COMMENT ⊗   VALID 00013 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00003 00002	[This file indicates changes considered and/or made to the
C00007 00003	* February 7 (more of Zapf's suggestions)
C00010 00004	* February 16
C00014 00005	* February 19
C00016 00006	* March 15
C00019 00007	* July 7 "Final changes" accumulated during "final revision" of volume 2
C00024 00008	* January 6, 1981
C00026 00009	* June 21-24 tuneup based on new photographic data
C00031 00010	Changes made during April visit of Richard Southall
C00033 00011	Things to be done:
C00037 00012	For next time:
C00039 00013	[transition to the new METAFONT, begun February 6, 1985]
C00040 ENDMK
C⊗;
[This file indicates changes considered and/or made to the
Computer Modern fonts since report CS780 was prepared.]

* January 28, 1980
1	The alpha is still too much like an italic a
2	The r and f get a little blurred at 252/inch resolution---add 1 pixel
3	Script K looks too much like an X
4	Lower case t correction to avoid hook shouldn't affect cmb10 etc.
* February 6 (the next big batch of changes were suggested by Hermann Zapf)
5	Xi bars lighter and serifs must not overlap
6	Right stem of M slightly thinner
7	Top of loop of Q to be raised a little
8	Less heavy stems on V and W
9	More open space at upper right of the bowl of an e
10	Bar a little longer on f, and make it the same weight as serifs
11	Stem of variant g to be a little higher, also italic g and a
12	Classic g: bulb link a bit higher, lower bowl lighter and more open
13	Stem of q to be a little higher in roman
14	Bar of t to be same weight as serif
15	Serifs of z to be longer and less slanted
16	More serif at left of right stroke in &
17	open the shoulder of the 6
18	make the 7 a quarter unit wider at each side
19	longer nose on the 1
20	arc of ? should be more superelliptical
21	thinner hairline on *
22	Make hachek and circumflex longer and thinner
23	The tilde needs to be stretch out too, also fix bug in angles
24	More length to the arrowheads of the arrow accent
25	Descender of cedilla to be a little lower
26	Less stem on A side of common stem of AE
27	More stem on E side of common stem of OE, also move E middle bar right a bit
28	Exclamation point not so tall
29	fl and ffl, make the l triangular at the top
30	Make Danish \o and \O less heavy, also empty set symbol
31	More weight to right of bowl in italic e, ae, oe
32	Italic f stem should stay thin longer; its bar line to be longer and thinner
33	Slightly more weight in the endv subroutine
34	Middle of italic w should not overshoot; also u and y right stems
35	Raise loop of italic &, and give more weight to its t bar
36	Italic 4 and 7 a bit wider; give hint of loop at left of the 4
37	Raise middle of italic 8 and make it slightly less heavy
38	Less weight in right arc of stem in lower case sigma and variant theta
39	More weight on bottom of lower case omega and variant omega
40	Redesign italic ae, the a to be distorted
* February 7 (more of Zapf's suggestions)
41	More weight at points 11 and 12 of Aleph
42	Move stem of Fraktur R .5u to the right
43	Small integral sign to be thinner at 4 and 6, also make these farther from 5
44	Crossbar of sterling sign to be .5u longer at the right
45	Script A more open (move 2,3,4,5,8 left by .5u)
46	Raise point 4 about .25u in script G
47	Script F should extend to right of point 2 a little (= European J otherwise)
48	Script V one unit wider; script W two units wider
49	Boldface a should not have a hook
50	German soft s: one unit thinner; the dot touches now
51	Italic ss: change as italic f, add a spur; more weight at pts 7, 9-11

* February 8
52	Tail of R should come oo below baseline
53	Change the partial sign to be like the 6
54	File symbol.mf split into two parts so it can be edited at PARC
55	Changed erasures in upper case Omega so that they don't need huge pens

