home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.text.tex
- Path: sparky!uunet!pipex!doc.ic.ac.uk!warwick!pavo.csi.cam.ac.uk!camcus!cet1
- From: cet1@cus.cam.ac.uk (C.E. Thompson)
- Subject: Re: Forcing expansions in Tex
- Message-ID: <1993Jan6.014020.6570@infodev.cam.ac.uk>
- Sender: news@infodev.cam.ac.uk (USENET news)
- Nntp-Posting-Host: apus.cus.cam.ac.uk
- Organization: U of Cambridge, England
- References: <MANDAL.93Jan5124200@rose.cis.ohio-state.edu> <EIJKHOUT.93Jan5155258@cupid.cs.utk.edu>
- Date: Wed, 6 Jan 1993 01:40:20 GMT
- Lines: 27
-
- In article <EIJKHOUT.93Jan5155258@cupid.cs.utk.edu>, eijkhout@cupid.cs.utk.edu
- (Victor Eijkhout) writes:
- |>
- |> No bug. The \uppercase command is *not expandable* in the
- |> technical sense, and that's the one that counts.
- |>
- [...]
- |>
- |> The behaviour of \uppercase is often counterintuitive.
- |> See a couple of cute examples in section 3.6.2 of the
- |> book by yours truly.)
-
- No argument about that. Does anyone understand the pre-history of this?
- In general, TeX82 does more things by expansion than TeX78 did: for example,
- conditionals were moved from the stomach to the mouth in TeX82 version 0.8
- (change #45 in tex82.bug). Why did \uppercase and \lowercase get left as
- mode-indepedent commands?
-
- There is a design constraint in TeX that expansion mustn't have certain
- side-effects (in particular, it mustn't alter anything in the |eqtb| array).
- (In METAFONT, by the way, there is no such constraint: expanding a token can
- cause almost anything to happen.) But there would seem to be no problem with
- \uppercase and \lowercase in this respect.
-
- Chris Thompson
- JANET: cet1@uk.ac.cam.phx
- Internet: cet1@phx.cam.ac.uk
-