On Tue, 2008-11-04 at 11:41 -0600, Nicolas Williams wrote: > On Tue, Nov 04, 2008 at 06:30:55PM +0100, Mark Phalan wrote: > > > > On Tue, 2008-11-04 at 11:17 -0600, Nicolas Williams wrote: > > > On Tue, Nov 04, 2008 at 06:16:35PM +0100, Mark Phalan wrote: > > > > Can I take it that you've reviewed the changes and are happy with them > > > > (at least for the libtelca support)? > > > > > > Yes. > > > > Thanks. > > Did you look at the other (very minor) changes for 6763503 and 6704459 ? > > The changes for 6763503 look good. > > I think the changes to 6704459 look good, but you have to update the > suggested fix in the CR (it doesn't match the webrev).
Oops. I'll update. > > I think it would be very useful if assert() could log file names > relative to the workspace root, not absolute paths. That would probably > require a compiler change though, unless there exists a suitable > compiler option already; have you checked? I just had a quick look now in the man page for cc but nothing jumped out at me. I think it's a general rule not to ship with assert()s in production code - so if the assert()s with the full path are in DEBUG bits only I don't see it as much of a problem. I guess an option to the compiler to make the asserts relative to the ws root would be nice though. We probably need to bring this up on tools-discuss. Thanks for the review. -M
