HISTORY of the (unsupported) concmath package. ============================================== by Ulrik Vieth (vieth@xerxes.thphy.uni-duesseldorf.de) -- 1995/02/28 Hi folks! Here is something I hacked together recently when I was preparing a poster presentation for a conference. There may be room for improvements and the documentation is incomplete, but nevertheless I wanted to make it available, as it might be useful for others. * Some history how it started ... When I was preparing a poster, I was unhappy about the output with Computer Modern (CM) fonts turning out so extremely thin at high magnifications, and PostScript fonts weren't an alternative either because of the lack of matching math fonts. Therefore, I decided to give the Computer Concrete (CC) fonts a try. As they were originally developed by DEK to be used in conjunction with the AMS Euler fonts (see ``Typesetting Concrete Mathematics'' in TUGboat 10#1), they don't (well until now they didn't) come with math fonts either, but it turned out that it was easy to make them as the CC fonts are technically based on CM. * ... and the recipe how I did it Here is the recipe how I went about to create Concrete math fonts. I haven't done any fine tuning whatsoever, but the results seem OK. Of course, the real MF gurus are invited to improve on this... First, I had a look at ccmi10.mf and compared it to ccr10.mf. Despite its name, it is not a math font but rather a text font (with math_fitting:=false) containing the old-style numbers and some funny accents. Concerning the MF parameters there are only a few differences such as the font identifier, the slant parameter (slant:=.25), the italic-style lowercase g (variant_g:=true) and the call of a different driver file (generate mathit). To make it a real math font, all I did was to add `math_fitting:=true' and call the result xccmi10.mf to make the distinction clear. Likewise, I proceeded to make xccmi[5-9].mf by applying similar changes to ccr[5-9].mf. Next, I turned to the math symbol font. A comparison of cmmi10.mf and cmsy10.mf revealed that the only difference apart from the font identifier and the different driver file (generate mathsy) are the additional fontdimen parameters needed for a \textfont2. Thus, all I did to make xccsy10.mf was to copy xccmi10.mf, apply the usual changes, and insert the fontdimen code from cmsy10.mf. Likewise, I proceeded to make xccsy[5-9].mf from xccmi[5-9].mf and ccsy[5-9].mf. I'm not really sure if the fontdimen formulas from cmsy are fully appropriate for ccsy, but they are certainly better than using the values from cmsy.tfm calculated using CM parameters instead of CC parameters. Now you can guess how to make a math extension font. Again I did a comparison of cmex10.mf with cmmi10.mf and then with cmr10.mf. It turned out that cmex is in fact based on cmr except for the different font identifier and the fondimens for a \textfont3. Thus, to make xccex10.mf, I copied ccr10.mf, applied the usual changes, and copied the fontdimen code from cmex10.mf. That's it. This completes the set of Concrete math fonts. Finally, I also tried to make Concrete versions of the AMS symbol fonts, which I called xccam10.mf and xccbm10.mf. Again, I did a comparison with cmsy10.mf, noted the differences, took a copy of ccsy10.mf and applied the changes. For xccam10.mf I used the lower.parameters from cccsc10.mf for the circled R and S. For xccbm10.mf nothing special was needed. I didn't care to make Concrete versions of msam[5-9].mf and msbm[5-9].mf, but it shouldn't be difficult to make them. Note that there were some inconsistencies between the 10pt versions and the smaller sizes regarding the slant parameter, but that seems to be fixed now in version 2.2 of the AMS fonts released a few weeks ago. As you've seen, it doesn't take an expert METAFONT craftsman to make Concrete variants of math fonts. Some common sense and some familiarity with the organization of the parameter and driver files of the Computer Modern font family is enough. Unfortunately, the story ends here. For anything else that's more than just another CM variant you're out of luck, I'm afraid.