--- Junio C Hamano <junkio@cox.net> wrote:
quoted text
> > I like the fact that the "data part" of blame is text, and
> > that the commit-8 is on the left, and color-chunked.  Sometimes
> > people simply _remember_ a number of commit-8's and thus the
> > layout of blame is intentional, since they can look to the left
> > and recognize a commit-8...
> 
> It is not only the initial commit.  A substantial rewrite and
> new development also has the same issue.
> 
> I think you are also contradicting yourself by saying that
> people would recognize a commit-8, and at the same time you do
> not like the chunk code that makes sure you do not get too few
> of them.  If people _do_ recognize commit-8 (I seriously doubt
> that), then wouldn't it help to make sure you have them on every
> couple-dozen lines so that the user would see the familiar one
> even when scrolled?

It is not that I don't like it.  For example if we didn't have
the block-per-commit-coloring, then we'd make use of this, but it
seems that the block-per-commit-coloring exists for the purpose to
show conglomerations of same-commit lines, thus obviating the need
to repeat it (commit-8) every so many lines.

The other question is how many lines should the repeat-chunk be?

In my case I'd like to set it to infinity, since the
block-per-commit-coloring gives me the same information.

The other extreme case is set it to 1, thus obviating the need to
use block-per-commit-coloring.

The middle ground as it seems to me, neither infinity nor 1, is
to just use the block-per-commit-coloring and use your idea of printing
the commit-8 only on the leading block row with mouse-over author
and date info. That's an excellent idea.*

    Luben
* I've three patches which implement your excellent idea but without
using a "stack-like" chunk, eliminating the concern of rare but present
files with 100s of 1000s of lines with only an initial commit.  Should I
post them?

-
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html