As you have found out, English text tends not to contain strings like \win 1\para, so adding such strings to your training text helps. Having "Language model adapts as you write" turned on and writing such strings teaches dasher about them. And another option is to try a higher value for "smoothing" (also on the "Language" tab in preferences). A higher "smoothing" value reduces the strength of the predictions, so low probability symbols such as the Paragraph symbol end up being allocated a higher probability than they deserve.
(Anyway, this isn't really a dasher bug as such, so in the strict "bug" sense, can be closed...) -- Newline (¶) only selectable after punctuation or space https://bugs.launchpad.net/bugs/579181 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
