G. Branden Robinson
2020-08-28 11:01:26 UTC
URL:
<https://savannah.gnu.org/bugs/?59031>
Summary: the \[Im] and \[Re] escapes are wrong
Project: GNU troff
Submitted by: gbranden
Submitted on: Fri 28 Aug 2020 11:01:24 AM UTC
Category: Core
Severity: 3 - Normal
Item Group: Incorrect behaviour
Status: Need Info
Privacy: Public
Assigned to: gbranden
Open/Closed: Open
Discussion Lock: Any
Planned Release: None
_______________________________________________________
Details:
Blackletter/fraktur fonts are NOT generally used for the names of standard
number sets (naturals, integers, rationals, reals, complex numbers). Instead,
double-struck letters are typically used for this purpose. And in turn,
double-struck and blackletter faces are not the same.
See <https://mathworld.wolfram.com/Doublestruck.html>.
Given that:
I submit that at least one of the following is true.
1. The descriptions of these glyph names in groff_char(7) is wrong.
\[Im] \e[Im] Ifraktur u2111 Gothic I, imaginary
\[Re] \e[Re] Rfraktur u211C Gothic R, real
2. Their association with the Ifraktur and Rfraktur names in the Adobe Glyph
List (AGL) is incorrect.
3. The glyph names themselves are wrong to imply idiomatic usage as indicators
for the sets of real and imaginary numbers. (The complex field ℂ is more
commonly referred to than the set of imaginaries, which, being a 1-space
anyway, can be represented by simply by iℝ (where i is the imaginary
unit).
Unfortunately, this problem goes all the way back to the beginning.
^351da0dc doc/chars.tr (James Clark 1991-06-02 04:20:34 -0500 564)
.C2 Im Ifraktur "Fraktur I, imaginary"
^351da0dc doc/chars.tr (James Clark 1991-06-02 04:20:34 -0500 565)
.C2 Re Rfraktur "Fraktur R, real"
A variety of solutions suggest themselves.
A. Retire (delete, unsupport) the glyph names and let the user use
Unicode-form special character escapes if they need these symbols. As they
already do if they want the correct glyph for the reals.
B. Retain the mnemonics of the glyph name, remapping \[Im] and \[Re] to
U+1D540 and U+221D, respectively. Note that the former is outside the Basic
Multilingual Plane. This means dropping the AGL associations and the "Gothic"
part of the glyph descriptions.
C. Retain the mappings but rename them to be less misleading. We could go
ahead and use long glyph names for this purpose. I share the general
consensus expressed on the groff list in recent years that the groff glyph
list should be frozen, but this is an outright error and should be corrected.
D. Continue to be wrong and further cede the field of mathematical typography
to TeX.
Solutions A, B, and C would merit NEWS items. Solution D requires only an
admission of error in the groff_char(7) man page. And perhaps a suggestion
for .char remappings that one can add to one's documents.
groff mavens--what say you?
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?59031>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
<https://savannah.gnu.org/bugs/?59031>
Summary: the \[Im] and \[Re] escapes are wrong
Project: GNU troff
Submitted by: gbranden
Submitted on: Fri 28 Aug 2020 11:01:24 AM UTC
Category: Core
Severity: 3 - Normal
Item Group: Incorrect behaviour
Status: Need Info
Privacy: Public
Assigned to: gbranden
Open/Closed: Open
Discussion Lock: Any
Planned Release: None
_______________________________________________________
Details:
Blackletter/fraktur fonts are NOT generally used for the names of standard
number sets (naturals, integers, rationals, reals, complex numbers). Instead,
double-struck letters are typically used for this purpose. And in turn,
double-struck and blackletter faces are not the same.
See <https://mathworld.wolfram.com/Doublestruck.html>.
Given that:
I submit that at least one of the following is true.
1. The descriptions of these glyph names in groff_char(7) is wrong.
\[Im] \e[Im] Ifraktur u2111 Gothic I, imaginary
\[Re] \e[Re] Rfraktur u211C Gothic R, real
2. Their association with the Ifraktur and Rfraktur names in the Adobe Glyph
List (AGL) is incorrect.
3. The glyph names themselves are wrong to imply idiomatic usage as indicators
for the sets of real and imaginary numbers. (The complex field ℂ is more
commonly referred to than the set of imaginaries, which, being a 1-space
anyway, can be represented by simply by iℝ (where i is the imaginary
unit).
Unfortunately, this problem goes all the way back to the beginning.
^351da0dc doc/chars.tr (James Clark 1991-06-02 04:20:34 -0500 564)
.C2 Im Ifraktur "Fraktur I, imaginary"
^351da0dc doc/chars.tr (James Clark 1991-06-02 04:20:34 -0500 565)
.C2 Re Rfraktur "Fraktur R, real"
A variety of solutions suggest themselves.
A. Retire (delete, unsupport) the glyph names and let the user use
Unicode-form special character escapes if they need these symbols. As they
already do if they want the correct glyph for the reals.
B. Retain the mnemonics of the glyph name, remapping \[Im] and \[Re] to
U+1D540 and U+221D, respectively. Note that the former is outside the Basic
Multilingual Plane. This means dropping the AGL associations and the "Gothic"
part of the glyph descriptions.
C. Retain the mappings but rename them to be less misleading. We could go
ahead and use long glyph names for this purpose. I share the general
consensus expressed on the groff list in recent years that the groff glyph
list should be frozen, but this is an outright error and should be corrected.
D. Continue to be wrong and further cede the field of mathematical typography
to TeX.
Solutions A, B, and C would merit NEWS items. Solution D requires only an
admission of error in the groff_char(7) man page. And perhaps a suggestion
for .char remappings that one can add to one's documents.
groff mavens--what say you?
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?59031>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/