On Mon, Jul 18, 2011 at 12:04:40PM +0200, Siarzhuk Zharski wrote: > On Sun, 17 Jul 2011 22:59:48 -0700, pete.goodeve@xxxxxxxxxxxx wrote: > >>> /BFS_User/haiku/haiku/src/kits/tracker/ContainerWindow.cpp:159: > >>> error: 'class BView' has no member named 'Wi' > >>> [....] > >>> ... everything seeming to follow from that "null character". > >>> However, I scanned the file for a null, even did od -c and looked > >>at the line in question. There *is* no null character! > >> > >>did you check line 159 > >Yes, of course. It is just: > > && view->Window()) > > I observe this usually when occasionally put cyrillic characters into > the source code. Which ASCII code has this "n" character? It should be > 0x6e. You can check this line in DiskProbe. > Hmmm -- when I look with DiskProbe, I see the exact point in question shows up as a 'line-break' between offset 0x1af ("i") and 0x1b0 ("n") in block 8 of the file. Coincidence...? Otherwise there is certainly no questionable character there. (The "n" is 0x6e as it should be.) The whole file doesn't contain anything but basic ASCII -- standard 7-bit printing chars, newlines, and tabs. I think it's a gcc4 glitch of some kind, but hard to prove. What's the current version? I'm using gcc-4.4.4-haiku-100705, which seems to be dated 2010-07-08. Maybe I need to update? -- Pete --