* February 11
56	made hh a separate parameter for upper case height
57	changed halfd to pdd
58	added sserif subroutine and the ssd parameter
59	kern +.5pu between o's (Zapf)
60	ligature ft desirable (Zapf)
61	v and y are too wide
62	tail of t a bit more to the right (Zapf)
63	lower case a to be .25u to the right in its box (Zapf)
64	Left thin stem of N should be a bit to the right (Zapf)
65	Large square roots had bug (y2 never above baseline)

* February 15 (based on first Alphatype output)
66	added parameter for "crisp" serifs
67	must make serifs longer, they disappear on Alphatype and Xerox output
* February 16
68	new pas parameter for arm serifs
69	moved v and y slightly left, with more serif on the right
70	left bulb of ff not touching
71	longer tail on t
72	top of AE, make little straight part to help out the A (Zapf)
73	point 6 of the "5" should be a little lower
74	stem of 4 lighter, bar line a bit longer (Zapf)
75	oe a unit wider, makes the o more round
76	added dserif routine for more weight on light serifs
77	period and comma must be darker
78	Make C curve like G , one unit narrower (Zapf)
79	Uppercase overshoot is more than lower case (Zapf)
80	Digits not as tall as letters in cmr fonts
81	6 and 9 points 5,6,10 moved off center
82	Middle arm of E to be a little longer (Zapf)
83	Longer serifs at base of T and I (Zapf)
84	Make 3 wider and its upper dot heavier (Zapf)
85	Bar line of E and H should be slightly above center (Zapf)
86	Sans-serif C to curve more
87	italic f to have more curve, less straight part (Zapf)
	[this took about 6 hours to get right and to change the ligatures]

* February 17
88	italic sharp s 1u wider and modified as the f; also add another bulb (Zapf)
89	italic x should be more rounded (Zapf)
90	italic ae too dark in middle
91	more weight to stems of italic 4 and 7 (Zapf)
92	condense cmti fonts to 92 per cent of their present set width (Zapf)
93	italic 0 a little wider
94	bold f and r should let their bulbs stick out more (Zapf)
95	script U, V, and Y are too narrow (Zapf)
96	more weight at endpoints of script V, T, X, L (Zapf)

* February 18
97	Redesign script M and N to be more neutral style (Zapf)
98	In typewriter style, the e is too wide and the o is too narrow
99	Use thinner pens on fixed-width M, W, and w. (Zapf)
100	In sans-serif styles: A, M, N, V, W strokes not to touch at point (Zapf)
101	bar lines in f and t should be drawn with hpen, not cpen
102	a little letter spacing in the small caps font (Zapf)
103	bulbs in j and J to be further left in boldface versions
* February 19
104	Slope in 5 to be sharper
105	use only .5del in y
106	move the 4 .25u left in its box
107	adjust for interference in serifs of k, m, w

* March 5
108	additional corrections for low resolution in fontbegin
109	cmtitl required changes to variant g, and to 9 and ?, to avoid touching
110	bug in Upsilon, bot0 changed to bot6

* March 8
111	Fraktur R, need to reinforce diagonal after erasure
112	Stem on A is too heavy
113	ft ligature looks too crowded at top
114	Upper part of 3 shouldn't be an ellipse
115	The 6 and 9 still weren't right; more fiddling
116	5 too crowded at top
117	More serif at sidebars of W

* March 11
118	Made ff more like ffi in certain cases
119	Typo in pretzels (20/9 for 25/9)
120	Bottom of pretzels had 4..5 instead of 1..4
121	Lower case a slightly wider at point 10

* March 13
122	Need to round the points at the tips of arms
123	Arms of T made sharper (.25u instead of .5u)
124	Slightly more weight to stem of x
125	Sheared serifs at the mean line to have raised stems
126	Sheared serifs on u instead of serifs; also on i parts of fi ligatures
127	Large summation signs not so heavy at bottom
128	Large product signs with more bracketing on the serifs
* March 15
129	Have to draw stem before the sheared serif in b
130	Link comes too high in the g

* March 20
131	Adjust vertical positioning of large integral signs

* March 21
132	serifs on 2 and 7 to be like ucs not like as

* March 22
133	v is .5u less at right when there are serifs
134	more kerning between v and a

