You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Karl Fogel <kf...@newton.ch.collab.net> on 2003/02/05 23:30:07 UTC

Re: sander's diff library

Ben Collins-Sussman <su...@collab.net> writes:
> Sander -- I'd like to propose that you integrate your code into trunk
> permanently.  Create a new cmdline switch (--difflib?) that causes the
> commandline client to to use the internal diff library instead of
> calling out to GNU diff/diff3.  Testing your code then becomes a
> runtime issue, rather than a compiletime one.  I think that will
> significantly lower the bar for testing.
> 
> Eventually, the --difflib switch can be deleted.  But it would great
> for now.  (IIRC, this is method by which we first started getting i18n
> code exercised.)

+1  Make it easy to test, and it will be tested :-)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: sander's diff library

Posted by Florin Iucha <fl...@iucha.net>.
On Wed, Feb 05, 2003 at 05:30:07PM -0600, Karl Fogel wrote:
> Ben Collins-Sussman <su...@collab.net> writes:
> > Sander -- I'd like to propose that you integrate your code into trunk
> > permanently.  Create a new cmdline switch (--difflib?) that causes the
> > commandline client to to use the internal diff library instead of
> > calling out to GNU diff/diff3.  Testing your code then becomes a
> > runtime issue, rather than a compiletime one.  I think that will
> > significantly lower the bar for testing.
> > 
> > Eventually, the --difflib switch can be deleted.  But it would great
> > for now.  (IIRC, this is method by which we first started getting i18n
> > code exercised.)
> 
> +1  Make it easy to test, and it will be tested :-)

I think it would be a good idea to write a small utitily to take an
existing repository A, get through each revision and copy it onto B,
then go through both repositories and compare them. Then the guys with
big repositories might test this easily and without risking production
data.

florin

-- 

"NT is to UNIX what a doughnut is to a particle accelerator."