getComputedTextLength always returns 0 with tspan elements. FF3 does not.
Created attachment 21997 [details] Example Should alert the same number twice. Instead will be 42 (or so) and then 0.
Yup. Looks like a real bug. Thanks for the simple example.
Can confirm this bug. We use the tspan.getComputedTextLength() function for a manual word wrap of a label. Due to the returned 0 our word wrap does not work at all in Webkit. This is a major issue for us and we'd appreciate it to be fixed somewhere in the near future...
Created attachment 88380 [details] Testcase Bug is resolved in trunk since a while, we forgot to close this bug report. Nevertheless, I've modified the attached testcase to be a real layout test, and I'd like to land it. Dirk, can you check?
Comment on attachment 88380 [details] Testcase r=me
Comment on attachment 88380 [details] Testcase Clearing flags on attachment: 88380 Committed r83035: <http://trac.webkit.org/changeset/83035>
All reviewed patches have been landed. Closing bug.