On Thu, Jul 27, 2017 at 11:27 AM, Junio C Hamano <gitster@xxxxxxxxx> wrote: > Stefan Beller <sbeller@xxxxxxxxxx> writes: > >> Well, we could also try a "zebra" as in sb/diff-color-move to show blocks >> with the same fancy border detection. > > Luckily, blame output has places for metainfo on each line, unlike > diff output, so there won't be a need for painting border > differently. Right. > For example, if we decide that metainfo is shown only > for the first line for each block, then by definition, the border > between blocks is just before a line with metainfo and nowhere else. > So in that sense, the problem should be a lot easier to solve ;-) Yes, also the code already provides structured blocks, so internally it is also easier. The question is whether we decide that showing the first line is a good choice. We could also put it at the approximated middle and draw lines up and down. In an advanced world, we would not use a dumb pager, but e.g. ncurses, such that the line displaying meta information may come along when you scroll down a page. Maybe the discussion here can also feed back into the "machine readable move coloring" that Aevar seemed to be interested in. > > Unluckily, a block may span several pages, so "only at the first > line" might not be very useful. I wonder if we can print in black > ink on black background and let selection by mouse or in screen > still reveal useful information? This is what the dark color approximates, except you don't need a mouse selection to read?
