14

Recently I have been investigating some history of the mathbb(mathematical blackboard bold) letters of AMS, which can be called by invoking amsfonts or amssymb package.

By the content on page 35 of Handbook of Typography for the Mathematical Sciences, Donald Knuth does not like the mathbb letters, hence mathbb letters are not a standard part of TEX in 1979. Note that [KNU] refers to The TEXbook (Donald Knuth 1984).

enter image description here

While by the content on page 4 in User’s Guide to AMSFonts Version 2.2d, AMS rescue mathbb letters which were implemented in TEX in 1984.

enter image description here

So

  1. the mathbb letters should be created during 1979-1984?
  2. and what was the exact time for designing it and who designed it?
18
  • 1
    the older msxm and msym were written in the old metafont-in-sail, msam and msbm were re-implementations in what we now know as metafont, not sure who designed the originals. – David Carlisle Sep 4 '20 at 15:50
  • 2
    Yes I pinged Barbara in Chat, she was there at the time..... – David Carlisle Sep 4 '20 at 16:15
  • 1
    @David Carlisle It's better to give a detail and some references if possible. And I also have an advice that the LATEX team should write some matierals about the history of some important symbols or packages brought about. – M. Logic Sep 4 '20 at 16:19
  • 1
    @Mogic The first edition of “The Joy of TeX” (the manual for AMS-TeX) is of 1982. AMS-TeX had the blackboard bold letters in the msym fonts. However, they were quite different from (and better than, in my opinion) the letters in msbm that was issued as part of AMS-TeX 2.0 in 1990. The work for AMS-TeX 2.0 fonts, which include the Euler fonts, had been started earlier on. The Euler fonts have been designed by Hermann Zapf and digitized by the AMS. I'd not be surprised that Zapf has also something to do with the new blackboard bold. – egreg Sep 4 '20 at 16:57
  • 2
    @egreg -- I can assure you that Hermann Zapf had absolutely nothing to do with the msbm blackboard bold. In fact, I suspect he would have disliked them quite strongly. (I would have liked to know what he would have designed though. I don't think it would have matched the Monotype model.) – barbara beeton Sep 4 '20 at 19:30
14

When metal type composition was still in use, a "block style" blackboard bold font was available from Monotype. Much later, this style (and I believe the identical shapes) was used in the font charts from Unicode for the blackboard bold alphabet that appears in the plane 1 block of "Mathematical alphanumerics" (U+1D538-U+1D63B in https://unicode.org/charts/PDF/U1D400.pdf) except for a few letters that had been defined previously in the "letterlike symbols" block (https://unicode.org/charts/PDF/U2100.pdf, namely C, H, N, P, Q, R and Z).

The first public announcement of the AMS font containing a blackboard bold alphabet appeared in TUGboat 6:2 (1985) in the article "Mathematical symbols and Cyrillic fonts ready for distribution" (https://tug.org/TUGboat/tb06-2/tb12beet.pdf). A font chart of the egregiously misnamed euym contains the alphabet (uppercase only). A revised version of this article appeared in the next issue (https://tug.org/TUGboat/tb06-3/tb13beetcyr.pdf), with the euxm and euym fonts renamed to msxm and msym and an explanation of why the eu prefix was a bad choice. (The Euler fonts were designed by Hermann Zapf, who had no part in developing msxm or msym. To even hint at such a possibility was an insult.)

The blackboard bold letters in msym, although not shown in the revised announcement, are the same as those appearing in the first. They are blocky in appearance, somewhat similar to those in the Monotype blackboard bold, but of much lower quality. (It's no surprise that Knuth did not like them.)

With the advent of Metafont84, the "extra symbols" fonts were recreated, and a different style was used for the blackboard bold. As described in the quote that @Davislor has exhumed from comp.text.tex, a different style was requested from the designer (whose identity is unknown to me), namely that the shapes and weights of the letters be compatible with Times bold, as AMS was intending to set their books and journals with a variant of Times. (This was the font used on the Autologic APS-5; at that time, each typesetter manufacturer had their own proprietary version of most common fonts, since it was from fonts, not the equipment, that the manufacturers made their income.) At the time, this was considered novel, and although many authors are happy to use it, there are still strong feelings about what a blackboard bold font should ideally look like, as evidenced by the numerous packages on CTAN.

The announcement in TUGboat 10:3 (1989) (http://tug.org/TUGboat/tb10-3/tb25ams.pdf -- thanks to @egreg for finding this) appears to be the first mention that msbm is available. No additional information is provided by the user's guide for the amsfonts package, version 2.2d (January 2002) (http://mirror.ctan.org/fonts/amsfonts/doc/amsfndoc.pdf).

