On Sat, May 27, 2006 at 12:22:47PM -0600, Jeremy Huntwork wrote: > > I believe that, previously, Archaic was creating these build logs with > his personal scripts, which up to now has been sufficient.
Not just previously. Still. > Now, however, we have perhaps a opportunity to increase the usefulness > of these logs. With jhalfs, which appears to be being used by a variety > of people, especially our devs, we should be able to easily glean a > consistent style of build logs from a number of people. I'd like to > suggest that we make available a means to upload (correctly used) jhalfs > build logs for view by people following the book. We can sort them first > by book version, then by cpu class. This would likely help us see at a > quicker pace, the differences/similarities between different builds. As > SVN progresses, old svn logs can be rotated out. I would think intel/amd is sufficient if even that is necessary. It would be easy to go overboard on this, though. As for jhalfs versus my logs, if the test output was separated from the rest, then it is feasible, other wise we are talking about several megabytes of log files (just the check output is around 5 MB for the 6.1.1 release). I don't see the actual benefit though because it's still just a simple make check > logfile 2&>1. i.e. nothing that requires a special tool, but if the code to separate out check logs was indeed added to jhalfs, then we could include others' logs. Some way of verifying that the builds were strictly by the book would be required, though, and I always use the previous release built strictly by the book as a standard host system so that even chapter 5 logs have something to be based on. -- Archaic Want control, education, and security from your operating system? Hardened Linux From Scratch http://www.linuxfromscratch.org/hlfs -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
