Hmm scrap that, haven't read the other 'Concerning the CellML repository' emails yet. Tommy just said that variants are getting scrapped...
James Lawson wrote: > My $0.02 on this is (please forgive me if I get some of the technical > stuff mixed up): > > The current naming scheme is as it translates to the web address is: > author(s)_date_versionXX_variantXX > > I think it should be author(s)_date_variantXX_versionXX instead, since > IMO, one should be thinking in terms of versions of variants, rather > than variants of versions. > > Also, I think that if there were perhaps some metadata that could > pertain to what version and variant a cellml file is, and also some > 'sub'metadata under variant to say what the variant represents, whether > it's a particular cell type or what. > > I realise that metadata isn't supposed to be added to a model for the > sake of a repository or for any non-generalised purpose, but I think > that version/variant metadata would be useful. > E.g. for 1.1 models, a simulator could pick this metadata up. So you > could bring up window in which the software could tell you that, for > example, you are embedding this version of this markov model of an > L-type Ca++ channel, by such and such et al., into a variant 02 - > "epicardial cell" Pandit et al. cardiac cell model, etc. etc. > Another example would be working with CellML 1.1 models in an era where > we have a library of components that people can use. We might have a > GPCR component which has a large number of variants, and it would be > crucial for the simulation/editing programs like PCEnv to know, and be > able to tell the user, which version and variant of each component they > are using. People might want to swap in different variants to see how if > affects their model etc. > > And of course this version/variant metadata would obviously be highly > useful (IMO) for the repository. Maybe subversion could automatically > write this metadata. > > What I'm really trying to say is that I think there is justification > for version/variant information to be stored in metadata as well as the > URI naming scheme, since, unless I'm missing something, there is useful > information (both for repositories and simulator software) that can't be > stored in the URI. > > James > >>> - Version/Variant >>> It already clogged up the system. There is no proper revision control >>> mechanism, what we have now is an ad-hoc emulated system. >> I don't think it has clogged the system I just think it has been >> improperly used both by authors and by the user interface. > > Ideally the users and authors shouldn't be presented the option to make > mistakes like this, should they? Most people, I would imagine, don't > care about the versions of a model unless they are actually working > on/with it. > > > This is no >> fault of the authors, there is simply a specification for versioning >> that is missing. The hope is that subversion applies well to this. > > > _______________________________________________ > cellml-discussion mailing list > cellml-discussion@cellml.org > http://www.cellml.org/mailman/listinfo/cellml-discussion _______________________________________________ cellml-discussion mailing list cellml-discussion@cellml.org http://www.cellml.org/mailman/listinfo/cellml-discussion
