 |
 |
Malcolm J. Currie wrote: > Yesterday I looked at all the logs that failed and there was little > consistency, to indicate a common failure, and nothing I saw explicitly > to do with the SC changes. However, from the timing and looking at the > CVS commit logs I suspected the new docs/sc/* component.xml.in files now > merged in the top-level componentset. [BTW somebody at York thought > that I needed to alter some other top-level/global file, if I understood > correctly, but I could not locate anything obvious and the build system > picked up the new docs fine. Please clarify.] > > I had looked at SSN/78 for guidance on special characters---it said > nothing---before I left the ampersands in. Upon further inspection > today I found a reference to SSN/70 that mentions it under Entity > References "which you cannot type on your keyboard". I think that > sentence needs rewriting. Note it's easy to miss the existence of > SSN/70 because this document is not built and installed such that we can > examine it with a browser. > See, http://www.w3schools.com/xml/xml_cdata.asp for XML and special characters. > What's the way to determine the _first_ error from the nightly logs? > By that I mean both what order are the elements built, and what to > search for in potentially very long log files? For applications and > libraries failure I go to the end of a log to see what's derailed. > Here that didn't help. > There is a log file in /home/vmwareshare/rhel30linux_i386/starlink on venus called cronout.txt that logs the nightly build and what is happening, but not the output from the software component builds, as they are published on the build web pages. > I fixed and committed the two component.xml.in files. > > Malcolm