From: "Bertram Ludaescher" <ludaesch AT ucdavis.edu>
Date: Wed, 13 Dec 2006 08:59:00 -0800
| Threading: | ↑ RE: [provenance-challenge] New provenance queries from Washington workshop from deelman AT isi.edu • This Message |
>>> On Wed, 13 Dec 2006 06:52:32 -0800 >>> "Bertram Ludaescher" <ludaesch AT ucdavis.edu> ↵ wrote: ... BL> If we just dump our internal RWS and COMAD provenance records on the BL> wiki site, however, it might not be useful for others (no effort was BL> spent on making the format interoperable, e.g. as XML -- so it might BL> be hard for PC-2 folks to do sth with it). I was just reminded about the fact that COMAD provenance info is available in XML. So we can get that part immediately. However, COMAD uses a different model of computation (MoC) than other approaches, so the fact that XML is used probably won't help as much as one might hope. Also COMAD by default embeds provenance info in the data stream. Externalization should be possible though. RWS on the other hand is using a MoC (dataflow process network) that is a bit closer to the standard DAG model (in fact generalizing it) and thus might be easier to integrate. Generating XML from the current internal syntax (Prolog) should be easy. Bertram