* March 27
135	not so much spacing around display operators
136	adjust math font spacing
137	lower case lambda, the join point should be lower
138	raise the lower bulb of the 3
139	lower case pi to be one unit wider,.5 unit more space in middle
140	upper case xi to be thicker on sans serif
141	the t should curl more in cmssb
142	bar of 7 and 2 is too heavy in cmssb
143	w and v too heavy in cmssb and cmtt, also shouldn't go below baseline
144	serifs too long on cmtt s
145	o too wide in cmtt
146	ear of g lost in boldface
147	cmti wants slightly thicker hairlines
148	longer lower arm on Z
149	changed arm to be thin for half its length
150	dotless italic math j had wrong stated height, also variant phi

* March 29
151	middle bar of E and H was too high in sans-serif styles
152	middle bar of E too short in sans-serif styles
153	added overcorr parameter to the various modes
154	fixed inconsistent overshoots on numerals

* April 4
155	ff less exaggerated

* June 3
156	meet and join in cmathx were interchanged
* July 7 "Final changes" accumulated during "final revision" of volume 2
157→134	kern -1pu between va and ka (-1.5pu was too much)
158	Less serif on I
159	Slightly less slope on / stroke of 8, but its endpts further from center
160	Bowl of 6 is rounder
161	ff still more symmetrical
162	More kerning after W and D; also before W
163	Less kerning between T and y in serif fonts
164	D one unit narrower, but the bowl only .5u narrower
165	math italic d is .5 unit narrower
166	lower case x not symmetrical when thick ≠ thin
167	ft ligature to be kerned like t
168	Polish l accent is wider and more horizontal
169	Removed old point "9" from letter a; it wasn't necessary or quite right.
170	diagonal stroke of z slightly heavier
171	macron accent longer, slightly darker; breve slightly darker to match
172	lower case alpha made one unit wider and more distinguishable from italic a
173	script G and Y have descenders
174	script T has longer bar
175	script X and Z not to be crossed in the middle
176	lowercase sigma narrower bowl, longer bar
177	dollar sign was funny looking; made it narrower and more symmetrical
178	top of lowercase t to be governed by top of pen, not middle of pen
179	tail of script Q to be lower, so the character isn't so dark
180	large parentheses moved slightly "inwards"

* September 15
181	Corrected oversight causing positive kern between "ho", "mo", "no"

* November 23
182	Depth of uppercase Theta should have been 0

* November 26
183	Typo in uppercase Delta, should say bot9 not bot8
184	Upper bowl of lowercase a to be drawn by arc subroutine
185	Use good6 for middle bar of E and F
186	Serif subroutine had ddraw that wasn't filling properly
187	Added crsbreak ∞ to characters that are built up
188	Adjusted upper right and lower left strokes on sans-serif S and s
189	Typo in varchar specification of extensible square root sign

* November 28
190	Point 4 above point 3 in K and k, a little less inner serif in k.
191	6, 9, and partial: less weight on the stem
192	Fixed bug in computation of serifs on w and W (del removed)
193	Lowercase r will be wider if w3<1.5u
194	7 should not be influenced by the x-height
195	Smaller dots in exclamation point and question mark
196	Fixed italic correction on colon and semicolon
197	Avoided serif fillin in h and n
198	Italic q made 9 units wide instead of 10 (so it's like p and the others)

* December 5
199	Used "ssdoo=min(ssd,oo)" for lower case sheared serifs, instead of ssd
200	Top serif of s to go up to the point of overshoot

* December 10
201	The new MF requires fontidentifier, ptsize, codingscheme attributes
202	Added Ramshaw's conventions for magnified fonts
203	Made new file for "caps and small caps"
204	Modifications to the stroke of the "2", maybe I can be happy with it now

* December 16
205	Script Q given less weight
* January 6, 1981
206	Small caps fonts to be heavier
207	positive kern when f or ff is followed by ' or ! or ?
208→203	maxht was wrong in the csc fonts

