has commented on this directly within cellml-discussion. I
> understand you hold some opposition to finalising what you believe to be a
> draft specification, Andre. Could you please elaborate?

I just think that because the current draft has been clearly marked as
a draft that it is not unreasonable to correct known deficiencies
prior to releasing the a final 1.0 specification. I'm also a bit wary
of setting a precedent of "finalising" a specification with known
problems...

I guess I would prefer to see the current 1.0 draft not be finalised,
at least until a draft of the next version of the specification is
available. Then the status of the 1.0 draft would become more of a
historical reference which clearly directs users to a specification
which is much more useful.

> The current consensus (for reasons Andrew mentions below) is that we should
> proceed with this with a view to creating the next version of the metadata
> specification as soon as possible so we can fix the problems with the
> current spec. Peter Hunter's proposed for this work timeframe is by April
> 2009, in time for the planned CellML / SBO / MIASE, BioPAX combined
> workshop.

As I say above, I think work should begin on the next version without
waiting to release the 1.0 specification. People citing or adopting a
draft specification do so knowing that it is likely to change, and in
fact I have had reviewers ask for more detail to be specified when
referencing draft specifications in order to allow readers to be clear
what is "standard" and what is draft, and what version drafts were
used in the work to ensure readers are able to find the appropriate
specifications.

I think the proposed timeframe is reasonable to get the next version
of the specification, at least in draft form, and then at that point
the 1.0 specification could be finalised with the link to the new
(1.1?) specification draft.

Having said that, if releasing the current draft as-is is the only way
to make progress on the next version, then I have no strenuous
objections to doing so as the actual defects in the specification are
quite minor. If this approach is taken, then I would like to see the
status section of the specification note this and link to appropriate
tracker item(s) detailing the work being done to define a more useful
metadata specification.


Andre.

---

• [cellml-discussion] Finalising CellML Metadata 1.0 specification, Andrew Miller, 10/01/2008
  • [cellml-discussion] Finalising CellML Metadata 1.0 specification, James Lawson, 10/09/2008
    • [cellml-discussion] Finalising CellML Metadata 1.0 specification, David Nickerson, 10/09/2008
      • [cellml-discussion] Finalising CellML Metadata 1.0 specification, James Lawson, 10/10/2008

---

Archive powered by MHonArc 2.6.18.