COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	So sorry, I should have FTPed your sources to here before fiddling around. Now
C00006 ENDMK
C⊗;
So sorry, I should have FTPed your sources to here before fiddling around. Now
BOTH of our files are wrong.
It seems my changes were much less extensive than yours, so I will list them here.
You can incorporate them into the PARC sources, and I shall FTP them to here when
I hear from you. (You may wish to break up the large files to BRAVO size too while
you're at it?)

Changes to TEXHDR: (I'll surround the changed lines by unchanged lines)
define otherchar=12 # none of the other character types;
define metaescape=13 # chars that get preceded by the escape character;
define parend=13	# end of paragraph;

define maxopcode=ifmode # the largest code number;
define charcodes=metaescape+1 # number of distinct codes allowed in chartype;
define texpars=10 # number of distinct parameters settable by \chpar command;

Changes to TEXSYN:
id←lop(curbuf) # remove first character;
comment The next two lines to be restored only if necessary! --DEK, Jan 1980;
comment if curbuf=0 and chartype('15)=carret
	then curbuf←'15 # don't remove the '15 at end of curbuf;
comment Note that in, e.g., \% the % should not be treated as a comment delimiter;

Finally, insert the following into the case statement after innerswitch in getnext:
	[midline+metaescape][skipblanks+metaescape][newline+metaescape] begin
	curbuf←escapechar&curchar&curbuf; go to innerswitch end;

The METAFONT sources still live on [maxc1]<knuth>. I need to correct a few things
in them (e.g., crsbreak has to maintain a list of breaks, not just make a single one,
also the sharp turn suppressed warnings sometimes refer to the wrong points in a
ddraw, also I want spens and epens to be remembered longer, etc.) In fact, I'm
planning to make one incompatible extension to the language: instead of using
hpenht to govern also the height of lpens and rpens, I will introduce lpenht and
rpenht. This seems a better choice and I can make it now before too many users
are affected.

I was going to hack these MF improvements on the weekend, but am traveling
unexpectedly to Washington DC! So it will be another few days before I get to them.