* January 13
209	Points at tips of arms should be "good" not rounded
210	The letter T should have forced left-right symmetry
211	Decreased bulb size on ff ligature

* January 20
212	Draw upper right serif first in u so it won't erase anything else
213	sserif routine needs to do ddraw not just draw

* January 22
214	first case of arm subroutine shouldn't compute a new point 2
215	program for t used "u" before calling charbegin

* January 28
216	typo in Q program, w9 should be w8

* March 22
217	increased "blacker" for Alphatype
218	decreased aspect ratio in cmr, cms, cmi

* March 30
219	increased stem widths in italic fonts

* April 5
220	made the "3" heavier

* April 13
221	correction for low resolution, aspect not to be substantially decreased
222	correction for low resolution vis a vis bulbs of f and r not too close

* May 26
223	Infinity made less lopsided in fixed-width styles
* June 21-24 tuneup based on new photographic data
224	decreased xheight
225	less contrast between roman and italic
226	lower case r is tangent at stem
227	lower case i and j have slightly bigger dots
228	numerals 0,1,3,4,6,8,9 bolder (from upper case not lower case pens)
229	`2' is bolder and wider, bulb is lower, upper right edge higher
230	`2' also has slope adjustment at point 10
231	`5' is bolder, upper part narrower, lower part wider
232	`7' is bolder, has serif at top left, base moves left a bit
233	`s' is half unit wider with less sidebar at left
234	`s' without serifs is full unit wider
235	new parameter added to arm subroutine, controls thickness of the arm
236	horizontal extensible braces: adjusted rounding to make them touch right
237	`T' is wider and thicker at top, has longer serifs at the bottom
238	letter spacing introduced on upper case as well as lower case of cmcsc
239	depth of `Q' had been set to zero
240	lower arm of `E' and `L' longer
241	upper arm of `E' and `F' repositioned
242	`R' is made narrower, its tail is cleaned up
243	extensible sqrt sign was too high at the top
244	sans serif `C' gets more curly
245	`D' made narrower, gets a normal bowl at the right
246	vector accent raised
247	serifs reduced on `x', also it gets narrower
248	kerning between P and a,e,o
249	"unbundling" of characters so that other character codes can be used
250	+, -, +/-, -/+ made smaller
251	upper left arm of `z' and `Z' ends a little sooner
252	corrected bug on `≠', the diagonal wasn't axially symmetric
253	added SAIL character set driver
254	added special case of `↔' for fixwidth

* July 8
255	ft ligature made slightly lighter

* July 11 (after seeing Alphatype proofs of the above)
256	all ligatures (except ae,oe) take account of letterspacing now
257	upper bowl of a less open and lighter
258	longer serifs on lower inside of x
259	decreased aspect rations on bold and small size fonts
260	raised upper dot and turning point on 3
261	less kerning for case kkk
262	zero is less bold and a little wider
263	r is made wider when the bulb would otherwise be crowded

* July 13
264	cmssb, cmss8, cmsss8 made a trifle lighter

* July 20
265	letterspace parameter now affects texinfo (e.g. space between words)

* July 24
266	0 and 6 and 9 tuned up

* August 7
267	italic s made with smaller bulb at top

* October 2
268	added +1 in denominator of white-space ratio calculation in l.c. s
269	changed aspect in cmb5 from .8 to .7

* October 20
270	horizontal strokes in 2, 7, 8 are lightened in sans-serif case
Changes made during April visit of Richard Southall
(summarized briefly, as there are many of them)
New parameters introduced for slight corrections to weights.
Bowed serifs.
New pen thinner than hairline used for details that are otherwise too heavy.
Major revision of the hstroke routine: all one stroke, ddrawn.
Major revision of arc routine: ddrawn.
Letter o to avoid vertical symmetry except at low resolution.
Other letters like p to come as close to sidebar as the o.
Letter m at low resolution to have chardw for consistent sidebars at both sides.
Letters d and u get reverse sheared serif at the bottom.
... Oh well, I tried to do that for the first few hours, but there were
	thousands and thousands more changes, so I gave up recording them.

