X-VM-v5-Data: ([nil nil nil nil nil nil nil t nil]
	[nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil])
Date:         Sun, 25 Feb 90 22:58:00 PST
Reply-To: LaTeX-L Mailing list <LATEX-L%DHDURZ1@unido.bitnet>
Sender: LaTeX-L Mailing list <LATEX-L%DHDURZ1@unido.bitnet>
From: DHOSEK@HMCVAX.bitnet
Subject:      Additional thoughts on verification
To: Rainer Schoepf <BK4@DHDURZ1.uucp>
Status: R
X-Status: 
X-Keywords:                  
X-UID: 30

Hmm, the more I think about it, the less of a problem
verification seems to be:

Verification problem 1: The kernel.
  For this, we should create a file latrip.tex and latrip.sty
  which between them at least attempt to use every piece of code
  in the latex.tex, lfonts.tex and lplain.tex files. One thing
  that is important and has been absent in previous version
  numbering of LaTeX: the version number has only changed when
  LaTeX's _user_ interface has changed. It should also change
  when the _style_ interface changes. Perhaps something like
  appending a letter to the version number indicating what
  version the style interface is at, e.g., (note that the future
  history is just for expository purposes)

     LaTeX 2.09a   where we are now.
     LaTeX 2.09b   LaTeX with a new \@startsection command, say
     LaTeX 2.10b   some new user commands have been added (note
                   that it's still b)
     LaTeX 2.10c   New version of list environment.
     LaTeX 2.11d   Additional users commands that affect style
                   files.

  make sense?

Verification problem 2: the styles. Or perhaps, more accurately,
  the meta-styles. We could have arttrip.tex, reptrip.tex,
  bktrip.tex, lettrip.tex, memtrip.tex, ... This should work with
  any style in its class, e.g., I could run arttrip.tex with
  documentstyles of article, ltugbot, physrev, etc. reptrip.tex
  with report, hmcthesis, hmcmathclinic, hmcengclinic, etc.

Are there any other classes of verification problems that I've
missed?