David Nickerson wrote: > Hi Andrew, > > I have just starting looking at using git and checking your current > draft of the specification. I have made a few changes and attach the > patch (generated with git diff -p > andre.patch). Not really sure the > best way to do things in order to share changes - I guess if there are That way is probably okay for small changes. I believe git has ways to e-mail patches while keeping the change identifier - that might make it easier to merge back and forwards in the future, because if I commit your change and you update, it would be good if there was no need to re-merge that patch. I have never actually tried git's e-mail features, although I believe that they are used extensively for Linux kernel development.
> going to be many its worth signing up at the same place hosting your > draft? but then I'll probably not have to much time to devote to this... > > Anyway, about the only significant change I made was to turn all the > sect1 and sect2's into section's - while this is probably more a > personal preference, it seems to be a widely used one. Hardcoding > section levels just seems a bad thing to be doing to me, especially when > you're using XInclude to include multiple documents into a single > document....although maybe you are trying to force those sections to a > specific depth regardless of where they are imported into? Thanks, using section seems to make sense. > > And then a couple of other things I changed that we can maybe discuss > more on cellml-discussion. You are using 'CellML File' as the base unit > whereas I generally think of them as documents - especially in the > context of generated data which may never exist in an actual file. > Again, just my preference :) I guess the only issue is that file implies it is always a single XML file, while document could potentially be more than one file. > > And the second point, which I'm sure has been discussed in the past but > I can't quite recall is the statement that CellML processing software > should not attempt to process documents containing elements in the > CellML 1.0 namespace. Perhaps you could just refresh my memory as to > where that came up before? The idea is that CellML 1.0 and CellML 1.1 namespaces shouldn't be mixed. CellML 1.x (x > 1) processing software can also process CellmL 1.0 files, but it must do so in accordance with the CellML 1.0 specification, and not in accordance with any of the future specifications, because CellML 1.0 documents are not valid under those later specifications. The annotated specification will probably need to explain this, but that explanation is not necessary to define CellML 1.2 and so not really appropriate in a normative-only version of the document. Best regards, Andrew > > > Thanks, > Andre. > _______________________________________________ cellml-discussion mailing list [email protected] http://www.cellml.org/mailman/listinfo/cellml-discussion
