- From: matt.halstead at auckland.ac.nz (Matt )
- Subject: [cellml-discussion] curation annotation framework
- Date: Thu, 12 Apr 2007 15:04:15 +1200
On 4/12/07, James Lawson <j.lawson at auckland.ac.nz> wrote:
>
So, for example in
>
> http://www.cellml.org/models/tentusscher_noble_noble_panfilov_2004_version01_variant01
>
> I'm guessing you have fixed the model to work in PCEnv (hence the star)
>
> but the model status still states "This is the original unchecked
>
> version of the model imported from the previous CellML model repository,
>
> 24-Jan-2006". Given there is still no curation annotation framework I
>
> think we still need to be using such a plain text description of the
>
> status of each model and thus you should update the documentation to
>
> reflect what you actually have done. This would also be the place to
>
> justify the use of a variant rather than version.
>
>
This particular model isn't one I've actually fixed - just one that I've
>
checked and annotated with a star to denote that it runs in PCEnv. As
>
such I don't know exactly where that version has come from - I can only
>
assume that it came from the old repository.
So this is a variant of a version that does not exist anywhere? It may
have been that someone thought the default for name for a new model
needed to include a variant too.
>
The presence of the star
>
means that it shouldn't say that it is unchecked, however. I'll change
>
that now. Of course the star system remains ambiguous because not all of
>
the models that lack stars don't run in PCEnv - they simply haven't been
>
checked, and not all of the models with stars have been fixed by myself
>
personally - I have just certified that they run in PCEnv.
In my opinion the stars should be removed for this very reason.
>
>
In the case that I have in fact created a new version, I have updated
>
the model status. Since there's currently no annotation framework as
>
such, and since I have really just started and am learning what needs to
>
be said etc as I go, some of the annotation I've put to models that I
>
fixed earlier isn't consistent with what I'm doing now. The model
>
documentation can't be changed from the page - the file needs to be
>
downloaded, changed and then reuploaded.
And that changes the version which is ok.
>
>
At the moment I'm primarily saying what version the current version was
>
updated from, by who, and when (if I know). Some of my earlier
>
annotations included the error that the previous version was producing.
>
This can produce a lot of text though, particularly when it requires
>
listing scores of variables that couldn't be defined etc. I do have all
>
this information (regarding how a particular model was fixed, that is,)
>
documented on my computer, however.
I think include it anyway, it's basically your commit message for the
new version.
>
>
Does anyone have any comments or proposals, formal or informal,
>
concerning what information needs to be included in the model status
>
documentation? The more consistent I can be now, the less I'll have to
>
go back and redo in the future.
I presume you are talking explicitly about the element <section
id="sec_status"> in the
http://cellml.org/tmp-documentation namespace?
e.g.:
<section id="sec_status">
<title>Model Status</title>
<para>
This is the original unchecked version of the model
imported from the previous
CellML model repository, 24-Jan-2006.
</para>
</section>
I would keep this brief as is, but those verbose changes that you
mentioned that described what the reason was for the change - i.e.
what it was fixing should use cellml metadata, specifically:
4.6 Modification History
The modification history lists changes that have been made to the
CellML document. Each separate alteration can be described in its own
<cmeta:modification> element. Each modification listed should also
include the name of the person who made the changes in a
<cmeta:modifier> element and the date the document was modified in the
Dublin Core date qualifier element, <dcterms:modified>. The definition
of modification history metadata is demonstrated in Figure 16
see
http://www.cellml.org/specifications/metadata/cellml_metadata_1.0#sec_general_metadata
cheers
Matt
>
>
James
>
_______________________________________________
>
cellml-discussion mailing list
>
cellml-discussion at cellml.org
>
http://www.cellml.org/mailman/listinfo/cellml-discussion
>
- [cellml-discussion] model variants(?): e.g. where multiple cell types are described by the same model in different files, (continued)
[cellml-discussion] model variants(?): e.g. where multiple celltypes are described by the same model in different files, David Nickerson, 04/12/2007
Archive powered by MHonArc 2.6.18.