Text in headlines does not wrap, as it does in NS4.6. A table problem?

Troy, the block frame (at the bottom of the debug output) is getting reflowed with reason=resize and returning a max element size too large (23865). This may not fix the bug, but needs to be fixed before I can proceed. Please reassign to me when this is fixed. TO::Rfl en 01A1D3F0 rea=0 av=(8880,UC) comp=(9000,UC) T::Rfl en 01A1D350 rea=0 av=(9000,UC) comp=(9000,UC) TRG::Rfl 01A1D090 rea=0 av=(UC,UC) comp=(UC,UC) TR::Rfl en 01A1E840 rea=0 av=(UC,UC) comp=(UC,UC) TC::Rfl 01A1E2B0 rea=0 av=(UC,UC) comp=(UC,UC) Area::Rfl en 01A1E220 rea=0 av=(UC,UC) comp=(UC,UC) TO::Rfl en 01A1F3A0 rea=0 av=(UC,UC) comp=(UC,UC) T::Rfl en 01A1F300 rea=0 av=(UC,UC) comp=(UC,UC) TRG::Rfl 01A20B70 rea=0 av=(UC,UC) comp=(UC,UC) TR::Rfl en 01A20610 rea=0 av=(UC,UC) comp=(UC,UC) TC::Rfl 01A20080 rea=0 av=(UC,UC) comp=(UC,UC) Area::Rfl en 01A21F90 rea=0 av=(UC,UC) comp=(UC,UC) Area::Rfl ex 01A21F90 des=(23895,600) maxElem=(1050,315) TC::Rfl ex 01A20080 des=(23925,630) maxElem=(1080,345) TR::Rfl ex 01A20610 des=(23925,630) maxElem=(1080,345) TR::Rfl en 01A22A00 rea=0 av=(UC,UC) comp=(UC,UC) TC::Rfl 01A22470 rea=0 av=(UC,UC) comp=(UC,UC) Area::Rfl en 01A223E0 rea=0 av=(UC,UC) comp=(UC,UC) InstantiateEmbededPlugin for application/x-java-vm Area::Rfl ex 01A223E0 des=(4620,675) maxElem=(4620,675) TC::Rfl ex 01A22470 des=(4650,705) maxElem=(4650,705) TR::Rfl ex 01A22A00 des=(4650,705) maxElem=(4650,705) TRG::Rfl ex 01A20B70 des=(UC,1335) maxElem=(4650,345) TRG::Rfl 01A20B70 rea=2 av=(23985,UC) comp=(23985,UC) TR::Rfl en 01A20610 rea=2 av=(23985,UC) comp=(23985,UC) TC::Rfl 01A20080 rea=2 av=(23925,UC) comp=(23895,UC) Area::Rfl en 01A21F90 rea=2 av=(23895,UC) comp=(23895,UC) Area::Rfl ex 01A21F90 des=(23895,600) maxElem=(23865,315)

Troy, in my tree I have changed nsTableFrame so that during a pass2 reflow the max element size is set to nsnull (since we already have it from pass1). After doing that, I think the real issue with this bug surfaced. The block frame in the 1st cell of the nested table contains a line that is too long (from dump frames). I looked at the reflow debug output and the last time the block frame was reflowed it returned a desired size that was too big (see below). I'm planning to check that change in tonight, but I think the problem reported in my last comment will eventually bite us at some point, so maybe it needs to be fixed before getting my changes. I'm not sure if the problem I'm reporting here (after the change) is yours or Kipp's. TC::Rfl 01A21F10 rea=2 av=(8850,UC) comp=(8820,UC) Area::Rfl en 01A21E80 rea=2 av=(8820,UC) comp=(8820,UC) Area::Rfl ex 01A21E80 des=(23895,600) TC::Rfl ex 01A21F10 des=(8850,630)

Note: If you eliminate the HR element the test reflows properly. Somehow the springy hr is mucking up the works...

And I seem to remember that removing the applet results in a correct reflow as well.

Interesting. Didn't we had problems with springy HRs in the past?

This seems to be fixed now.

Looks okay to me, too. Marking FIXED because one of the many recent changes must have fixed this

Verified fixed.