[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn cat (was: Re: svn up -p)

From: Benjamin Pflugmann <benjamin-svn-dev_at_pflugmann.de>
Date: 2002-11-06 21:47:57 CET

Hello.

On Wed 2002-11-06 at 20:59:25 +0100, Branko ??ibej wrote:
[...]
> > >Another thought that's flashed through my mind is that
> > >the soon-to-come "blame" is very similar to this
> > >functionality.
> >
> > Funny. While reading the thread, I just came up with the
> > same idea. So, I second that.
> >
> These are very strange and disturbing associations.
>
> "cat" would fetch a file and pipe it to the standard output. "blame"
> would print a line-by-line annotation, but only for text files.

And? There are a lot of UNIX command line tools that only work for
text files (reasonably). E.g. the --number-nonblank, --show-ends,
--number, --squeeze-blank options of (GNU) cat only works for text
files, either.

The fact that currently diffs (that is what blame boils down to) are
hard to do for non-text files is not a failure in the association.

> I fail to see how those are even remotely similar, except in that
> they write something to stdout.

Oh, they have more in common. They do not only write something to
stdout. That "something" is the same, indeed: The content of the
file. blame/cat-v is just more verbose by printing out some metadata,
additionally. For the idea it is irrelevant that this metadata may be
hard to generate.

Or the other way around: cat makes no sense for binary data, except
for redirecting it into some file or other program. The terminal is
not able to handle binary data, e.g. a MS Word document. You could
consider the terminal to be an implicit display-tool for text. And we
have a text-oriented diff. Having both, we can generate blame/cat-v
output for text, already.

In theory, there is nothing that prevents implementing a diff for MS
Word documents and therefore implementing blame/cat-v by generating a
new MS Word doc at the end (the metadata being inserting into the
document itself). MS Word or a Word viewer could be the display tool.

I do not mean to really go that route by any way. I just wanted to
show, that IMHO the concept is not that strange. Currently, we "just"
lack the tools to have blame/cat-v work on other formats.

Maybe the different opinions come from why we use blame. I use cvs
annotate just as what I described, a more informative listing of the
file at hand. You seem to me more concerned about the metadata or the
task you try to solve with it. Or I did I misunderstand you?

Regards,

        Benjamin.

  • application/pgp-signature attachment: stored
Received on Wed Nov 6 21:48:55 2002

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.