Re: [SED-ML-discuss] more schema updates

[Nicolas Le n. <le...@eb...>]

On 09/06/10 08:52, Lucian Smith wrote:
> * Nicolas Le novère<le...@eb...>  [2010-06-09 08:44] writes:
>> On 08/06/10 19:07, Frank Bergmann wrote:
>>
>>> I'm still missing the most important one ... the fix to the ChangeXML ...
>>> implementing adding of elements by having to replace an existing model
>>> element with itself and another one is just plain wrong!
>>
>> It is not a "fix", and replacing the containing element is not "plain
>> wrong". It is just verbose. But we are talking of computers here. verbosity
>> is not the main concern. Unambiguity is. We may want to separate the
>> semantics of the XPath for addition and removal. But not doing so does not
>> break SED-ML.
>
> I would say that semantically, the current method of adding a new element
> to a file is indeed wrong, because you must replace the element just
> before it with itself plus a new element.

That is not what we do. We replace:

<x>
   <y />
</x>

by

<x>
   <y />
   <z />
</x>

> What if you wanted to add an
> AssignmentRule to a model that had no AssignmentRules?  You replace the
> RateRules with the exact same set of RateRules, plus a new set of
> AssignmentRules?  That's just weird.  Adding 'add' is a clear semantic
> fix.

Yes. It is not weird. It is verbose. To bring you some ammunitions, it gets 
REALLY verbose if you do not have rules at all, because then you would have 
to replace the model element ...

OK, let's go for typing then: addition, deletion and replacement.


-- 
Nicolas LE NOVERE, Computational Neurobiology, EMBL-EBI, Wellcome-Trust
Genome Campus, Hinxton CB101SD UK, Mob:+447833147074, Tel:+441223494521
Fax:468,Skype:n.lenovere,AIM:nlenovere,MSN:nle...@ho...(NOT email)
http://www.ebi.ac.uk/~lenov/, http://www.ebi.ac.uk/compneur/, @lenovere