In my opinion, blackboard bold should blend with the principal fonts used in a document while remaining clearly distinct. Times does not really "match" Computer Modern, but neither does the Monotype style nor any of the various font packages posted on CTAN that I'm familiar with. Maybe it's time for a re-think.

4
  • I agree that mathbb of AMS used now doesn't match Computer Modern since the latter with some blending strokes is softer than the former. And I think mathbb of AMS used now is also not consistent with Times New Roman used popular because the latter has higher weight strokes. But I think it maybe consistent with Times with a lower line weight or other fonts with a similar style, and so it's also useful. – M. Logic Sep 5 '20 at 2:11
  • Maybe we can design a new mathbb with a similar way to define dsfont: ctan.org/pkg/blackboard . I mean use the same ways to define, not the same typeface, and either I don't like the dsfont a bit since its stroke weights are not balanced. – M. Logic Sep 5 '20 at 2:28
  • @Mogic New math fonts are, for the most part, OpenType, although it would also be possible to just make a double-struck decorative font and load it with \setmathfont[range=bb]{My New Font} in unicode-math. – Davislor Sep 5 '20 at 3:36
  • @Davislor I know that, and the fonts that I am using and whose otf files has been improved a lot by myself are invoked in this way. – M. Logic Sep 5 '20 at 3:40
11

Here’s what @barbarabeeton wrote about it back in 1994. I hope she’ll swing by here and fill us in.

michael downes (comp.text.tex, november 29) asks if i am sure about the model that was used for the blackboard bold in msbm.

in an earlier message, i said ... the model we used for the blackboard bold in the msbm font was a set of shapes we had been using previously in another system, created by a u.s. company. i am not sure specifically what they used for a model, but it would not surprise me if it were some traditional typographic in-line or open face typeface.

and michael points out, partly correctly, that ... I overheard at the time that the so-called blackboard bold font in msbm is merely a disemboweled Times Roman bold, as that was precisely what the AMS asked of the font designer

ams did ask that the overall shapes and weights of the letters be compatible with times bold, since that is what we intended to use in our journal and book production. but if there was any doubt, the locations of the doubled strokes were to conform to what had been used by the typesetting system we had been using previously, a system developed by science typographers (sti), and which had been reimplemented in msym (in a not particularly refined manner; some strokes that should have been parallel were not, and there were other very bad flaws in that alphabet, though not in the original sti implementation. since i created the first few letters for the msym \Bbb -- using the sti openface as a strict model -- as part of my attempt to learn mf79 under knuth's tutelage, i am abundantly aware of their limitations).

i assure you that the doubled strokes in msym, in the sti openface font, and in msym are all in exactly the same places -- i checked all three from contemporaneous source documents before replying. i will be happy to produce the evidence, which is part of my permanent library.

as you see from my original statement, i don't know what sti used for a model -- it may very well have been an inline version of times or century -- but in fact we used that font at ams from probably about 1973 until about 1983, and the doubled lines were always in the same locations in each letter as what's in msbm, quite different from what's in bbold.

i too am sorry that there's not more commentary in the mf code, but since i installed the original sti font here at ams, and was somewhat involved with the development of both msym and msbm, i do know for sure what i've stated above about its provenance. -- bb

5
  • 7
    Good sleuthing. (I wouldn't even have thought of looking there.) I believe I never did know who actually did the coding of that alphabet, and in the absence of actual records I have no way of checking. But it is true that the model was Times of some ilk. The "more traditional" style, supplied in metal by Monotype, has been used as the model in the Unicode charts (U+1D538-U+1D63B in unicode.org/charts/PDF/U1D400.pdf with a few in the "Letterlike symbols" block). This style was explicitly not used for the model of \Bbb. – barbara beeton Sep 4 '20 at 19:21
  • @Mogic -- Although I was involved in the decision of what should go into these fonts and their testing, I did not arrange for their creation or guide the developer. Any Metafont work that I did was entirely experimental, and was not included in the final font. Taking into consideration the length of time needed to package a font for distribution, my best guess is that a test version of the msbm font mentioned in the 1989 announcement was available in 1988. – barbara beeton Sep 5 '20 at 2:03
  • @barbara beeton Can we have other ways to know the origination of mathbb of AMS used now, for example email to Knuth? – M. Logic Sep 5 '20 at 2:23
  • @Mogic -- I am pretty certain that correspondence with Knuth would shed no light on the origin of the AMS blackboard bold fonts. They were created entirely inhouse, and anyone involved is no longer at AMS. The original, blocky, fonts in msym can be compared with the blackboard bold in the Science Typographers system (I have a copy of their manual in an unmarked box, not easily accessible, but existing), but the origin of the "Times-like" shape in msbm seems undocumented in any available source. – barbara beeton Sep 5 '20 at 14:36
  • @barbara beeton So it is possible to become a puzzle for this...... – M. Logic Sep 5 '20 at 14:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.