From: Fredrik L. <Fre...@im...> - 2008-03-27 15:44:13
|
Hi Lennart, It looks very good, chromatograms are very welcome for SRMs and SIMs etc. As Matt writes, this brings up the question about ids again. I'd vote for making all ids (except the mzML id) data type xs:ID, even if this means that they have to be unique within the file (and that they cannot contain spaces etc), but it is very good for validation of uniqueness. References that are always within the file should be of type xs:IDREF to make sure that the referenced item exist by simple xsd validation. Or is there a good reason for permitting a chromatogram to have the same id as a spectrum or something else? Fredrik |