David Nickerson wrote:
> Hi Tommy,
> 
> That looks good - its all starting to make sense to me now.
> 
> I'm just wondering how your system would handle a case where two authors 
> independently encode the same published model. The first author to 
> upload their encoding would get "ownership" of the publication alias (if 
> I have the terminology right). Is there any way for the second author to 
> get a similar alias to their encoding of the model? This is starting to 
> sound like a version/variant theme, but its probably a situation that 
> will crop up quite frequently...

I guess it depends on how those two model creators work.  If John and Mary work 
independently and two different models describing the same item were created, 
each will need have separate project directories.  If John did get the 
publication alias set up first it would obviously point to his model, but now 
Mary comes along and wants to have a separate model up also.  What could happen 
is this:

1) Publication alias is no longer an alias, but a directory holding aliases to 
users' models.
2) New model directory is created.  John and Mary's model directory is copied 
into there.

While outcome is similar, 1) separates publications from models a lot more, may 
reflect this situation when a paper with multiple models with each created by 
different people:

John, Mary and Ming creates on models a, b, c based on Doe's paper.  All three 
gets approved, and repos://publication/doe_2007_1/ is created containing 

repos://publication/doe_2007_1/pathway_a -> repos://!rev/45/home/john/a
repos://publication/doe_2007_1/pathway_b -> repos://!rev/60/home/mary/b
repos://publication/doe_2007_1/pathway_c -> repos://!rev/54/home/ming/c

created by their respective creators.  Each published model is treated 
differently, note their revision numbers.

2) has the benefit of encouraging model creators to work together, groups the 
same models in one place, and may reflect this situation:

A publication that describes multiple models with different people coding up 
each one could have a shared UUID named workspace, owned by the people working 
on it, with each separate models in its own directory.  The publication alias 
could be owned by the whole group that worked on the model.

I just flushed this out of my head, both of these suggestions may have very 
interesting consequences that is not noted here.

This was a very good question.

> 
> This is a slightly different example from your example workflow and 
> could be viewed as John and Mary both having "valid and correct" but 
> different encodings of the doe_2007_1 paper. Actually, I just saw the 
> '_1' on the publication link - is that some kind of version/variant that 
> would be _2 for Mary in my example? I had been assuming the 2007_1 meant 
> January 2007.
> 

It could conceivably mean the first paper John Doe published in 2007, or 
January, as that haven't been decided yet.

Thanks,
Tommy.

> 
> Thanks,
> Andre.
> 
> Tommy Yu wrote:
>> Hi,
>>
>> I thought Andrew's ideas here is worth expanding, and I wrote a page based 
>> on that.
>>
>> http://www.cellml.org/Members/tommy/BaseRepository
>>
>> Cheers,
>> Tommy.
> 
> _______________________________________________
> cellml-discussion mailing list
> [email protected]
> http://www.cellml.org/mailman/listinfo/cellml-discussion

_______________________________________________
cellml-discussion mailing list
[email protected]
http://www.cellml.org/mailman/listinfo/cellml-discussion

Reply via email to