Meeting Minutes 30 May 2007
Meeting Minutes 30 May 2007
Present: Sarala Dissanayake, Poul Nielsen, Jams Lawson, Matt Halstead, Catherine Lloyd, Randall Britten, Tommy Yu, Andrew Miller, Michael
Andrew has been working on fixing bugs prior to the upcoming release. He has classified all the feature bugs as deferred, and has put fixed the remaining bugs (except those that appeared late last week).
Tommy has finished the modification history editor, and has take a look at Michael's site tools, which can build the dependencies for the CellML site.
Over the next week, he plans to work on filtering the main listing, so that it will be based on the keywords in the RDF metadata. Currently, only some of the models actually have keywords in them.
Catherine noted that the current count of CellML models includes every version, which is not ideal. Tommy will look into how to adjust the count to not include versions. Catherine noted, however, that some publications which have multiple models are currently put in as multiple versions.
James noted that there is a need to move much of the data which is in XML comments into cmeta comments, which will work with Tommy's planned changes to display data from this.
Catherine asked about the uploading of images onto the CellML site. At present, it needs to be put into a flat directory. In the future, this sort of data should be stored alongside the model. Matt suggested that this should be done by making the top-level item in CellML a folder.
Matt advocates letting model submitters control how the models are arranged in the top-level folder. Andrew thinks that the top-level should have some structure imposed on it (perhaps limited to UUID, with the front-end providing aliases), although the structure within a model folder should be controlled by the authors. However, he thinks there should at least be a manifest file or similar describing how to run a simulation for a model and so on.
There was also some discussion about the details of how we could migrate to using Subversion or a similar backend and whether the system could and should be distributed. Matt agreed to write a preliminary discussion document based on the discussions at the meeting.
Andrew has been working on fixing bugs prior to the upcoming release. He has classified all the feature bugs as deferred, and has put fixed the remaining bugs (except those that appeared late last week).
Tommy has finished the modification history editor, and has take a look at Michael's site tools, which can build the dependencies for the CellML site.
Over the next week, he plans to work on filtering the main listing, so that it will be based on the keywords in the RDF metadata. Currently, only some of the models actually have keywords in them.
Catherine noted that the current count of CellML models includes every version, which is not ideal. Tommy will look into how to adjust the count to not include versions. Catherine noted, however, that some publications which have multiple models are currently put in as multiple versions.
James noted that there is a need to move much of the data which is in XML comments into cmeta comments, which will work with Tommy's planned changes to display data from this.
Catherine asked about the uploading of images onto the CellML site. At present, it needs to be put into a flat directory. In the future, this sort of data should be stored alongside the model. Matt suggested that this should be done by making the top-level item in CellML a folder.
Matt advocates letting model submitters control how the models are arranged in the top-level folder. Andrew thinks that the top-level should have some structure imposed on it (perhaps limited to UUID, with the front-end providing aliases), although the structure within a model folder should be controlled by the authors. However, he thinks there should at least be a manifest file or similar describing how to run a simulation for a model and so on.
There was also some discussion about the details of how we could migrate to using Subversion or a similar backend and whether the system could and should be distributed. Matt agreed to write a preliminary discussion document based on the discussions at the meeting.