I Steven, I very much like the idea of log that keeps track of operations. Still in many cases we would need information on unit/segment level. But this is another story.. As Bryan, Arle, and I suggested elsewhere we should defer discussion of core/module status of particular features.
I suggest that you record the log facility as a feature in the section 2 on the wiki.
Rgds
dF
Dr. David Filip
=======================
LRC | CNGL | LT-Web | CSIS
University of Limerick, Ireland
telephone: +353-6120-2781
mobile: +353-86-049-34-68
facsimile: +353-6120-2734
mailto:
On Tue, Nov 8, 2011 at 01:23, Steven R Loomis <> wrote:
Hello,
We discussed this a little bit in IBM today. Our view would still be that segmentation does not need to be in core for interchange.
However, we had discussed a little bit about the idea of a logging facility which would give a list of what operations had occurred on a particular document as it processes through a workflow. If the logging facility stored, within the document, for each such operation:
* timestamp
* type of operation ( from a namespace )
* perhaps free format text with a description of the operation
For example, (and please disregard specifics of the following markup, it is only given for the rough concept)
<log>
<logEntry
timestamp="2011-11-08 01:16:07 UTC"
operation="org.oasis-open.xliff.segmentation">Segmentation was performed</logEntry>
<logEntry
timestamp="2011-12-08 01:16:07 UTC"
operation="com.example.someOtherOp.specialTranslation">Some other operation was performed</logEntry>
...
</log>
* XLIFF could administer a namespace containing items such as org.oasis-open.xliff.segmentation ( Java form, or it could be a URI such as with DTDs )
* Or, a company could use their own namespace (com.example for example).
* This way we could answer questions such as 'has segmentation occurred?' and 'where in the workflow (sequentially, according to the timestamp) did it occur?
* the operations referenced in the <log> would not need to be core or even part of the currently referenced version of xliff - as long as the namespace was maintained
Steven
