[rfc] dumb terminal, progress bar and 'read knit index: ..'

Martin Pool mbp at canonical.com
Thu Dec 21 06:40:09 GMT 2006


On 30 Nov 2006, John Arbash Meinel <john at arbash-meinel.com> wrote:
> Alexander Belchenko wrote:
> > $ bzr diff >1.diff 2>stderr
> > $ cat stderr
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > read knit index: ..
> > 
> > I often launch bzr commands from inside FTE editor and every time I see
> > this useless messages.
> > 
> > Does 'read knit index: ..' is really needed?

I agree we should get rid of it, or at least make it display something
more interesting.  I see this too in vi.  Particularly when we're on a
dumb terminal (and so not able to clear the line, and possibly with no
user watching) I would lean towards just blocking without showing
anything.

I think the test progress is now much better in this case.

Also we should aim for progress messages which are meaningful to the
user, eg it would be much better to show 

[4/123] diff bzrlib/delta.py

> Not really, Robert pushed for it because it could be slow over sftp. In
> my branch which refactors the progress bars, I took it out.

Is that still live, or in progress, or ...?
 
> But after my changes, I think I have much more informative stuff about
> what is going on, and 'read knit index' is just noise.

I'll reply


-- 
Martin




More information about the bazaar mailing list