Re: Regression testing of tetex

[Junichi Uekawa <dancer@netfort.gr.jp>]

Hi,

> >> The process that cannot really be automated is:
> >>
> >> 3. making sure that the resulting text is same, or reasonable.
> >
> > Even this could be done in principle.  When the commands
> >
> > \pdfinfo{/CreationDate (2004-06-13)} % an arbitrary date
> > \year=1980 \month=9 \day=9 \time=10
> > \pdfcompresslevel0
> >
> > are included in a TeX document, pdf(la)TeX will produce identical,
> > grep'able files with different runs.  
> 
> diff'able is what I really meant.

This is a great news; it allows for some workable 'there should be a
change only when something should be changed' kind of workflow.

Certainly, some sustainable way of scripting the test is really
required.


regards,
	junichi
-- 
dancer@{debian.org,netfort.gr.jp}   Debian Project