DataMuseum.dk

top - metrics - download

    Length: 2494 (0x9be)
    Types: TextFile
    Names: »READ_ME«

└─⟦060c9c824⟧ Bits:30007080 DKUUG TeX 2/12/89
    └─⟦this⟧ »./tex82/ctex/trip/READ_ME« 

TextFile

The SAVE*.* files in this directory represent an actual run
of "make test" on an Ultrix (BSD4.2) system.  There are
only trivial operating system and date differences between
them and the archive files on SCORE.STANFORD.EDU
For purists, the SCORE_trip.typ file is provided so
that you can see what a 36-bit machine made of the test

In order to make the full set of tests under the makefile,
it is necessary to redirect the inputs, which come from
Unix_tripin?  
These inputs will not show up in the *.fot files, but they
are correct according to the more reliable *.log files.

The two runs of triptex are tee'ed so that you can see the
output on the screen as well as on the *.fot files.

You should really read the trip manual (tripman.tex), but here's a short
course on how to run the trip test.  NB You will get LOTS of error
messages!

First, make triptex.  This can be done by moving back to the base directory
(./tex82) in this distribution and typing 

        make default

If you do not already have recent versions of dvitype and pltotf, 
make dvitype and pltotf in the ./tex82/texware directory

Then come back to this directory and type

        make test

The trip.log file should be identical to either SAVEbig_trip.log or
SAVEtrip.log.  Big_trip.log was produced with the BIGTEX.PATCH applied,
and changing the type glueratio from float to double (in "site.h").
(If you need to use either SAVEbig_trip.log or SAVEbig_trip.typ, 
you will have to uncompress them.)

SAVEtrip.log was produced using the unmodified ctex.ch and site.h
files.  For reference, trip.log.knuth is the output Knuth obtained on
a 36-bit machine with 36-bit floats (in Pascal).  It will give you an
idea of what kind of differences are "legal".  Glue ratios are the
only thing TeX does with floating point arithmetic, so they are the
only calculations whose outcomes may vary from machine to machine;
usually, this variance is only in the last decimal place or two.

The output on your terminal when running the second phase of the test
should match trip.fot.

Dvitype will produce a file "dvitype.out", which the makefile
will rename to trip.typ.  This file should be identical
to either big_trip.typ or small_trip.typ, which were produced under
the same conditions as big_trip.log and small_trip.log.

For reference, "SCOREtrip.typ" is the equivalent file produced by Knuth
on a 36-bit machine.  For an explanation of how different your file
can be from trip.typ, see the trip test manual (tripman.tex).