Here is what I changed since ftp-ing a backup copy of the files to PARC
on May 2: (See Canon proofs for a record of what was ftp-ed)
ROMANL, -.5uc removed from width of k,x; -uc removed from width of v,y.
ROMANU, theta introduced into arc calls in U. (This not in Dover fonts.)
Things to be done:
1. Adjust darkness of brackets, parentheses, braces (large sizes)---DONE
2. math italic U has too much sidebar at the right.---DONE
3. lower case m in cmr10 rounded badly on dover (too narrow)---DONE
4. @ sign to be simpler in fixed-width fonts---DONE
5. λ in cmsa10 to be slanted; font to become TeX layout, not SAIL.
6. ft ligature should not have the bars run together--no, ft is eliminated.
7. Spacing too wide between O and f in text italic--no, I was looking at $Of$!
8. sqrt signs to have baseline below the rule and height=thickness---DONE
9. make sure that extensible characters run through the same sequence---DONE
10. Coproduct and backslash into mathex font---DONE
11. new parameter for the dots, w3 to be just for bulbs.---DONE
12. kerning in math italic fonts: make ,./()!:; unslanted in ami10
	df (and d[others]?)
	and eliminate letterspace parameter in the / and \
NOTE: This may change some examples of Chapter 17!
13. y came out wrong (stem too oblique) especially in ambx10. why? [rt for lft] DONE
14. s is topheavy in ambx10. Should not use the e-height after all in the s.
15. Change character arrangement: \o ↔ \l, @→\O→\t→\vec; \vec→symbols? #,$ in.
	And in \tt, delete ∞ and ⊗ ....---DONE (differently)
16. Positive kern between (','') and (?,!).---DONE
17. Make extensible vertical arrows (single and double).---DONE
18. Brackets in cmtt shouldn't look like boxes.---DONE
19. en-dash and em-dash are a bit too low. Cf v3 p222---DONE
20. Check for kerning between ro, ci, ed? (Cf controlling, italicized)
	Also between n and ' in can't (acc to Lamport)
Note: Part of the problem was in low resolution; fixed it by changing
the setting of incx in charbegin.
21. Need varepsilon in math italic font. Also varsigma. Also varrho--DONE
22. \widehat,\widetilde accents to be put into amathx, in three sizes---DONE
23. Put big/ and big\ as extensible character with empty extender for \langle
	(the empty extender should have the correct width)--can't, no room
24. Sharp, flat, and natural into math italic.--DONE
25. Wreath product and proportional sign and nearrow etc in symbols font--DONE
26. Make sure the = blends exactly with double arrows.---DONE
27. The bigg parentheses are too dramatic (see manual proofs).
28. Improve the script L.---DONE
29. Barline on f: why not more to left?---Tuning DONE
30. Polish slashed ell: modified. This changes \hbar in plain.tex!--DONE
31. Minus sign to be a little narrower so that it "looks" as wide as the plus.DONE
For next time:
There's a glitch on cap G in the Dunhill-like font PARTY.MF
And besides, PARTY.MF, XMAS.MF, PARTYI.MF generate bad fonts!

From ARK (12/12/83):The arms in the + are shorter than the arms for the
times.  This is because the code for + uses lft10 and rt10, which the code
in times assumes and uses good10 for placement of points.  I discovered
this when using your code as a base for a box and a diamond.

sym.mf[mf,sys]/7p "bot6y3" should be "bot10y3"!

The psmid parameter affects lower case s only; upper case needs
a correction too (e.g. when making an extremely condensed font).

Need kern in names like Kafka, Korfhage -- no, that made it worse

Dotless i under \u doesn't look right? (see Tchaikovsky in TeXbook index)
[transition to the new METAFONT, begun February 6, 1985]
After a long hiatus, I converted the old sources to MF84, silently
making lots of revisions as I went. The file CM85.BUG shows changes that
were made after I got the first full set of proofs.