James Carlson <james.d.carlson at sun.com> writes: > I've (finally!) finished testing of my fix for this issue: > > 440 Need exactly one implementation of scm sniffing. > > This was originally a code review comment from Bill Sommerfeld. The > webrev for this change is here: > > http://cr.opensolaris.org/~carlsonj/webrev-440/ > > Testing consisted of: > > - Full builds on x86 and PSARC (naturally). > > - Manual testing of each of the modified binaries (xref, which_scm, > sdrop, nightly, webrev, ws, xref) in hg and Teamware workspaces.
Subversion? (I know, that's a pain, since that means building Companion somehow) > - Manual testing of modified binaries with CVS (quagga) workspace, > RCS, and plain SCCS. > > - Netinstalled an AMD64 box, installed the SUNWonbld package, and > did full builds of onnv with an hg parent, and a Teamware parent. > (The teamware one did a bringover; hg is harder to automate > because of the closed binary nonsense.) I have an env file that uses Bill's hooks work to fetch closed bins too, but yes, it's flaky and fragile. That said, in your position I'd bringover the closed tree out of Hg too, and do a full build with bringover, if that was your goal. (/net/onnv.eng/export/gate-hg/usr/closed) I don't really mind if you choose to run again, and do that, but it'd be nice to know it worked properly, I have no idea how many people even know where the closed side of the bridge is, never mind have used it. -- Rich
