> > > problems: > > > a) refers to freed memory (same as the previous report), and > > > b) doesn't repaint the display properly. Unlike the previous > > case, this time it doesn't repaint the parent properly - half > > the time. > > > I'm considering moving your fixed HTLine's (tmp_long_line) into the document > > structure, and seeing how your original patch does that way. From debugging, I > > can see that the problem with redisplay occurs with the second entry > > of that array. > > I am curious about the difference when fixed HTLine's are globals > vs HText members (in case of difference the color styles should also be > broken).
I did test with/without color style, and with/without caching. None of that seemed to make a difference. Though I noticed the problem with the color-styles since I use that more often, I configured my debugging disabling default-colors, enabling the debugging and leak-checking options. The contents of my lynx.cfg didn't seem to affect it either. (Debugging stuff like this goes very slowly even when I cut the screensize down). > The following proposed: 1) use ALLOC_IN_POOL and POOL_FREE as functions so > it will be easier to debug the stack; and 2) add one more new_line() into > HText_endAppend() so we switch self->last_line from the fixed one (if any) > into the one allocated in pool (extra bottom lines will be removed either > way). > > > > ; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED] > -- Thomas E. Dickey <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> http://dickey.his.com ftp://dickey.his.com ; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED]
