[Codestriker-user] RE: Bug with TOC when using '-b' with diff

[John G. <jgo...@mo...>]

The "diff -b" without diffs seems to work, now.  It seems to not list
files with empty diffs.

Though we just noticed another problem with the TOC;  binary files that
differ.
Sample diff output, assume this is sandwhiched between other diffs.  The
TOC ends with the binary file.

-JT.
In this situation, the binary file was new...

Index: lib/my.jar
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
RCS file: my.jar
diff -N my.jar
Binary files /dev/null and my.jar differ


-----Original Message-----
From: David Sitsky [mailto:si...@us...]
Sent: Friday, May 17, 2002 2:13 PM
To: cod...@li...
Cc: John Goldstone
Subject: Re: Bug with TOC when using '-b' with diff


Should be fixed now in the latest CVS (as is the cc problem - although I

couldn't reproduce it here, but could see where the problem might be).
Also=20
fixed some rendering/anchoring bugs in popup parallel mode, here are the

CHANGELOG entries:

* Fix bug in popup parallel coloured-diff mode where the anchor names
  weren't being generated properly, resulting in some browsers not
  centering their view in the correct place.

* Fixed bug in pop parallel coloured-diff mode where changes right at
  the end of the file weren't rendered.

* Support for parsing files generated with cvs diff -b.

Cheers,
David

On Friday 17 May 2002 10:52, John Goldstone wrote:
> There is a bug with the way the diffs are parsed, if there is an empty
> diff (such as may be caused by uinsg the -b flag with diff), the TOC
is
> not properly created.
> Attached is diff showing the problem.
> This is output from "cvs diff -uNb" in the codestriker directory.
> Whitespace diffs were made to codestriker.pl, the others had
additional
> changes.
>
>
>
> -J.